Numerical and experimental optimization of peak power

reduction control strategies

Nikolaos Stathopoulos

To cite this version:

Nikolaos Stathopoulos. Numerical and experimental optimization of peak power reduction control
strategies. Thermics [[Link]-ph]. École Nationale des Travaux Publics de l’État [ENTPE],
2015. English. �NNT : 2015ENTP0001�. �tel-01559651�

HAL Id: tel-01559651

[Link]
Submitted on 10 Jul 2017

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
N° d’ordre 2015-ENTP-0001 Année 2015

THÈSE
pour l’obtention du grade de
Docteur de l’Ecole Nationale des Travaux Publics de l’Etat

Université de Lyon

Ecole doctorale : MEGA (Mécanique, Energétique, Génie Civil, Acoustique)

Spécialité : Génie Civil
Préparée au Laboratoire Génie Civil et Bâtiment (LTDS/UMR 5513)

Présentée et soutenue publiquement par

Nikolaos STATHOPOULOS

Optimisation numérique et expérimentale

de stratégies d’effacement énergétique

Soutenue le 27/02/2015 devant le jury composé de :

M. EL MANKIBI HDR Directeur de thèse

H. EL QARNIA Professeur Examinateur
G. FRAISSE Professeur Rapporteur
C. GHIAUS Professeur Examinateur
F. HAGHIGHAT Professeur Rapporteur
P. MICHEL HDR Co-directeur de thèse
2
Abstract
Numerical and experimental optimization of peak power reduction control strategies
Nikolaos Stathopoulos, 2015

Keywords:
Building, Energy, Peak power reduction, Control strategies, Phase Change Materials, Thermal Storage,
Heat exchanger, Thermal Simulation, Thermal Comfort, Indoor Air Quality

Considering the current French energy context, two major challenges are emerging. In the short term,
significant peak power consumption has been observed in the past few years during the winter season.
These peaks are strongly linked to electrical space heating and have important economic, environmental
and social implications. In the long term, ambitious environmental goals have been set at national and
European levels, requiring thermal storage technology and efficient management of the built
environment. As part of the solution, Phase Change Materials (PCM) and heat exchanger applications
offer promising results through thermal storage and load shifting techniques.
Within this framework, the objective of this thesis is to develop load shifting solutions which also take
into account the thermal comfort of the occupants and the indoor air quality. To achieve this, two tools
were necessary: an experimental heat exchanger unit (prototype) and a numerical model that accurately
simulates its behavior.
The exchanger contains macroencapsumated PCM (paraffin) and is conceived in a way that facilitates its
integration in a ventilation system. It is aimed to shift space heating electrical consumption from peak to
off-peak period. The unit was experimentally characterized, using an important amount of sensors
through full thermal cycles (charging and discharging) and was coupled to an experimental test cell,
which led to the testing of preliminary control strategies.
The numerical model is based on the heat balance approach and the apparent heat capacity method,
using finite differences for differential equation solution under Matlab/Simulink environment. After
validation with experimental data, the model was used to optimize the performance of the exchanger.
Several parameters were investigated, including heat exchanger dimensions, PCM quantity and
properties, seeking the configuration with the optimal compromise between stored heat and the time
needed for the charging / discharging process.
The numerical model was coupled to a building simulation model and an 80m2 dwelling was conceived
for control strategies implementation and evaluation, by investigating different scenarios over a one-
month winter period. The scenarios vary with increasing complexity, first considering load shifting and
thermal comfort, then adding the final price of electricity consumption and finally taking into account
the indoor air quality with the presence of a four-person family.
This study has been conducted within the framework of a project funded by the French National
Research Agency (Stock-Air: ANR-Stock-E) and was also financially supported by the French Ministry of
Sustainable Development.

3
4
Résumé
Optimisation numérique et expérimentale de stratégies d’effacement énergétique
Nikolaos Stathopoulos, 2015

Mots clés:

Bâtiment, Energie, Effacement énergétique, Stratégies de contrôle, Matériaux à changement de phase,

Stockage thermique, Échangeur Thermique, Simulation thermique, Confort thermique, Qualité de l’air
intérieur

Dans le contexte énergétique français actuel, deux principaux enjeux émergent. À court terme, des
pointes de consommation électrique croissantes sont observées les dernières années pendant la période
hivernale. Ces pointes sont fortement liées au chauffage électrique et ont des conséquences
économiques, environnementales et sociales importantes. Dans un long terme, des objectifs
environnementaux ambitieux ont été fixés au niveau national et européen, nécessitant la technologie de
stockage thermique et une gestion efficace de l'environnement bâti. Les Matériaux à Changement de
Phase (MCP) ainsi que les dispositifs de type échangeurs thermiques offrent des résultats promettant
grâce au stockage thermique et le déplacement des consommations.
Dans ce cadre, l’objectif de cette thèse est de développer des solutions de déplacement des
consommations énergétiques qui prennent en compte le confort thermique des occupants et la qualité
de l’air intérieur. Pour ce faire, deux outils sont nécessaires: un échangeur thermique expérimental
(prototype) et un modèle numérique capable de simuler son comportement.
L'échangeur contient du MCP macroencapsulé (paraffine) et est conçu de manière à faciliter son
intégration dans un système de ventilation. Il a comme but de décaler la consommation due au
chauffage électrique vers la période hors pointe. Le dispositif a été caractérisé expérimentalement lors
des cycles thermiques complets (charge et décharge) en utilisant une quantité importante de capteurs. Il
a ensuite été couplé à une cellule expérimentale, afin de tester des stratégies de contrôle préliminaires.
Le modèle numérique est basé sur la discrétisation spatiale et l’établissement du bilan de chaleur des
couches considérées, la méthode de la capacité thermique apparente, ainsi que l’utilisation des
différences finies. Après validation à l’aide des données expérimentales, le modèle a été utilisé pour
optimiser la performance de l'échangeur. Plusieurs paramètres ont été étudiés, y compris les dimensions
de l'échangeur, la quantité et les propriétés du MCP, en cherchant la configuration avec le compromis
optimal entre la chaleur emmagasinée et le temps nécessaire pour la charge et la décharge.
Le modèle numérique a été couplé à un modèle de simulation du bâtiment et un logement de 80m2 a
été conçu pour la mise en œuvre et l'évaluation des stratégies de contrôle, en investiguant différents
scénarios sur une période hivernal d’un mois. Les scénarios varient avec une complexité croissante,
d'abord en considérant l’effacement énergétique et le confort thermique, ensuite en ajoutant le prix
final de la consommation électrique et enfin en prenant compte la qualité de l'air intérieur avec la
présence d'une famille de quatre personnes.

5
Cette étude a été menée dans le cadre d'un projet financé par l'Agence National de la Recherche (Stock-
Air: ANR-Stock-E) et a également été soutenu par le ministère de l'Ecologie, du Développement durable
et de l'Energie.

6
Acknowledgments

First, I would like to thank M. El Mankibi for his patient and constant guidance during the last three years
and for providing a healthy environment for the accomplishment of this PhD dissertation. I would also
like to thank M. Michel for the same reasons. Special thanks to Riccardo Issoglio for his great help and
advice on all aspects of the experimental part of the study. I acknowledge the financial support from the
French National Research Agency (ANR) and the fruitful collaboration with the members of the research
project Stock Air.

I am also grateful to Professors Haghighat and Fraisse for accepting to examine this work and participate
in the defense jury. My sincere gratitude also goes to Professors El Qarnia and Ghiaus for their
willingness to participate in the jury.

Special credits to Henri Odin for his numerous corrections and advice on the proper English writing of
this manuscript.

Thanks are also due to all members of the LGCB laboratory for their support and all the joyous moments
I spent around them.

My heartfelt gratitude goes to my father, mother and sister for providing me with excellent education
conditions and supporting my choices throughout my life. Without them I would not have come so far.

Lastly, I would like to thank my friends in Athens, Lyon and wherever they may now be living for always
being there for me, for their constant positive influence and advice and making the world go round.

7
8
Contents

General introduction ........................................................................................................................ 24

Context .................................................................................................................................................... 24
Objectives and approach ......................................................................................................................... 24
Chapter contents ..................................................................................................................................... 25
Chapter 1 : Energy context and built environment ............................................................................. 28
1.1 Introduction....................................................................................................................................... 28
1.2 At the international and EU levels..................................................................................................... 28
Built environment: a big energy consumer ......................................................................................... 28
1.3 At the French level ............................................................................................................................ 29
1.3.1 Energy context and policies........................................................................................................ 29
1.3.2 Built environment and energy consumption ............................................................................. 29
1.3.3 Electrical heating ........................................................................................................................ 32
1.3.4 Peak power demand and thermo-sensitivity of French electrical consumption ....................... 34
1.3.5 Peak power demand implications .............................................................................................. 36
1.3.6 Applied electricity load-shifting methods .................................................................................. 39
1.4 Conclusion ......................................................................................................................................... 39
Chapter 2 : Energy storage, PCM and HVAC integrated applications ................................................... 41
2.1 Introduction....................................................................................................................................... 41
2.2 Energy storage: necessity and methods............................................................................................ 41
2.2.1 Thermal energy storage ............................................................................................................. 42
2.2.2 Electrical energy storage ............................................................................................................ 47
2.3 Phase Change Materials .................................................................................................................... 47
2.3.1 Classification ............................................................................................................................... 49
2.3.2 Phase change phenomena and properties................................................................................. 50
2.3.3 Containment ............................................................................................................................... 51
2.3.4 PCM application in buildings ...................................................................................................... 53
2.3.5 PCM enhancement techniques .................................................................................................. 55
2.3.6 Limitations/barriers .................................................................................................................... 56

9
 2.4 HVAC integrated PCM applications ................................................................................................... 56
2.4.1 Demonstration examples ........................................................................................................... 56
2.4.2 Comparative table of existing studies ........................................................................................ 72
2.5 Design and operation of active LHS ventilation systems .................................................................. 73
2.5.1 Main elements of an active latent heat storage system ............................................................ 73
2.5.2 Operation principles ................................................................................................................... 73
2.5.3 Design philosophy of a LHS system ............................................................................................ 74
2.5.4 Design Methods.......................................................................................................................... 75
2.6 Conclusion ......................................................................................................................................... 75
Chapter 3 : Development of an experimental PCM-Air heat exchanger .............................................. 78
3.1 Introduction....................................................................................................................................... 78
3.2 Experimental set up description ....................................................................................................... 78
3.2.1 Thermal energy storage as a solution ........................................................................................ 78
3.2.2 Elaboration of a PCM-Air heat exchanger .................................................................................. 79
3.2.3 Experimental platform ............................................................................................................... 84
3.2.4 Instrumentation, data acquisition and control-command interface ......................................... 85
3.2.5 Operation principle and associated physical phenomena ......................................................... 90
3.2.6 Main components of the developed experimental system ....................................................... 91
3.2.7 Coupling with an experimental test cell ..................................................................................... 91
3.3 Characterization study ...................................................................................................................... 92
3.3.1 PCM temperature evolution curves ........................................................................................... 92
3.3.4 Characterization results ............................................................................................................. 94
3.3.2 PCM leakage ............................................................................................................................... 96
3.3.3 Ameliorated plates ..................................................................................................................... 96
3.3.3 Verification tests......................................................................................................................... 98
3.3.4 Characterization results ........................................................................................................... 100
3.4 Coupling with an experimental room and discussion ..................................................................... 103
3.5 Conclusion ....................................................................................................................................... 104
Chapter 4 : Numerical model development ..................................................................................... 106
4.1 Introduction..................................................................................................................................... 106
4.2 State of the art of existing models .................................................................................................. 106
4.3 Numerical model development: Heat balance approach and apparent heat capacity method .... 107

10
 4.3.1 Matlab/Simulink and S-function choice ................................................................................... 107
4.3.2 Heat balance approach............................................................................................................. 108
4.3.3 Apparent heat capacity method............................................................................................... 109
4.3.4 Nodal discretization and application of the heat balance approach ....................................... 109
4.3.5 Energy balance equation leading to matrix formulation ......................................................... 110
4.3.6 Model assumptions .................................................................................................................. 123
4.4 Numerical model calibration, specific heat capacity importance ................................................... 123
4.4.1 DSC Method.............................................................................................................................. 123
4.4.2 Creation of adequate heat capacity values .............................................................................. 127
4.4.3 First approach: isosceles triangle ............................................................................................. 127
4.4.4. Second approach: two peaks curve for the discharge phase .................................................. 129
4.4.5. Third approach: 2 separate triangles for the discharge phase ............................................... 131
4.4.6 Final approach: heat capacity values for different positions along the PCM layer.................. 133
4.5 Numerical model validation ............................................................................................................ 134
4.6 Coupling with existing building model ............................................................................................ 138
4.6.1 Building’s model: Hybcell 1.2 ................................................................................................... 138
4.6.2 Definition of the experimental test cell ................................................................................... 138
4.6.3 Reproduction of an experimentally tested scenario and comparison ..................................... 140
4.7 Conclusion ....................................................................................................................................... 143
Chapter 5 : Heat exchanger optimization and control strategies implementation ............................. 145
5.1 Introduction..................................................................................................................................... 145
5.2 Conception of a low energy house .................................................................................................. 145
5.2.1 Materials and properties of the house..................................................................................... 145
5.2.2 Climate definition ..................................................................................................................... 146
5.2.3 Heating needs of the house ..................................................................................................... 147
5.3 Testing of the initial exchanger ....................................................................................................... 149
5.4 Optimization first approach ............................................................................................................ 153
5.4.1 Definition of the optimization parameters .............................................................................. 153
5.4.2 Testing protocol........................................................................................................................ 154
5.4.3 Selection of three configurations ............................................................................................. 154
5.4.5 First scenario: simple load shifting ........................................................................................... 156
5.4.6 Second scenario: charging of the heat exchanger during the night ........................................ 159

11
 5.4.7 Third scenario: occupancy and indoor air quality .................................................................... 163
5.4.8 Conclusions of the first approach............................................................................................. 169
5.5 Optimization second approach ....................................................................................................... 169
5.5.1 Introduction.............................................................................................................................. 169
5.5.2 Definition of the modified parameters .................................................................................... 170
5.5.3 Testing of the modified parameters ......................................................................................... 171
5.5.4 Results of the second optimization approach .......................................................................... 174
5.5.3 Simple scenario, second approach ........................................................................................... 175
5.5.4 Night charging scenario ............................................................................................................ 180
5.5.5 Occupancy scenario .................................................................................................................. 184
5.6 Conclusions of the second approach............................................................................................... 189
Conclusive summary and future work ............................................................................................. 191
Synopsis ................................................................................................................................................. 191
Perspectives .......................................................................................................................................... 192
Texte de synthèse (français) ........................................................................................................... 195
References ..................................................................................................................................... 220
Appendices .................................................................................................................................... 227

12
List of figures

Figure A.1: The methodology followed along this thesis ............................................................................ 26

Figure 1.1: Evolution of energy consumption per sector ............................................................................ 30
Figure 1.2: Evolution of CO2 emission per sector ........................................................................................ 30
Figure 1.3: Final consumption of the residential/tertiary sector ................................................................ 31
Figure 1.4: Evolution of final electricity consumption per sector ............................................................... 31
Figure 1.5: Energy distribution within the residential sector...................................................................... 32
Figure 1.6: Percentage of electrical heating in newly built housing ........................................................... 33
Figure 1.7: Peak power demand around 19:00, before and during the cold wave of February 2012 ........ 35
Figure 1.8: Electrical consumption and equivalent thermo-sensitivity part, first half of February 2012 ... 36
Figure 1.9: Demand coverage during the cold wave of February 8th 2012 ................................................ 37
Figure 1.10: Implications of peak power demand....................................................................................... 38
Figure 2.1: Internal and external solicitations in a building, possible heat sources ................................... 42
Figure 2.2: Temperature profile variation with stored heat evolution ....................................................... 45
Figure 2.3: Classification and applications of EES (Fernandes et al. 2012) ................................................. 47
Figure 2.4: PCM temperature and heat flux evolution during solidification, 1: liquid phase 2: melting
phase 3: solid phase (Baetens et al., 2010) ................................................................................................. 48
Figure 2.5: Classification of PCMs (IEA, Annex 23, 2013) ............................................................................ 49
Figure 2.6: Classification of PCMs according to their meting temperature range and enthalpy (IEA, Annex
17, 2005)...................................................................................................................................................... 49
Figure 2.7: Subcooling effect during solidification (Cabeza et al. 2008) ..................................................... 50
Figure 2.8: Macroencapsulation solutions, from left to right: aluminum profiles with fins, coated
aluminum plate, capsule stripes (Cabeza et al., 2008)) .............................................................................. 52
Figure 2.9: PCM construction panels: microencapsulated PCM (left), shape stabilized plate (right) (IEA.
Annex 23) .................................................................................................................................................... 53
Figure 2.10: Borderon et al. heat exchanger unit ....................................................................................... 57
Figure 2.11: Outlet air temperature experimental and simulation results (Borderon et al.) ..................... 58
Figure 2.12: Zalba’s et al. experimental set up (left) and heat exchanger unit (right) ............................... 59
Figure 2.13: Zalba’s et al. unit’s heat transfer during phase change for different configurations ............. 59
Figure 2.14: Zalba’s et al. unit’s heat exchange during phase change for different configurations ........... 60
Figure 2.15: Heat exchanger developed by Arkar et al. .............................................................................. 61
Figure 2.16: Experimental and simulation confrontation for the outlet air temperature of Arkar’s et al.
unit during the melting stage (left) and the solidification stage (right) ...................................................... 61
Figure 2.17: Nagano et al. experimental unit .............................................................................................. 62
Figure 2.18: Temperature variation for different experimental configurations, Nagano’s et al. unit ........ 63
Figure 2.19: Lazaro et. al LHS system and first encapsulation method: aluminum pouches ...................... 64
Figure 2.20: Lazaro et. al LHS second encapsulation method (aluminum panels)...................................... 65
Figure 2.21: Cooling rate evolution with constant rise of inlet air temperature (left) and constant heating
power for Lazato’s et al. prototype 2 .......................................................................................................... 65

13
Figure 2.22: Dolado et al. modifications in Lazaro et al. system ................................................................. 66
Figure 2.23: Inlet and outlet air temperature, ambient air, and surface temperature distribution for a full
cycle of Dolado et al. proposed unit ........................................................................................................... 66
Figure 2.24: Air, plate and PCM temperatures of a single plate for a full cycle, Dolado et al. proposed unit
..................................................................................................................................................................... 67
Figure 2.25: Lin et al. experimental house and enhanced floor heating system. ....................................... 68
Figure 2.26: Lin et al. experimental vs simulation results for different configurations (a) indoor
temperature without air supply, (b) indoor temperature with air supply, (c) PCM surface temperature
without air supply and (d) PCM surface temperature with air supply........................................................ 69
Figure 2.27: Turnpenny et al. LHS unit, overall view (left) and heat exchange pipes (right) ...................... 70
Figure 2.28: Air, PCM and external temperature for 1 and 2 kW heat input, Turnpenny et al. LHS unit ... 71
Figure 2.29: Upstream-downstream and upstream-PCM temperature difference for 1 and 2 kW heat
input, Turnpenny et al. LHS unit.................................................................................................................. 71
Figure 3.1: Microtek 37D in solid form........................................................................................................ 79
Figure 3.2: Aluminum plates used for PCM containment ........................................................................... 80
Figure 3.3: Aluminum fin ............................................................................................................................. 81
Figure 3.4: Fins between the PCM plates .................................................................................................... 82
Figure 3.5: The heat exchanger during construction phase ........................................................................ 82
Figure 3.6: The heat exchanger through the assembly process.................................................................. 83
Figure 3.7: Electrical duct heater (left) and air diffuser (right) ................................................................... 84
Figure 3.8: Experimental platform .............................................................................................................. 85
Figure 3.9: PT100 temperature sensor........................................................................................................ 85
Figure 3.10: Nozzle Pitot Flow Sensor ......................................................................................................... 86
Figure 3.11: Insertion of temperature sensors inside the PCM .................................................................. 87
Figure 3.12: Air and PCM properties monitoring interface ......................................................................... 88
Figure 3.13: PCM and surface temperature monitoring interface.............................................................. 89
Figure 3.14: Data acquisition and command implementation interface .................................................... 89
Figure 3.15: PCM during liquefaction.......................................................................................................... 90
Figure 3.16: The experimental test cell Hybcell .......................................................................................... 92
Figure 3.17: Downstream PCM temperature evolution .............................................................................. 93
Figure 3.18: Downstream surface temperature evolution ......................................................................... 93
Figure 3.19: Inlet and outlet air temperature evolution ............................................................................. 94
Figure 3.20: Characterization results: outlet air temperature for varying airflow rates, charging phase .. 95
Figure 3.21: Characterization results: outlet air temperature for varying airflow rates, discharging phase
..................................................................................................................................................................... 95
Figure 3.22: PCM leakage ............................................................................................................................ 96
Figure 3.23: The improved aluminum plates used as PCM containers ....................................................... 97
Figure 3.24: The improved aluminum plates used as PCM containers, sketch........................................... 97
Figure 3.25: Deformation on containers linked to the extrusion process .................................................. 99
Figure 3.26: Average, maximum and minimum outlet surface temperatures for 14 plates (airflow rate of
300 m3/h)..................................................................................................................................................... 99

14
Figure 3.27: Characterization results without fins: outlet air temperature for varying airflow rates,
charging phase .......................................................................................................................................... 100
Figure 3.28: Characterization results without fins: outlet air temperature for varying airflow, discharging
phase ......................................................................................................................................................... 101
Figure 3.29: Characterization results with fins: outlet air temperature for varying airflow, charging phase
................................................................................................................................................................... 101
Figure 3.30: Characterization results with fins: outlet air temperature for varying airflow, discharging
phase ......................................................................................................................................................... 102
Figure 3.31: Integration of AIR-PCM exchanger and HVAC system of Hybcell ......................................... 103
Figure 4.1: S-function block under Simulink environment (left) and parameters definition block (right) 108
Figure 4.2: Schematic representation of the considered layers for temperature evolution calculation . 109
Figure 4.3: Upper view of the nodal representation of the layers: air (a), aluminum (p) and PCM (m)... 111
Figure 4.4 Centralized matrix equation ..................................................................................................... 122
Figure 4.5: Representation of the DSC method (Borderon, 2012) ............................................................ 124
Figure 4.6: DSC obtained heat capacity curves for three different heating/cooling rates ....................... 124
Figure 4.7: Simulation results for inlet and outlet air temperature using DSC obtained heat capacity
curves and comparison with experimental data ....................................................................................... 126
Figure 4.8:Simulation results for PCM temperature at node 5 using DSC obtained heat capacity curves
and comparison with experimental data .................................................................................................. 126
Figure 4.9: Fictive Heat capacity values, first approach ............................................................................ 127
Figure 4.10: Simulation results for inlet and outlet air temperature using first approach’s heat capacity
curves and comparison with experimental data ....................................................................................... 128
Figure 4.11: Simulation results for PCM temperature at node 5 using first approach’s heat capacity curves
and comparison with experimental data .................................................................................................. 128
Figure 4.12: Fictive Heat capacity values, second approach ..................................................................... 129
Figure 4.13: Simulation results for inlet and outlet air temperature using second approach’s heat capacity
curves and comparison with experimental data ....................................................................................... 130
Figure 4.14: Simulation results for PCM temperature at node 5 using second approach’s heat capacity
curves and comparison with experimental data ....................................................................................... 130
Figure 4.15: Fictive Heat capacity values, third approach ........................................................................ 131
Figure 4.16: Simulation results for inlet and outlet air temperature using third approach’s heat capacity
curves and comparison with experimental data ....................................................................................... 132
Figure 4.17: Simulation results for PCM temperature at node 5 using third approach’s heat capacity
curves and comparison with experimental data ....................................................................................... 132
Figure 4.18: Surface temperature difference per minute at different parts of the plate ......................... 133
Figure 4.19: Calibrated cp values created for the numerical model ......................................................... 134
Figure 4.20: Numerical and experimental results for inlet air temperature, for 100, 300 and 500 m3/h 135
Figure 4.21: Numerical and experimental results for outlet air temperature, for 100, 300 and 500 m3/h
................................................................................................................................................................... 135
Figure 4.22: Numerical and experimental results for surface temperature at node 1, for 100, 300 and 500
m3/h .......................................................................................................................................................... 136

15
Figure 4.23: Numerical and experimental results for surface temperature at node 5, for 100, 300 and 500
m3/h .......................................................................................................................................................... 136
Figure 4.24: Numerical and experimental results for surface temperature at node 10, for 100, 300 and
500 m3/h ................................................................................................................................................... 136
Figure 4.25: Numerical and experimental results for PCM temperature at node 1, for 100, 300 and 500
m3/h .......................................................................................................................................................... 137
Figure 4.26: Numerical and experimental results for PCM temperature at node 5, for 100, 300 and 500
m3/h .......................................................................................................................................................... 137
Figure 4.27: Numerical and experimental results for PCM temperature at node 10, for 100, 300 and 500
m3/h .......................................................................................................................................................... 137
Figure 4.28: Hybcell 1.2 model for building simulation ............................................................................ 138
Figure 4.29: Visual representation of the modeled experimental cell...................................................... 139
Figure 4.30: Composition and properties of Hybcell’s north-east wall..................................................... 139
Figure 4.31: North-east wall definition, including openings ..................................................................... 140
Figure 4.32: Coupling of the heat exchanger’s model with the building’s model Hybcell 1.2 under Simulink
environment .............................................................................................................................................. 141
Figure 4.33: Numerical and experimental results for the heat exchanger unit and the Hybcell room
coupling during a 4 days test..................................................................................................................... 142
Figure 4.34: Numerical and experimental results for the heat exchanger unit and the Hybcell room
coupling during a 4 days test..................................................................................................................... 142
Figure 5.1: The conceived mono zone house, as presented in Hybcell 1.2 model ................................... 146
Figure 5.2: Outdoor temperature for the tested month (January 2013) .................................................. 147
Figure 5.3: Indoor air temperature using conventional heating, setpoint of 20°C ................................... 148
Figure 5.4: Heating needs during 18:00-20:00 for the conceived house .................................................. 148
Figure 5.5: Performance of the initial exchanger, destocking 18:00-20:00, inlet exchanger temperature
38°C ........................................................................................................................................................... 150
Figure 5.6: Indoor temperature for different inlet exchanger air temperatures ...................................... 151
Figure 5.7: Indoor temperature for different load shifting periods, initial exchanger ............................. 151
Figure 5.8: Representation of the modified parameters (PCM and air layers, height of the unit) ........... 153
Figure 5.9: Indoor air temperature, 3,5 days testing, optimized exchanger 1.......................................... 155
Figure 5.10: Indoor air temperature, 3,5 days testing, optimized exchanger 2 ....................................... 155
Figure 5.11: Indoor air temperature, 3,5 days testing, optimized exchanger 3 ....................................... 156
Figure 5.12: Performance of the heat exchanger, days 3-8, configuration 1, simple load shifting scenario,
first optimization approach ....................................................................................................................... 157
Figure 5.13 Performance of the heat exchanger, days 3-8, configuration 2, simple load shifting scenario,
first optimization approach ....................................................................................................................... 158
Figure 5.14: Performance of the heat exchanger, days 3-8, configuration3, simple load shifting scenario,
first optimization approach ....................................................................................................................... 158
Figure 5.15: Heating needs from 11 to 13 January ................................................................................... 159
Figure 5.16: Performance of the heat exchanger, days 3-8, configuration 1, inlet exchanger air
temperature=38°C, night charging scenario, first optimization approach ............................................... 161

16
Figure 5.17: Performance of the heat exchanger, days 3-8, configuration 1, inlet exchanger air
temperature=44°C, night charging scenario, first optimization approach ............................................... 161
Figure 5.18: Performance of the heat exchanger, days 3-8, configuration 2, inlet exchanger air
temperature=44°C, night charging scenario, first optimization approach ............................................... 162
Figure 5.19: Performance of the heat exchanger, days 3-8, configuration 1, inlet exchanger air
temperature=38°C, night charging scenario, first optimization approach ............................................... 162
Figure 5.20: Weakly occupancy program .................................................................................................. 164
Figure 5.21: Incoming air temperature for an indoor temperature of 18 and 20 °C ................................ 165
Figure 5.22: Performance of the heat exchanger, days 3-8, configuration 3, inlet exchanger air
temperature=38°C, occupancy scenario, first optimization approach ..................................................... 166
Figure 5.23: Performance of the heat exchanger, days 3-8, configuration3, inlet exchanger air
temperature=44°C, occupancy scenario, first optimization approach ..................................................... 167
Figure 5.24: Performance of the heat exchanger, days 3-8, configuration 1, inlet exchanger air
temperature=44°C, occupancy scenario, first optimization approach ..................................................... 167
Figure 5.25: Performance of the heat exchanger, days 3-8, configuration 2, inlet exchanger air
temperature=44°C, occupancy scenario, first optimization approach ..................................................... 168
Figure 5.26: Performance of the heat exchanger, days 3-8, case 1, inlet exchanger air temperature=44°C,
simple load shifting scenario, second optimization approach .................................................................. 177
Figure 5.27: Performance of the heat exchanger, days 3-8, case 28, inlet exchanger air
temperature=44°C, simple load shifting scenario, second optimization approach .................................. 177
Figure 5.28: Performance of the heat exchanger, days 3-8, case 38, inlet exchanger air
temperature=38°C, simple load shifting scenario, second optimization approach .................................. 178
Figure 5.29: Performance of the heat exchanger, days 3-8, case 38, inlet exchanger air
temperature=44°C, inlet house air temperature= outlet exchanger –4°C, simple load shifting scenario,
second optimization approach .................................................................................................................. 179
Figure 5.30: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air
temperature=38°C, inlet house air temperature= outlet exchanger –4°C, simple load shifting scenario,
second optimization approach .................................................................................................................. 179
Figure 5.31: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air
temperature=38°C, inlet house air temperature= outlet exchanger –4°C, simple load shifting scenario,
second optimization approach .................................................................................................................. 182
Figure 5.32: Performance of the heat exchanger, days 3-8, case 38, inlet exchanger air
temperature=44°C, ameliorated insulation, night charging scenario, second optimization approach .... 182
Figure 5.33: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air
temperature=38°C, initial insulation, night charging scenario, second optimization approach .............. 183
Figure 5.34: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air
temperature=38°C, ameliorated insulation, night charging scenario, second optimization approach .... 183
Figure 5.35: Performance of the heat exchanger, days 3-8, case 38, inlet exchanger air
temperature=44°C, initial insulation, occupancy scenario, second optimization approach .................... 185
Figure 5.36:Performance of the heat exchanger, days 3-8, case 38, inlet exchanger air temperature=44°C,
ameliorated insulation, occupancy scenario, second optimization approach .......................................... 186

17
Figure 5.37: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air
temperature=38°C, initial insulation, occupancy scenario, second optimization approach .................... 187
Figure 5.38: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air
temperature=38°C, ameliorated insulation, occupancy scenario, second optimization approach .......... 187
Figure 5.39: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air
temperature=40°C, initial insulation, occupancy scenario, second optimization approach .................... 188
Figure 5.40: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air
temperature=40°C, ameliorated insulation, occupancy scenario, second optimization approach .......... 188

18
List of tables
Table 1.1: List of annual electrical consumption peaks (RTE, 2012, a) ....................................................... 35
Table 2.1: Comparison between sensible and latent heat storage (Farid et al., 2004) .............................. 44
Table 2.2: Comparison of different types of TES based on various performance parameters, (Abedin et
al., 2011) ...................................................................................................................................................... 46
Table 2.3: Main desirable characteristics of PCMs (Hale et al., 1971) ........................................................ 48
Table 2.4: Comparative table of the different LHS units ............................................................................. 72
Table 3.1: Properties of the Microtek 37D paraffin .................................................................................... 79
Table 3.2: PCM mass calculation ................................................................................................................. 81
Table 3.3: Description of the instrumentation ............................................................................................ 87
Table 3.4: PCM mass calculation for the new heat exchanger ................................................................... 98
Table 4.1: Enthalpy change during heating/cooling of the PCM based on DSC results ............................ 125
Table 4.2: Enthalpy change for the three part discretization heat capacity curves (heating and cooling)
................................................................................................................................................................... 134
Table 5.1: Properties of the materials used for the definition of the building ......................................... 145
Table 5.2: Composition of the building's surfaces .................................................................................... 146
Table 5.3: French electricity price rates during the day and during the night .......................................... 152
Table 5.4: Monthly energy consumption and cost for different inlet air temperature values and load
shifting periods for the initial exchanger .................................................................................................. 152
Table 5.5: Characteristics of the three selected configurations ............................................................... 154
Table 5.6: Simple load shifting operation description, first optimization approach................................. 157
Table 5.7: Monthly performance of the three tested configurations, simple load shifting scenario, first
optimization approach .............................................................................................................................. 159
Table 5.8: Night charging scenario operation description, first optimization approach .......................... 160
Table 5.9: Monthly performance of the three tested configurations, night charging scenario, first
optimization approach .............................................................................................................................. 163
Table 5.10: Occupancy scenario operation description, first optimization approach .............................. 166
Table 5.11: Monthly performance of the three tested configurations, occupancy scenario, first
optimization approach .............................................................................................................................. 169
Table 5.12: Modified parameters for the second optimization approach ................................................ 171
Table 5.13: Classification of the obtained results for the charging of the unit according to faster charging
time, airflow rate of 300m3/h ................................................................................................................... 174
Table 5.14: Comparison between initial and optimized version of the heat exchanger .......................... 175
Table 5.15: Simple load shifting operation description, second optimization approach ......................... 176
Table 5.16: Monthly performance of the tested configurations, simple load shifting scenario, second
optimization approach .............................................................................................................................. 180
Table 5.17: Night charging operation description, second optimization approach .................................. 181
Table 5.18: Monthly performance of the tested configurations, night charging scenario, second
optimization approach .............................................................................................................................. 184
Table 5.19: Occupancy scenario operation description, second optimization approach ......................... 185

19
Table 5.20: Monthly performance of the tested configurations, occupancy scenario, second optimization
approach.................................................................................................................................................... 189

20
List of symbols
Upper case
Cp Specific heat capacity [J/(kg·K)]
E Energy [J]
H Enthalpy [J/kg]
R Thermal resistance [(m2K)/W]
L Latent heat of fusion (J/kg)
P Power [W]
Patm Standard atmospheric pressure [Pa]
Psat Saturation pressure [Pa]
Q Heat [J]
S Surface [m2]
T Temperature [°C, K]
̅
T Average temperature [°C, K]
V Volume [m3]

Lower case
dx Length of node [m]
dy Thickness of node [m]
dz Height of node [m]
h Convection coefficient [W/m2·K]
m Mass [kg]
ṁ Airflow rate [kg/s, m3.h-1]
ps Static pressure
pt Stagnation pressure
t Time [s]
w Mixture ratio of water vapor mass and dry air mass

Greek symbols
ε Efficiency [%]
λ Thermal conductivity [W/(m·K)]
ρ Density [kg/m3]
φ Relative humidity [%]

Subscripts
Air, a Related to air / position in the air stream
Inlet/in Heat exchanger inlet point

21
Outlet/out Heat exchanger outlet point
app Refers to apparent
exp Refers to experimental
i Node position lengthwise
j Node position crosswise
l Referring to liquid PCM
m Referring to PCM layer
n Total number of nodes lengthwise
p Referring to aluminum layer
q Total number of PCM nodes crosswise
s Referring to solid PCM
sim Refers to simulation

Acronyms
ADEME French Environment and Energy Management Agency
ANR French National Research Agency
CHS Chemical heat storage
COP Coefficient of performance
DSC Differential Scanning Calorimetry
EDF Electricity of France
EES Electrical energy storage
ES Energy storage
EU European Union
GHG Greenhouse gas
HVAC Heating, ventilation and air conditioning
IEA International Energy Agency
INSEE French National Institute of Statistics and Economic Studies
LHS Latent heat storage
MEDDE French Ministry of Sustainable Development
PCM Phase change material
PEB Positive energy building
RT Thermal regulation Act
RTE French Electricity Transport Network
SHS Sensible heat storage
TES Thermal energy storage
UN United Nations

22
General introduction

23
General introduction

Context
Human-induced global warming, fossil resources rarefaction, oil price shocks, increase in world and
urban population constitute an alarming list of issues which require serious debates, rigorous decision
making and innovative solutions. Several agreements and policies have already been put into practice at
the international and national levels and further initiatives are being prepared for the future, reflecting
the rising world concern and the need for action.

Regarding these considerations, the building sector plays a key role, as it is one of the major contributors
worldwide concerning energy consumption and greenhouse gas emissions. In view of these facts,
European and French policies are considering imposing zero or even positive energy buildings regulations
in the near future. Two main problems have to be dealt with regarding those goals: renewable energy
intermittence and daily or seasonal variations of energy consumption. A significant amount of research
has been performed concerning Energy Storage, proving that this field can provide pertinent applications
as an answer to these drawbacks.

In general, France faces the above stated challenges but it also presents a specific particularity: high peak
power demands. These peaks are mostly encountered during the late afternoon winter period and are
closely related to electrical heating. A project was formulated under the aegis and founding of the
French National Research Agency aiming at developing and controlling an active PCM-Air heat exchanger
coupled to a heat pump, as a technical solution to the augmented electricity demand during the 18:00-
20:00 time slot. The project partners include four research units (LGCB, CETHIL, ARMINES – CEP and EDF
R&D) and an industrial one (RIBO).

Objectives and approach

It is within this context that the here presented PhD thesis seeks to investigate the potential of a PCM-Air
heat exchanger and proposes control strategies that will perform the needed shift in electrical
consumption without degrading thermal comfort and indoor air quality.

The general idea is to develop a heat exchanger unit integrated in a ventilation system, in which a
sufficient amount of heat will be stored in order to substitute the use of electrical heating during the
peak power demand period. Three major objectives were identified:

 The development of experimental and numerical tools that will enable the elaboration, testing
and evaluation of the control strategies;
 The optimization of the heat exchanger’s performance;
 The elaboration of load shifting control strategies for different scenarios.
The approach that was followed in order to achieve these goals is a retroactive process based on two
tools: an experimental and a numerical one (figure A.1). A PCM-Air heat exchanger was designed,
constructed, instrumented, characterized and coupled to an experimental test cell. In parallel, a
numerical model was developed using the apparent heat capacity approach and the finite difference
method. After calibration and validation with the use of experimental data were conducted, the model

24
was coupled to a building model. An optimization study was performed focusing on the geometry of the
exchanger and the PCM properties in order to obtain faster charging/discharging periods with minimum
effect on the stored energy. Finally, the optimized version of the exchanger was used for the elaboration
of control strategies.

Chapter contents
The thesis is divided into 5 chapters, presenting the various elements that had to be linked together in
order to achieve the established goals.

The first chapter briefly presents the energy and built environment context in an International, European
Union and French level. Special focus is given on the French peak power demand issue and its strong
association with electrical heating. The implications of the peak power demand are discussed, along with
existing load shifting management methods and technologies.

The first part of the second chapter deals with energy storage, presenting different methods through
which it can be achieved. It focuses on thermal energy storage processes and targets phase change
materials by demonstrating existing applications in buildings, enhancement techniques as well as
limitations and barriers to their use. The second part of the chapter describes the main elements,
operation principles and design philosophy and methods of active latent heat storage units. Finally,
existing applications and studies are discussed.

The third chapter presents the experimental part of the study. The assembly procedure of the
experimental platform is demonstrated and all the important choices concerning design are explained
including PCM, geometry, instrumentation and data acquisition. In a second phase, characterization
results are presented, as well as the coupling with an officelike experimental cell and a preliminary
control strategy.

The fourth chapter presents the elaborated numerical model that is used to simulate the heat
exchanger’s behavior. Experimental results are used for the model’s calibration and validation; following
that, the model is coupled to an existing building simulation model.

Chapter five presents the optimization study of the heat exchanger, performed through the numerical
model. Several parameters are investigated (geometry, PCM properties) in order to obtain an
ameliorated performance of the heat exchanger in terms of minimum charging time and augmented
power. The optimized versions of the exchanger are then used in realistic scenarios in order to achieve
load shifting while maintaining acceptable indoor thermal comfort and air quality conditions.

Finally, the last section of this thesis presents the conclusions of the study and discusses the perspectives
that can be tackled in the future.

25
 Methodology

Peak power demand

(linked to electrical heating)

Thermal storage
Load shifting strategies

Tools

Experimental Numerical
 Heat exchanger elaboration  Apparent heat capacity approach
 Characterization  Finite difference method
 Coupling with Hybcell Validation  Coupling with building model

Load shifting strategies Heat exchanger optimization

 Substitute electrical heating  Geometry, PCM parameters
 Charging/discharging time
 Thermal comfort
 Load shifting
 Indoor air quality  Augmented energy storage

Figure1: The folowed

Figure
methodology
A.1: The methodology
for the development
followedofalong
load shifting
this thesis
control strategies

26
Chapter 1
Energy context and built environment

27
Chapter 1 : Energy context and built environment

1.1 Introduction
For almost half a century energy-related issues have gained increasing concern within societies all over
the world due to their significant economic and environmental importance. Starting from the oil price
shocks in the 70s and until the current awareness about global warming, energy production and
management is now a prominent subject in international, national and local debates and policies.
This chapter starts by presenting important international, EU and French policies that reflect the global
climate change awareness and sets a significant number of goals for the near and far future. Special
focus is then given to the French context and more specifically to the influence of the built environment
in energy consumption and greenhouse gas emissions. The considerable share of electrical heating in
augmented peak power demand is demonstrated using statistics and figures. Finally, the implications of
elevated electricity demand are debated and existing load shifting methods are mentioned as an
introduction to energy storage.
Overall, this chapter aims to pass from global energy concerns to the French energy context and the
electrical heating-related peak power demand in order to clearly state the problem in question and
underline the necessity for efficient solutions.

1.2 At the international and EU levels

Several agreements have been signed between countries worldwide, aiming to reduce energy
consumption and greenhouse gases (GHG) emissions. The cornerstone of these efforts was the Kyoto
Protocol (United Nations, 1998), an international agreement under the United Nations Framework
Convention on Climate Change which sets emissions reduction targets to its participating parties.
Concerning the European Union, the ‘Climate and Energy Package’ (European Commission, 2008) was
adopted in December 2008, focusing on emissions cuts, renewable energy production increase and
energy efficiency. The Package sets three key objectives (known as the ‘20-20-20’ targets) for 2020: 20%
reductions of GHG emissions comparing to 1990 levels, 20% energy consumption from renewable
resources and a 20% improvement in energy efficiency. Before that, the Energy Performance of Buildings
Directive (European Commission, 2002) implemented enhanced building regulations and introduced
energy certification schemes for buildings in all EU countries. Furthermore, the Communication ‘Energy
Roadmap 2050’ (European Commission, 2011) was adopted in 2011, showing the EU’s commitment to
reducing GHG emissions to 80-95% below 1990 levels by 2050.
The above agreements and legislations demonstrate the increasing concern for environment friendly
policies as well as the demand for more efficient ways of energy utilization.

Built environment: a big energy consumer

One of the most influencing factors when it comes to the energy domain is the built environment:
according to the International Energy Agency buildings represent about 40% of primary energy
consumption among its 29 member countries. Furthermore, it is estimated (UNEP SBCI, 2009) that

28
buildings contribute up to 33% of the global GHG emissions, mostly due to fossil fuels utilization during
the operational phase. These numbers clearly demonstrate the significant environmental impact of
buildings and the key role they can play in reducing energy consumption and GHG emissions.

1.3 At the French level

1.3.1 Energy context and policies

Until 1973, oil constituted the main means of energy production in France, in order to support the
industrial development and the declining use of coal. Between that year and the late ‘80s, the oil price
chocks led to a better control of energy consumption and oriented its production to other sources. First
the nuclear sector became the dominant producer –it is still so nowadays. An increase of renewable
sources production could be observed after the ‘90s, mostly from hydraulic energy but also from
photovoltaic and wind energy (INSEE, 2012). The annual global electricity consumption has tended to
stabilize since the ‘00s, presenting, however, increasing peak power demands, mainly during the winter
period. A significant dependence on temperature variation is also observed during winter, mainly due to
electrical heating (RTE, 2012a, 2013).
The Kyoto Protocol was implemented through the ‘Climate Plan 2004-2012’ (MEDDE, 2006) aiming at a
10% reduction of CO2 emission as well as aiming to raise awareness within society and among public
and private actors etc. It was reinforced by the commitments established by the First and Second
Grenelle Acts (MEDDE, 2009, 2012) following the ‘Grenelle de l'environnement’ debate, a multi-party
debate launched in 2007. The main commitments include the generalization of standards of low
consumption in new housing and public buildings, the renovation of housing and building heating,
transportation measures favoring low polluting means, the development of renewable energy, as well as
other actions in the health, agricultural and biodiversity domains.
As far as the building sector is concerned, several Thermal Regulations acts have been voted (1974, 1988,
2000, 2005, and 2012), with an increasing interest observed in the last 15 years. They aim to set upper
limits on the energy consumption of new buildings for heating, ventilation and air conditioning (HVAC),
lighting and hot water consumption and to promulgate the renovation of existing residential and tertiary
buildings, with a 38% reduction goal in the consumption of the existing park.
It is worth noting that the new Thermal Regulation Act (RT 2020, currently under discussion) will
introduce the concept of positive energy buildings (PEB). It will aim to divide energy consumptions by
four and oblige all new buildings to achieve PEB standards.

1.3.2 Built environment and energy consumption

The built environment and energy consumption in France is not an exception compared to the European
and international situation, as this sector represents a significant part of energy utilization and GHG
emissions (ADEME, 2013, a, b). More specifically, the housing and tertiary sector is to account for 44% of
the final energy consumption (figure 1.1), 20% of GHG and 22% of CO2 emissions (figure 1.2).

29
 Energy consumption

Figure 1.1: Evolution of energy consumption per sector

CO2 emission

Year
Agriculture / forestry
Industry
Energy transformation
Transport
Residential – tertiary
Net absorption of biomass

Figure 1.2: Evolution of CO2 emission per sector

30
 Electricity is the main energy source of French buildings representing 37% of final consumption,
followed by gas (32%), oil (16%), renewable/waste (15%) and coal (0,4%) (figure 1.3). It presents a 15%
increase since 1973 and along with renewable sources and gas it has contributed to the decrease of coal
and oil consumption.
In 2012 :

In 1973 :

Electricity
Renewable and waste
Gas
Oil
Coal

Figure 1.3: Final consumption of the residential/tertiary sector

Furthermore, final electricity consumption in buildings has been increasing with an annual rate of 8%
over the past 30 years (figure 1.4, MEDDTL, 2010). This can be explained by the population growth and
the subsequent increase of the building stock (+41% of houses in 30 years), the increase of the mean
housing surface, the augmented comfort expectations and the apparition of new needs linked to
technology advancements. These factors have contributed to a significant increase of electricity
consumption for HVAC, household/office appliances, lighting, etc.
TWh
Residential – tertiary
Electricity consumption

Industry

Steel and iron

Transport

Agriculture
Year Source SOeS
Figure 1.4: Evolution of final electricity consumption per sector

31
 Apart from the performance of the building itself, one also has to consider the related energy and
environmental impact on the construction sector (concrete, wood, steel and other components) and on
management policies (access to energy sources and networks, transport connection, etc).

1.3.3 Electrical heating

[Link] Residential sector

In 2012, the residential sector has risen to 33,4 million houses, out of which 27,8 million are main
residences. Since 1973, the average unitary consumption (including wood) has shown a significant
reduction of 58%, which can be attributed to renovation, new thermal policies and a more energy
efficient behavior among occupants (ADEME, 2013, a). Electricity and gas are the two main energy
sources of the residential sector. Figure 1.5 illustrates the energy distribution in this sector: almost 2/3 of
energy consumption is employed for heating needs, followed by specific usage (PC, drying, TV and other
appliances) (19,5%), hot water (12,1%) and cooking (7%).

KWh / housing

Apartment
buildings
Year

Individual
housing

Unitary consumption
Heating
Hot water
Cooking
Specific usage

Figure 1.5: Energy distribution within the residential sector

32
Electrical heating is used in 9,5 million households nationwide, a number that represents 35% of main
residences; 44% use gas and 14% fuel (RTE, 2012, b).
Even though the mean unitary housing heating needs have decreased by 44% since 1973, the overall
electricity consumption in main residences has increased by 130% in the same period. This can be
explained by the strong penetration of specific usage and by the augmenting usage of electrical heating.
During the 00’s electrical heating was largely preferred in the newly built buildings, mostly because of its
advantageous installation and maintenance cost and its moderate price evolution, when compared to
other heating means. The percentage of electrical heating use in newly built housing reaches almost 75%
of cases in 2008 (figure 1.6). The last Thermal Regulation (RT 2012) resulted in an important decrease of
this percentage, as it focuses on primary energy consumption and not on the final one (as was the case
with the former regulation RT 2005).
Nevertheless, according to some estimations (NégaWatt 2009), electrical heating reaches an annual
consumption of 60 TWh, a number that represents 27% of total electricity consumption of the residential
sector and 12,3% of the total annual electricity consumption in France.

%
Percentage of electrical heating (newly built)

Year
Figure 1.6: Percentage of electrical heating in newly built housing

33
These figures depend on various factors, one of the most important ones being the weather conditions
and the rigorousness of the winter period. For example, throughout the particularly cold winter of
2008/2009 the electrical heating consumption rose to 70 TWh, which corresponds to the total energy
consumption of a country like Switzerland. Furthermore, the French electrical Distribution Network (RTE)
estimated that during the cold wave of January 7th 2009 electrical heating represented 53% of total
residential consumption.

[Link] Tertiary sector

In 2011, the heating area of the tertiary sector accounts for 931 149 000 m2 (ADEME, 2013, a). As for the
residential sector, gas is the main energy heating source (46%), electricity coming second (25,9%) and
fuel last (18,1%). Although the tertiary sector ranks low in the French energy balance, it appears as the
most dynamic in terms of energy consumption increase, with an annual growth of 2% for electricity and
0,7% for all energies combined. This fact shows that it can have an increasing impact in the future hence
constituting a considerable aspect of energy management policies.

1.3.4 Peak power demand and thermo-sensitivity of French electrical consumption

For the past 15 years electrical consumption in France has shown increasing peaks in electrical
consumption (table 1). For the first time, the symbolic threshold of 100.000 MW was surpassed during
the cold wave of February 2012. Furthermore, according to RTE (RTE, 2012, a), the increase of these
peaks is faster and more dynamic than the increase of the annual electrical consumption. The maximal
values are mostly observed during the late afternoon hours of the winter period (19:00) (figure 1.7).
They are the result of a decreasing level of tertiary activities, increased public transportation needs and
the starting of evening household activities.

Annual electrical consumption peaks

Thursday 17/01/2013 92 600 MW

Wednesday 08/02/2012 102 100 MW

Tuesday 04/01/2011 91 820 MW

Wednesday 15/12/2010 96 710 MW

Wednesday 07/01/2009 92 400 MW

Monday 15/12/2008 84 420 MW

Monday 17/12/2007 88 960 MW

Friday 27/01/2006 86 280 MW

Monday 28/02/2005 86 020 MW

Wednesday 22/12/2004 81 400 MW

34
 Thursday 09/01/2003 83 540 MW

Tuesday 10/12/2002 79 730 MW

Monday 17/12/2001 79590 MW

Table 1.1: List of annual electrical consumption peaks (RTE, 2012, a)

One of the most influential factors of this phenomenon is electrical heating, as justified by the figures
presented in the previous sections. RTE underlines the strong dependence of electrical consumption on
outdoor temperature values, particularly during winter. Figure 1.7 also illustrates the increase in power
consumption during the cold wave of February 2012.

MW
Power consumption

Time

Figure 1.7: Peak power demand around 19:00, before and during the cold wave of February 2012

The term thermo-sensitivity is applied to describe this phenomenon, with a calculated 2300 MW
increase in consumption per Celsius degree temperature increase during the 2012 winter. This value
presents a growth of 35% when compared to the 2001/2002 one. This phenomenon is much more visible
in France than in other European countries as it represents almost half of the total thermo-sensitivity
observed in Europe. During periods of great cold, thermo-sensitivity is the predominant factor in
electrical consumption. For instance, if we consider the cold wave observed in February 2012, the

35
thermo-sensitivity part represented on average 40% of the consumption. Figure 1.8 illustrates the
electrical consumption and the equivalent thermo-sensitivity part for the first half of February 2012.

Consumption
MW Thermo-sensitivity part
Electrical consumption

Date
Figure 1.8: Electrical consumption and equivalent thermo-sensitivity part, first half of February 2012

Contrary to the increase in peak power demand over the last 15 year, the annual minimum consumption
observed in the summer remains stable during the same period. This results in increasing differences
between winter peak demands and summer minimal levels of consumption.

1.3.5 Peak power demand implications

The magnitude of the peak power demand has important economic, environmental and social
consequences.
The French energy management system faces increasing difficulties in responding to such events, as it
relies in great part on nuclear power plants, presenting mostly constant production rates of 63.130 MW,
that cover 77% of total energy consumption (RTE, 2014). As a result, France has to call upon other means
of energy production and/or address neighboring countries for importations (figure 1.9). In both cases
fossil fuel utilization is usually involved, leading to an increase of CO2 emission.

36
 Demand coverage (08/02/2012)
Imports
Hydraulic
Thermal (fossil fuel)
Other renewable
Wind
Nuclear

Figure 1.9: Demand coverage during the cold wave of February 8th 2012

According to RTE, CO2 emissions have been three times higher in winter than in summer over the past six
years. Estimations report that a change of electrical heating consumption would result in a change of CO2
emission by 500-600 g CO2/kWh (ADEME & RTE, 2007). Furthermore, RTE notes that France has been
importing electricity from Germany over the past few years, even though Germany has announced its
exit from the nuclear power and closed down seven nuclear sites in 2011.
At the same time, peak power demands lead to the starting of thermal power plants, the reduction of
energy exports and the increase of energy importations. In all cases, the energy provider is forced to pay
the KWh obtained at price market values, for example two or three times more than the equivalent gas
produced energy. The augmented price range is bound to be transferred to the consumer’s electrical bill.
In addition, the peak power consumption period represents a relatively small percentage of the total
annual consumption. Nevertheless, significant oversizing investments have to be made that are
practically not used for the greater part of the year. These costs are also likely to be passed on to the
consumers and more specifically to the fixed part of their bill that corresponds to infrastructure
development. This outcome penalizes above all small individual consumers, that is to say precarious or
energy efficient households.
Increasing bills can contribute to energy precariousness of low income households that cannot afford a
well-insulated dwelling or a higher energy budget. Energy precariousness is defined as the difficulty in
obtaining the necessary energy supply for the satisfaction of basic needs. This difficulty may occur in
conditions of the occupants’ insufficient resources or inadequate habitat conditions. According to the
French National Institute of Statistics and Economic Studies (INSEE, 2011), 3.8 million households spend
more than 10% of their income on energy supply, whereas 3.5 million declare that they suffer from cold
in their homes. 621.000 households suffer from both forms of precariousness.

37
Moreover, according to some assumptions (NégaWatt, 2009), the augmented seasonal demand leads to
an increasing concentration of the nuclear units stop in summer in order to obtain maximum power
plants operation during the mid-season and winter. Additionally, the French Electricity Company (EDF)
performs load-following strategies in some of these plants, resulting in lower production factors. These
two management choices differentiate from the original designed operation of the nuclear power plants
and may result in a more complicated exploitation, additional costs and degradation of the parc.
Finally, in extreme cases peak electricity demand can lead to partial or total failure in the distribution
system (blackout), with severe economic (production deficiency, material damage, etc) and social
consequences (transports, discontent among the electricity-depending population, etc).

Resort to other means of energy

production (non-nuclear)
Import energy from neighboring
countries Environnemental
implications

Oversizing investments of
energy production sites

Economic
implications

Differentiation from original

operation of nuclear energy
production sites

Social
implications
Eventual partial or total failure in
the energy distribution system

Figure 1.10: Implications of peak power demand

38
1.3.6 Applied electricity load-shifting methods
Load-shifting management methods and technologies constitute efficient solutions to the increasing
electricity peak demands, by storing energy and moving power loads to off-peak times in order to
maintain the supply-demand balance and ensure a good operation of the network. Energy storage can
be achieved through various processes: electrochemical (batteries, fuel cell), electrical (capacitor,
superconducting magnetic field), thermal (sensible, latent, chemical form) and mechanical (hydraulic,
flywheel) ones. Some of them provide only short term energy storage and others can achieve very long
term storage.
Concerning France, two main peak load shifting categories can be noted:
- The industrial one, that consists in reducing or even stopping the consumption of one or more
industrial sites.
- The scattered one that includes the sum of several load shifting application among individuals or
professionals.
In both cases, RTE signs contracts with the consumers in question and solicits them in case of important
electrical demand, providing advantageous measures in exchange of lower consumption. According to
RTE (RTE, 2013), the total load shifting capacity (all features combined) was nearly 900Mw during some
days of November 2013. Globally, for the year 2013, the annual activated load shifting volume rose to 20
GWh.

1.4 Conclusion
Throughout this chapter we discussed the increasing concern for energy related subjects, focusing on the
French context and the increasing peak power demands. We explained the influence of electrical heating
on higher energy needs along with the serious economic, environmental and social problems that may
be provoked by this issue. We briefly discussed existing solutions as a transition to the second chapter
where energy storage and latent heat storage systems are presented as efficient ways to deal with peak
electricity demand.

39
Chapter 2
Energy storage, PCM and HVAC

40
Chapter 2 : Energy storage, PCM and HVAC integrated applications

2.1 Introduction
In the previous chapter the international and French energy context was presented, emphasizing on the
peak power problem linked to electrical heating. This chapter discusses the benefits of energy storage
technology in terms of consumption shifting and more efficient management of the built environment.

At first, the field of application of such technology is presented, along with its different forms, focusing
on thermal energy storage. Furthermore, phase change materials are discussed as one of the most
frequent ways to achieve thermal storage (classification, properties, containment, applications,
enhancement techniques and limitations).

Then several existing studies using PCM integrated in HVAC systems are compared, in order to acquire
useful information regarding their development and utilization in the building. Finally, an attempt to
summarize the operation principles and design methods is performed, as a guide for the fabrication of a
PCM-Air heat exchanger.

2.2 Energy storage: necessity and methods

As discussed earlier, the built environment is one of the major contributors in what concerns energy
consumption and GHG emissions. Furthermore, scientific communities have been alerting regarding
human-induced climate change and its impact on average earth temperature elevation. At the same
time, the world population is increasing and is expected to rise up to 9.6 billion by 2050 (UN, 2013). A
growing percentage of the population will be living in cities, indicating that high density urban
landscapes will also increase in the future. One of the consequences of this fact will be the intensification
of the Urban Heat and Pollution Island effect that can negatively influence mortality rates and heat-
related diseases (Haja et al., 2002). Indeed, the 2003 summer heat wave was directly linked to 70.000
deaths in European countries (15.000 deaths in France only) (Robine et al., 2008).
In view of these facts, the need for more efficient energy buildings seems essential, the replacement of
conventional energy sources with renewable and waste ones being a highly promising perspective.
Nevertheless, implementing renewable sources in the building presents some drawbacks, the main
difficulty being the variability and intermittence of their availability (especially concerning wind and solar
energy).
In addition, as presented in the previous chapter, daily and seasonal variations in energy consumption
lead to peak power usage and tend to become an important issue for energy companies, France being
one of the most affected countries in Europe.

41
In light of these elements, Energy Storage (ES) can play a key role in energy efficient buildings by
providing a number of solutions for intermittent resources and mismatching demand, while maintaining
comfortable indoor environment. Furthermore, it can contribute to energy supply security, advanced
control strategies elaboration (associated with smart grids) and thermal inertia increase.
ES can be performed through various processes, including mechanical, electromagnetic, chemical,
biological and thermal ones. Some of them provide only short term energy storage and others can
provide very long term storage. This chapter will mainly deal with thermal energy storage, presenting the
main methods through which it can be achieved and focusing on phase change materials technology.

2.2.1 Thermal energy storage

Thermal energy storage (TES) allows the stocking of thermal energy in order to reduce indoor air
temperature fluctuation or to reinstitute it to the system (building, district, and town) at a later period
with an hourly, daily or even seasonal time lag. Excess storable energy can be the result of intermittent
power sources (mostly renewable ones) or waste heat from domestic appliances (cooking, ventilation,
etc.) (figure 2.1).

Figure 2.1: Internal and external solicitations in a building, possible heat sources

Depending on the application, the following elements are usually used in TES systems:

 A storage medium where heat or cold is stored.

 A containment component that is used as a reservoir for the storage medium.
 A circuit that allows the flow of the heat/cold transferring fluid (air or water).
 A heat/cold source that is used to heat/cool the circulating fluid.
 A fan or a pump that regulates the flow of the circulating fluid in the system (for some systems).

42
The operation of TES systems can be divided into three distinct stages: charging, storing and discharging
heat or cold. Thermal insulation is needed in order to avoid losses over these processes, as the storage
medium is at higher or lower temperatures compared to the surrounding environment. The operational
temperature range depends on the applicable case and presents limitations related to the heat source
and the storage material.
Parameters that ensure a pertinent implementation of TES systems in the building include:

 high energy density of the storage element

 favorable heat transfer and compatibility between transfer fluids
 container and storage medium compatibility
 stability of the storage material during the charging/discharging process
 long-lasting reversibility of these processes
 limited heat losses, sufficient power load, cost effective system design and storage material
TES can be achieved through three major methods: sensible heat storage, latent heat storage and
chemical heat storage.

[Link] Sensible heat storage and building applications

Sensible heat storage (SHS) occurs in a medium when its temperature changes within a fixed range and
without phase change taking place. The storage medium can be a liquid such as water (the most
common use) or thermo-oil, a solid such as brick, concrete, sand and soil or even metals such as
aluminum and steel. SHS is usually simpler and less expensive to implement than latent heat storage. It
presents however lower energy efficiency and storage capacity, which leads to rather large volumes of
TES systems. As already explained, in SHS, thermal storage occurs only through an apparent temperature
change of the storage medium. Likewise, heat/cold release also manifests at varying temperatures, a fact
that can be a drawback if a specific temperature range is sought for.
The capacity of a storage system is defined by the specific heat capacity and the mass of the storage
material. The following equation is used to describe the SHS process:
𝑄 = 𝑚𝐶𝑝 (𝑇2 − 𝑇1 ) equation 2.1

Where Q is the amount of stored sensible heat (J), m is the storage medium mass (kg), C p is the specific
heat of the storage medium at constant pressure (J/Kg.K) and T1 and T2 are the temperatures before and
after stocking/destocking.
SHS applications can be found in wall insulation, inertia increasing techniques, solar or gas water heaters,
load levelling, night cooling etc.

[Link] Latent heat storage and building applications

Latent heat storage (LHS) occurs when the storage medium undergoes a phase change, from solid to
liquid (usually) or liquid to gas or solid to gas, and vice versa. The heat stored or released is used for the
phase change, thus allowing to obtain much higher storage capacity and a narrower temperature
operating range when compared to SHS systems. This can be demonstrated by considering the melting

43
of 1kg of ice, a process that requires 333KJ of energy. The same amount is needed to raise the
temperature of 1 kg of water from 0°C to 78°C.
The storage media used in such systems are called phase change materials (PCM), the most common
being paraffin, hydrated salt and fatty acids. Table 2.1 presents a comparison of SHS (rock bed and water
tank) and LHS technology (organic and inorganic PCM) (Farid et al., 2004). The advantages of LHS are
evident in terms of storage density and volume.

Rock Water Organic PCM Inorganic PCM

Density, kg/m3 2240 1000 800 1600

Specific heat, kJ/kg 1.0 4.2 2.0 2.0

Latent heat, kJ/kg - - 190 230

Latent heat, kJ/m3 - - 368 368

Storage mass for 106 J, kg 67,000 16,000 5,300 4,350

Storage volume for 106 J, m3 30 16 6.6 2.7

Relative storage mass 15 4 1.25 1.0

Relative storage volume 11 6 2.5 1.0

Table 2.1: Comparison between sensible and latent heat storage (Farid et al., 2004)

On the other hand, LHS systems present practical difficulties linked to PCM properties: low thermal
conductivity, density (and volume) change, chemical instability and other disadvantages that will be
discussed in detail later. In most LHS systems SHS also occurs before and after the phase change. The
equation that describes the total heat transfer process is as follows:
𝑄 = 𝑚𝐶𝑝,𝑠 𝛥𝑇 (𝑠) + 𝑚𝐿 + 𝑚𝐶𝑝,𝑙 𝛥𝑇 (𝑙) equation 2.2

Where Q is the total amount of stored heat (sensible and latent) (J), m is the storage medium mass (kg),
Cps and Cpl (J/Kg.K) are the specific heat of the storage medium in solid and liquid form, ΔT is the
temperature difference and L is the specific latent heat (J/kg). The first term represents the sensible heat
exchanged during the solid phase, the second one the latent heat absorbed or released during the phase
change and the last one the sensible heat exchanged during the liquid phase.
Figure 2.2 demonstrates a theoretical temperature profile variation as a function of stored heat during a
sensible and a latent heat exchange process. Practically, a small temperature increase is also observed
during the latent phase due to the impurity of materials and convective heat transfers.
Concerning the built environment, LHS technology can be introduced into the building structure (wall,
ceiling, façade) or in separate heat and cold storage devices usually related to HVAC systems. The first

44
category constitutes the passive elements of the building, whereas the second one is an active
component that can be used on demand.

Figure 2.2: Temperature profile variation with stored heat evolution

LHS applications have also been developed in a variety of domains including medical or food transport
technologies, spacecraft thermal systems, vehicle thermal comfort, protection of electronic devices,
computer cooling, clothing, etc.

[Link] Chemical heat storage and applications

Chemical heat storage (CHS) can be divided into two categories: thermochemical processes and sorption
processes.
The first ones are based on the energy released and absorbed during the breaking and the formation of
molecular bonds for completely reversible chemical reactions. Thermal energy can be absorbed when
the endothermic dissociation of a material (C) takes place, this constituting the charging procedure. This
process leads to the formation of two separate components (A and B), which have different properties
and can be stored at ambient temperatures for long periods with few or no thermal losses. Finally, the
discharging process occurs when the reversed procedure is performed and the components A and B are
mixed through an exothermic reaction that leads to the reformation of the initial material C. The
described procedures can be represented by the following equation:
𝐶 + ℎ𝑒𝑎𝑡 ⇄ 𝐴 + 𝐵 equation 2.3
Where C is the thermochemical material and A and B are the reactants. Reactant A can be a hydroxide,
hydrate, carbonate, ammoniate and reactant B can be water, carbon monoxide, ammonia, hydrogen etc.
Usually material C is in liquid or solid form, whereas materials A and B can be in any phase.

45
Sorption systems involve adsorption and absorption processes. The first one takes place when a
molecular or atomic layer is shaped, following the accumulation of an adsorptive on the surface of an
adsorbent. The adsorptive can be a liquid or gas while the adsorbent can be a solid or liquid. Absorption
consists in the formation of a solution which is occurring when a substance is distributed into a liquid or
solid.
CHS is an emerging field presenting the advantages of high energy density, with compact and long term
storage potential. On the other hand, it is more expensive and technically complex to implement
compared with sensible and latent heat storage technology.
Table 2.2 resumes the main parameters of the three TES methods, as well as their advantages and
disadvantages (Abedin et al., 2011).
Type of Thermal Energy Storage
Performance
Chemical TES (Sorption
Parameter Sensible TES Latent TES
and Thermochemical)
Up to:
110 °C (water tanks)
20-40 °C (paraffins)
Temperature range 50 °C (aquifers and 20-200 °C
30-80 °C (salt hydrates)
ground storage)
400 °C (concrete)
Low (with high Moderate (with low
temperature interval): temperature Normally high:
Storage density
0.2 GJ/m3 (for typical interval): 0.5-3 GJ/m3
water tanks) 0.3-0.5 GJ/m3
Often limited due to Depends on reactant
Lifetime Long storage material degradation and side
cycling reactions
Available commercially Generally not available,
for some but undergoing
Technology status Available commercially
temperatures and research and pilot
materials project tests
High storage density
Low heat losses
Low cost Medium storage density (storage at ambient
Reliable Small volumes temperatures)
Advantages Simple application with Short distance transport Long storage period
available possibility Long distance transport
materials possibility
Highly compact energy
storage
Low heat conductivity
Significant heat loss
Corrosivity of materials
over time (depending High capital costs
Disadvantages Significant heat losses
on level of insulation) Technically complex
(depending
Large volume needed
on level of insulation)
Table 2.2: Comparison of different types of TES based on various performance parameters, (Abedin et
al., 2011)

46
2.2.2 Electrical energy storage
Electrical energy storage (EES) consists in transforming electrical energy into other forms that facilitate
its storage for later use. This can be achieved through a wide range of different technologies, including
mechanical, electromagnetic, chemical, biological and thermal ones. Several studies have been
performed in the field of EES (Chen et al., 2009), a detailed description of which exceeds the interest of
this thesis. Figure 2.3 illustrates a classification of the different forms of EES and the existing applications
that derive from them (Fernandes et al., 2012).

Figure 2.3: Classification and applications of EES (Fernandes et al. 2012)

2.3 Phase Change Materials

Phase change materials (PCM) constitute one of the main elements of latent heat storage systems. The
phase change process and the subsequent latent energy storage/release enables them to store
significant amounts of heat/cold in small volume and within a certain temperature range. In the majority
of PCM building applications, the solid-liquid phase change is preferred because of the small volume
change compared to liquid-gas or solid-gas transformations. Pure PCMs have a distinct temperature in
which the phase transformation occurs (Tf). When the material is heated to reach this temperature its
chemical bonds start to break up through an endothermic process that leads to a solid to liquid
transformation and absorption of heat. On the contrary, when the material’s temperature decreases
again, it regains its solid form during an exothermic process that releases heat. Figure 2.4 (left) shows
the cooling process of such a material over time, demonstrating the constant temperature values during
the melting phase. Figure 2.4 (right) illustrates the heat flux variation during liquefaction for a fixed
temperature difference, with an apparent increase during the phase change. The phase change process
of some PCMs is observed within a temperature range, instead of a distinct value. This is due to the
mixture of different components that is made in order to acquire PCMs with specific properties.

47
 Figure 2.4: PCM temperature and heat flux evolution during solidification, 1: liquid phase 2: melting
phase 3: solid phase (Baetens et al., 2010)

Table 2.3 depicts the main desired thermal, physical, chemical and economic properties for an efficient
implementation of PCMs in building applications (Hale et al., 1971).

Thermodynamic
Kinetic properties Chemical properties Economic properties
properties
Phase change
Non-toxic and non-
temperature meeting High density Cost (cheap)
polluting
application needs
High thermal Compatible with
Availability (abundant
conductivity (in liquid Low density variation storage containers/no
and easily available)
and solid phases)+ corrosiveness
High rate of crystal
High change of enthalpy growth to meet
Non-flammable
(near usage demands of heat
/no explosive
temperature) recovery from storage
system
High specific heat and High nucleation rate to
No phase separation
high density avoid super cooling
Congruent melting Chemical stability
Complete reversible
High latent heat of
freezing and melting
fusion per unit volume
cycles
Small volume changes
on phase
No degradation after a
transformation and
large number of
small vapor pressure at
freezing and melting
operating temperatures
cycles
to reduce containment
problem
Often not provided for
both phases
Table 2.3: Main desirable characteristics of PCMs (Hale et al., 1971)

48
2.3.1 Classification
Several criteria exist for the classification of PCMs based on their chemical composition, melting
temperature, latent heat of fusion, etc. The most frequent categorization divides them to three groups:
organic, inorganic and eutectics (figure 2.5). In fact, the last ones are mixtures of the first two categories
but it is now common to be considered as a specific group.

Figure 2.5: Classification of PCMs (IEA, Annex 23, 2013)

A great number of materials have been used and studied as PCMs in TES applications, including hydrated
salts, paraffin waxes, fatty acids and eutectics of organic and non-organic compounds. Figure 2.6
illustrates a classification of PCMs according to their melting temperature range and melting enthalpy.

A detailed list of commercially available PCMs can be found in Cabeza et al. (2011) publication.

Figure 2.6: Classification of PCMs according to their meting temperature range and enthalpy (IEA,
Annex 17, 2005)

49
2.3.2 Phase change phenomena and properties
Over the years, several problematic phenomena related to PCMs performances have been identified and
studied. This section presents the most common ones.

[Link] The subcooling phenomenon

Subcooling (or supercooling) occurs when the PCM remains in liquid form even when its temperature
decreases below the solidification threshold value (Tf). Normally, the solidification process starts with
the formation of initial crystals (nucleus) and in a second stage, the propagation of these crystals. In
some PCMs however (mostly salt hydrates and some eutectics) it has been observed that the nucleation
rate is low, resulting in a reduction of the nucleus. As a consequence, the PCM remains in liquid form,
solidification and heat release are delayed, and its temperature increases until it gets back to the
solidification value (figure 2.7). Subcooling can have a disrupting effect on the performance of the PCM
unit and must be taken into account in the design phase of TES applications. A solution to minimize this
phenomenon lies in the addition of suitable nucleating agents to the PCM.

Figure 2.7: Subcooling effect during solidification (Cabeza et al. 2008)

[Link] The segregation phenomenon

The segregation phenomenon is observed on salt hydrates PCMs and consists in the decrease of the
storage density over cycling. It occurs due to the congruent melting of most salt hydrates with the
formation of the lower salt hydrate. The effect is non reversible and lowers the thermal properties and
performance of the PCM. Solutions can be found in the inclusion of materials that can alter the
properties of the salt hydrate by preventing the heavier phases from sinking to the bottom. These
materials can be gelling (polymer) or thickening elements and are used to hold the salt hydrate together.

[Link] Thermal life cycle

Thermal life cycle refers to the number of cyclic phase change repetitions (melting-solidification) before
the PCM thermal properties begin to alter (decrease). The thermal life of most PCMs varies between
3000 and 10000 cycles (Hadjieva et al., 1992). Thermal life cycle affects the performance of an LHS unit
and can help determining the potential of its use.

50
[Link] Incompatibility with storage containers: Corrosion or volume change
Some PCMs present another disadvantage as they may not be compatible with the chosen storage
solution. This concerns corrosion and volume changes that could both lead to the degradation of the
container and the leakage of the PCM, rendering the application unusable in the long term.

[Link] Flammability and odor

PCM based materials or applications that are meant to be implemented in the building require
respecting building legislation related to fire risks and the occupants’ olfactory comfort. Some PCMs
present high combustibility values and require special treatment before usage in a real life application.
This can include the incorporation of fire retardant additives in the PCM and conducting fire risk tests.
However these procedures are bound to increase / raise the final price of the PCM and of its
implementation in the building.
Furthermore, some PCMs are not odorless and the inhalation of PCM vapor may cause irritation, both
facts being unacceptable by the occupant. Air-tight containment is thus required to overcome these
problems.

2.3.3 Containment
PCM containment is a very important issue in an LHS unit as it is related to the heat transfer process and
the general rigidity of the application’s structure. As PCMs present low thermal conductivity, proper
containment can help overpass this disadvantage, for example by providing a high exchange surface to
volume ratio or by having a small storage thickness. As discussed earlier, poor containment solutions can
lead to corrosion and/or leakage of the PCM and render the equipment non-operational. Several
techniques are proposed for the proper containment of PCMs, including macroencapsulation,
microencapsulation, direct incorporation, immersion and shaped stabilized PCMs.

[Link] Macroencaplsuation
Macroencapsualtion is perhaps the most commonly used technique. Containers usually larger than 1 cm
in diameter are used to enclose the PCM. The containers can be in the form of tubes, pouches, spheres
and panels; they are usually made of high conductivity and rigidity materials such as aluminum, copper
etc. Macroencaplusation offers the advantages of small leakage risks and of little affection of the
building’s structure. On the other hand, it presents poor thermal conductivity and in some cases
complicated integration into the building materials.

51
 Figure 2.8: Macroencapsulation solutions, from left to right: aluminum profiles with fins, coated
aluminum plate, capsule stripes (Cabeza et al., 2008))

[Link] Microencapsulation
Microencaplsualtion is the technique in which small spherical or rod shaped PCM particles are englobed
in a thin molecular polymeric film, obtaining a diameter smaller than 1 mm. The particles are then
incorporated in any matrix that is compatible with the film, in order to form a homogenous compound in
macro scale. This containment method increases heat transfer as it presents a high surface to volume
ratio and improves cycling stability as the phase separation is restricted to microscopic distances. This
method facilitates the incorporation of PCM into construction materials.

[Link] Direct incorporation

Direct incorporation consists in adding liquid or powered PCM to building constructions materials
(gypsum, concrete or plaster) during production. It presents the advantage of being the most economical
method. Leakage and incompatibility with construction materials are the main disadvantages of this
method.

[Link] Immersion
This technique is based on the sinking of porous building components such as gypsum board, brick,
concrete block, wood and plaster into liquid PCM. The PCM is then absorbed into the component’s pores
by capillary effect. Leakage risk is reported by some researchers. Immersion and direct incorporation
both consist in impregnating PCMs directly in typical construction materials.

[Link] Shape stabilized PCM

Shape- stabilized PCM is the result of a liquid mixture of PCM with a support material, usually high
density polyethylene (HDPE) or styrene-butadiene-styrene (SBS). The mixture is then cooled below the
glass transition temperature of the supporting material, leading to a solid component that can have up
to 80% PCM mass proportion. The observed low thermal conductivity can be increased by adding
materials such as ex-foliated graphite into the mixture composition.

52
Figure 2.9: PCM construction panels: microencapsulated PCM (left), shape stabilized plate (right) (IEA.
Annex 23)

2.3.4 PCM application in buildings

According to L. Cabeza (IEA Task 32, 2005), the first PCM application in the literature appeared in 1975 in
a research byTelkes. Since then, several PCM based applications have been proposed and a number of
publications has been produced to describe them (Farid et al. 2004, Baetens et al. 2010, IEA Task 32
2005, Pasupathy et al. 2008, Parameshwaran et al. 2012, Pomianowski et al. 2013, Sharma et al. 2009,
Soares et al. 2009, Zalba et al. 2003, Zhou et al. 2012, Salyer et al. 1990, Khudair et al. 2004). They can be
divided into two large categories: passive and active ones.

[Link] Passive systems

In passive systems PCM is incorporated into specific components of the building, aiming at several
objectives: increasing thermal inertia, minimizing the usage of mechanical heating/cooling devices,
lowering the indoor temperature fluctuation and preserving the occupant’s comfort. No extra
mechanical heating/cooling means are used for the latent heat storage activation of such systems.

The following list presents the main usage of PCM in passive building applications.

PCM in gypsum boards and wallboards

A vast variety of studies and applications has been realized on composites of PCM impregnated gypsum
boards and wallboards. Their main goal is to increase the thermal mass of lightweight or existing or
renovated buildings, in order to lower indoor temperature fluctuations (figure 2.9).

PCM in concrete
These applications aim to increase heat storage in heavy construction materials by combining the
augmented latent heat PCM capacity with the high density of heavy weight concrete.

53
PCM in bricks
Fewer studies have been performed on PCM-bricks composite constructive applications, showing
nevertheless the interest of such technology in decreasing indoor temperature fluctuations. It has been
noted that this technology is more suitable for warm climates, such as the Mediterranean one.

Floor and ceiling PCM panels

Floor and ceiling PCM panels have also been developed, aiming to absorb heat (solar related, in roofs for
example) and diminish indoor temperature variation.

PCM in glazing and shading shutters

PCM has also been used for optimized glazing and shading components. This technology allows the
absorption of a part of the solar heat and can reduce the indoor air temperature fluctuation.
Furthermore, some PCMs are translucent and can therefore allow light transmission through them and
into the building.

The advantage that emerges from the combination of PCM with the construction materials in question is
that the final component offers large areas for heat transfer and LHS potential can therefore be fully
exploited.

[Link] Active systems

Active PCM applications constitute an improved technology for LHS utilization in buildings. Contrary to
passive applications, thermal storage can be activated by a supplementary mechanical system (fan,
pump, etc), enhancing heat transfer performance and allowing stocking or destocking heat/cold on
demand. They can be operated on specific time slots, being able to correlate the needs with the energy
supplies and apply in such a way peak power reduction strategies.
In some active PCM systems, the same medium is used both for the heat transfer and storage, whereas
in others heat transfer is performed by a first medium (usually air or water) and a second one is used for
heat storage (PCM).
Several active PCM applications have been proposed and the main available technologies are briefly
presented.

PCM integration with HVAC units

A series of studies has been performed in order to integrate PCM technology in HVAC systems. This is
mainly performed in two ways: incorporating either a heat exchanger or a PCM panel board in the HVAC
system either PCM. The installation of the PCM technology can be performed in the false ceiling or in the
ventilation plenum. This fact usually imposes restrictions in terms of the volume of the unit and the final
PCM quantity.

A PCM based heat exchanger usually consists of a set of receptacles that are used to contain the PCM
(macro encapsulation); a fluid (air or water) is used as the heat transfer medium. Design considerations
include heat transfer optimization between the heat transfer fluid and the PCM. Concerning the PCM
panel boards, this consists in the use of passive PCM technology (as presented in the previous section),
with the advantage of the coupling with an active mechanical system.

54
PCM integration in HVAC systems is usually performed to enhance the cooling or heating performance of
mechanical ventilation units and for peak power reduction needs.

PCM in construction: ceiling gypsum and radiant floor panels

A thermally activated ceiling gypsum panel has also been developed allowing heat gains of up to
40W/m2 and an increase of the thermal mass of light weight buildings. An incorporated water capillary
tube system allows the active control of the thermal storage. Active PCM applications can also be found
in radiant floor panels. Granulated PCM was employed together with integrated pipes for heating and
cooling. Results demonstrated the benefits of this application in summer conditions.

A table that summarizes passive and active LHS applications in buildings is presented by Parameshwaran
et al. (2012), focusing on the applications’ potential, integration within the building, and functional
aspects.

2.3.5 PCM enhancement techniques

[Link] Heat transfer

As discussed earlier, one of the main disadvantages of PCMs is their small thermal conductivity values:
0,2 W/(mK) for most paraffins and 0,5 W/(mK) for salt hydrates. These values can be insufficient for
rapid charging/discharging phases of LHS systems. For example, it is common in buildings to have daily
operation cycles as heat or cold is stored in the night and is released during the day. Low conductivity
values can lead to limited charging of the unit and poor performance of the system. Therefore, several
enhancement techniques have been proposed in order to increase the heat transfer exchange, mostly
aiming to obtain higher PCM conductivity values:

 Microencapsulation of PCMs using a high conductivity material such as graphite, silicon carbide,
copper, etc.
 Dispersion of high conductivity material particles (copper, steel, aluminum, etc.) within the PCM
 Addition of high conductivity material into the macroencapsulation container
 Extension of the surface area using fins
 Design optimization to obtain the highest surface to volume ratio possible

Nevertheless, these techniques can have a positive effect on conductivity values but on the other hand
they degrade the unit’s heat storage capacity due to the replacement of part of the PCM with sensible
storage material.

[Link] Fire retardation

PCM enhancement also concerns fire risks limitation, as building regulations become stricter in this field.
Salyer and Sircar (1990) propose the following solutions to fire retard PCM imbibed plasterboard:

 Adding alternate non-flammable surface to the plasterboard (e.g., aluminum foil and rigid
polyvinyl chloride film).

55
  Sequential treatment of plasterboard, first in PCM and then in an insoluble liquid fire retardant
(e.g., Fyrol CEF). The insoluble fire retardant displaces part of the PCM and some remains on the
surface, thus imparting self-extinguishing characterization to the plasterboard.
 Using brominated hexadecane and octadecane as PCMs. It is anticipated that when these
halogenated PCM compounds are combined with antimony oxide in plasterboard, the product
will be self-extinguishing.
 Fire retardant surface coatings may be used to prevent effectively the wicking action of the
plasterboard paper covers.

2.3.6 Limitations/barriers
Despite the mentioned advantages and the numerous research activities, PCMs have not yet reached
large scale commercial usage. Most of the realized studies limit themselves to small laboratory
experimental units and few investigations have been performed in actual buildings and real life
operation scenarios. Commonly, the developed concepts remain in the prototype stage and in some
cases unrealistic scenarios are tested that lead to overestimated results. Moreover, these studies fail to
present an economic analysis of the project or a life cycle assessment of its components. The main
factors explaining the so far limited utilization of PCM based technology are as follows:

 Elevated price of the material and its implementation in the building

 Difficulties concerning the proper containment
 Physical phenomena that deteriorate the initial PCM performance
 Flammability risks and odor risks
 Absence of legislation concerning their implementation in the building

2.4 HVAC integrated PCM applications

In this section studies performed by various laboratories and associated with active LHS applications
integrated into HVAC systems are presented. In these types of systems, the needed heat or cold is stored
in specific apparatus (heat exchangers), thermally contained from the rest of the building and used only
on demand and not automatically (i.e. active LHS applications).
Firstly, a detailed description of each case study is performed focusing on the experimental apparatus
description, on the main results and the development of a numerical model (if any). Secondly, useful
conclusions drawn from this study are presented in order to summarize the main elements, the
operation principles, the design methods and the philosophy of active LHS ventilation systems.

2.4.1 Demonstration examples

[Link] Borderon et al. system

Borderon et al. (2012) proposed an LHS system aiming to cool a house by using the freshness of the cold
night air. It is composed of horizontal PCM plates placed in a heat exchanger and coupled with a
ventilation system. A fan forces the air through the exchanger so that the PCM absorbs or releases heat.

56
Indoor air is cooled by passing through the exchanger during the day and outside air is used to solidify
the PCM during the night. A diffuser is used to equally distribute the air between each plate. The PCM
used is the Energain® compound, a mixture of 60% microencapsulated paraffin in an ethylene based
polymer, with a melting point around 21°C and a latent heat of 72 kJ/kg for 5 to 30 °C. It is used in the
form of plates with dimensions of (1,20x1,20x0,05) m, each face of the plates being covered with an
aluminum foil preventing leakage and fire risks.
The exact composition of the proposed heat exchanger (as seen in figure 2.10) is a first insulation layer, a
combination of several Energain plates piled up one another, an air layer, a second PCM layer which
doubles the thickness of the first one, a second air layer, a third and last PCM layer identical to the first
one and finally a second insulation layer.

Figure 2.10: Borderon et al. heat exchanger unit

The two air layers are instrumented: the upper surface of the lower air layer and the lower surface of the
upper air layer are equipped with flow meters and thermocouples (Type K, Chromel/Alumel). An
omnidirectional (TSI™, 0-10 V) hot-wire anemometer (TSI™, 0-10 V) is placed in each air layer near the
outlet. Two relative humidity sensors are placed on both sides of the prototype; they consist of 4-20 mA
sensors associated with a 500 Ohm resistance and an output signal of 0-10 V.
In order to measure the heat flux between PCM and air, four flux meters were installed, two for each of
the instrumented sides. The less than 1 mm thin sensors send an electrical signal that is proportional to
the heat flow. For each of the flow meters, an air and a surface temperature measurement is made.
Other air and surfaces temperature measuring thermocouples are spread over the plates. Air
temperature is measured at the inlet and outlet of the exchanger and a differential thermocouple is used
to directly measure inlet and outlet difference temperature. Data acquisition is performed with a
National Instruments device and Labview software.
The experiments conducted with the developed heat exchanger allowed to issue temperature, relative
humidity and airflow profiles.

57
Parallel to the experimental study, a bidimensional Matlab model was created based on the finite
difference method and the equivalent heat capacity for the PCM. Experimental data are used to validate
the numerical model, concerning convective coefficients and the hysteresis phenomenon.
Figure 2.11 illustrates the confrontation between experimental and simulated results for the outlet air
temperature.

Figure 2.11: Outlet air temperature experimental and simulation results (Borderon et al.)

The validated model is then coupled to a TRNSYS model of house in order to evaluate the proposed
system’s behavior for different configurations and climates.

[Link] Zalba et al. system

Zalba et al. (2004) developed a system in order to test the performance of PCM in a cooling system that
allows storage of outdoor coolness during the night, to supply it to the indoor environment during the
day.
The proposed unit (figure 2.12) is a closed air circuit consisting of a fan for circulating the air, a device for
heating and cooling the air at the desired temperature, a flow meter and the thermal energy storage
system. The last one consists of flat plate rectangular encapsulation vesicles made with methacrylate to
allow visualization of the phase change.

58
 Figure 2.12: Zalba’s et al. experimental set up (left) and heat exchanger unit (right)

Two PCMs were selected:

- A molecular alloy consisting of 34% C16 and 66% C18, with a latent heat of 152 kJ / kg and a melting
range between 19.5 °C and 22.2 °C.
- The RT25 Rubitherm paraffin, with a latent heat of 164 kJ/kg and a melting range between 18,8°C and
24,1 °C.
Inlet and outlet air temperatures were measured using five calibrated Pt100 sensors at the entrance and
exit of the unit. Air flow was measured using a calibrated flow meter and customized software was used
for data acquisition and evaluation.
The two following figures (2.13 and 2.14) show the experimental results for Zalba’s et al. proposed unit.
Figure 2.13 illustrates the exchanged heat and figure 2.14 the exchanged energy, both for the melting
and solidification processes.

Figure 2.13: Zalba’s et al. unit’s heat transfer during phase change for different configurations

59
 Figure 2.14: Zalba’s et al. unit’s heat exchange during phase change for different configurations

Studies performed on the designed system showed that for the solidification and the melting processes
the most influential factors are the thickness of the encapsulate container, the inlet temperature of the
air, the air flow, and the interaction thickness / temperature. The researchers also performed a viability
analysis that showed that this kind of system can be developed in a real installation, regarding technical
feasibility as well as economic advantages (when compared to existing cooling systems).

[Link] Arkar et al. system

Arkat et al. (2006) developed a latent heat thermal energy storage device in order to improve the cooling
potential of a mechanical ventilation system. The device has a cylindrical form and is filled with PCM
encapsulated spheres; the PCM used is the RT20 paraffin from Rubitherm GmbH.
For the optimization of the LHS device several parameters were taken into account: PCM characteristics,
the ratio between PCM mass and airflow, and the ratio between the length and the diameter of the
exchanger.
The developed heat exchanger prototype (figure 2.15) consists of a cylinder with a diameter of 0,34m
and a height of 1,52m, filled with polyethylene spheres with a diameter of 50mm containing PCM. Arkar
et al. investigated a system with two heat exchangers integrated in the mechanical ventilation system of
a low energy building. The first heat exchanger was used for cooling the ambient air when it was hotter
than indoor air temperature. The second one was used when additional cooling was needed, using re-
circulated air when indoor air temperature was higher than a set point value.

60
Three operation modes were considered: a simple ventilation mode (constant airflow through the day),
night cooling mode (increased airflow during the night) and free cooling mode (using the two heat
exchangers).

Figure 2.15: Heat exchanger developed by Arkar et al.

Experiments were conducted in order to study the phase change process of the PCM and verify an
adapted numerical model. Figure 2.16 shows the comparison between experimental and simulation
results during charging and discharging phase.

Figure 2.16: Experimental and simulation confrontation for the outlet air temperature of Arkar’s et al.
unit during the melting stage (left) and the solidification stage (right)

The model was then used to form a multi-parametric thermal response function that calculates the heat
exchanger outlet temperature. Finally, a TRNSYS simulation program was developed, using the
exchanger outlet temperatures as input and being able to predict the fresh or re-circulated air supply
temperature.

61
With the numerical tool in hand, several simulations were performed altering the ventilation strategies,
the air flow rates and the size of the heat exchanger. The investigation demonstrated the advantages of
the proposed system in terms of reduction of the mechanical ventilation size, thermal comfort and fresh
air supply.
As a perspective of their work, Arkar et al. mention the usage of the developed system during the
heating season with the installation of a roof solar collector.

[Link] Nagano et al. system

Nagano et al. (2004) proposed a latent heat storage system, in which PCM granules consisting of porous
media and paraffin wax are applied. In particular, the GR25 PCM made by Rubitherm GmbH was used to
create granules in which paraffinic hydrocarbon was absorbed into granulated porous ceramics. The
particle had a diameter from 1 to 3 mm and its composition was 65% ceramics and 35% paraffin wax by
weight. The melting and the solidification points are 23.2 and 24.1 °C, respectively and latent heat is 45.3
kJ/kg.
The experimental setup (figure 2.17) consists of a cylindrical column made of chloroethene, with an
internal diameter of 50 mm and a wall thickness of 5mm. The PCM granules were introduced in the
column from 120 to 350 mm from the bottom. A constant air chamber supplied air to the bottom of the
column with the use of an air pump. Passing through the PCM packed bed, the air was exhausted from
the top of the column. A chloroethene plate with air holes was used to rectify it, placed at a height of 60
mm.

Figure 2.17: Nagano et al. experimental unit

62
PCM temperature measurement were performed using T-type thermocouples at 11 points and a
propeller vane anemometer was used for air velocity measurements which led to air flow rate
calculation.
Figures 2.18 illustrates temperature variations at different points of the PCM packed bed (1 cm interval)
for various face velocities and supply temperatures.

Figure 2.18: Temperature variation for different experimental configurations, Nagano’s et al. unit

63
The experimental unit was used to observe the melting and the solidification process of the PCM and to
verify DSC measurements. Further remarks were made considering the exchanged heat flux between the
PCM granules and air. A numerical model was developed and validated with the use of experimental
results.

[Link] Lazaro - et al. system

Lazaro et al. (2009) designed an experimental setup for free cooling purposes. Studies were made for
different air to PCM heat exchangers, using a closed air loop setup to simulate indoor conditions. The
developed system (figure 2.19, left) consisted of two different heat exchanger prototypes, an inlet air
conditioner for various operating modes of simulation, air ducts and gates as well as numerous sensors
to measure air/PCM temperatures and air flow.
The first prototype used aluminum pouches filled with an inorganic PCM (figure 2.19, right). A metallic
grid was used to assure the vertical position of the apparatus. The air flowed parallel to the pouches
from bottom to top. Tests with constant inlet air temperatures showed that the cooling rates were very
low and that the total melting times were double the melting designed time. It was discovered that this
was due to PCM leakage from torn pouches.

Figure 2.19: Lazaro et. al LHS system and first encapsulation method: aluminum pouches

In view of these events, a second prototype was developed using aluminum panels (as shown in figure
2.20) with organic PCM. As with the first prototype, the configuration was also vertical but this time air
flowed from top to bottom. The PCM used for this configuration presented lower thermal conductivity
than the first prototype and the total stored energy was also lower.
Concerning instrumentation, the inlet and outlet air temperature differences were measured with a six
junction thermopile and air flow was calculated using fan velocity measurements. Inlet and outlet air
humidity was also measured, showing a maximum variation of 0,006 kg/kgda.

64
 Figure 2.20: Lazaro et. al LHS second encapsulation method (aluminum panels)

The two following graphs (figure 2.21) show experimental results concerning the cooling rate for a
constant rise of inlet air temperature and constant heating power. The second prototype resulted in a
cooling capacity over 3KW for approximately 1,5 hour or approximately 1KW for more than 3h. This
information can be useful when designing the optimal operation mode for different applications.

Figure 2.21: Cooling rate evolution with constant rise of inlet air temperature (left) and constant
heating power for Lazato’s et al. prototype 2

Lazaro et al. investigated two cases of LHS units. The second unit’s optimized design led to higher cooling
power and shorter melting rates. The researchers conclude that enhancement in the design of the heat
exchanger can be more effective than the use of a higher conductivity PCM.

[Link] Dolado et al. system

Dolado et al. (2011) investigated the performance and thermal cycling of a real-scale PCM-Air heat
exchanger. The unit in question was a modified version of Lazaro’s et al. (2009, figure 2.19) experimental
set up. Modifications included the creation of a single branch closed loop, the relocation of the chiller

65
and the removal of the second branch of the duct system and allowed the realization of thermal cycles
(figure 2.22).

Figure 2.22: Dolado et al. modifications in Lazaro et al. system

Approximately 135 kg of the organic RT27 PCM were used in the LHS unit, macroencapsulated in 216
aluminum rigid slabs. These slabs were then assembled in a way to form parallels walls, separated by a 1
cm thick air layer. The LHS unit fan can be operated for three different speeds, varying from 675m3/h to
1550 m3/h, whereas the fun chiller was single speed.
Dolado et al.’s investigations led to the full thermal characterization of the LHS unit and the development
of empirical designing tools. They observed the capacity of the proposed system to assure PCM
solidification for specific time periods as well as the repetitive thermal cycling performance of the unit.
Figure 2.23 shows the evolution of the plate’s surface temperature for a melting-solidification cycle.
These temperatures were used as an indicator of the phase change advancement. A full melting process
occurred in less than two hours with a 9.2°C temperature difference between inlet air temperature and
average phase change of the PCM (ΔTair-PCM).

Figure 2.23: Inlet and outlet air temperature, ambient air, and surface temperature distribution for a
full cycle of Dolado et al. proposed unit

66
Investigations were also performed for the evolution of the temperature of the PCM and encapsulation
material for a single slab, with measurements occurring at three different points. The points are denoted
as PCM Up, Middle and Down and Plate, Up, Middle and Down in figure 2.24. A 3°C temperature
difference can be observed between PCM and surface temperature during the phase transition. It is
noted that both melting and solidification processes start in the upper part and move towards the lower
part of the plate.

Figure 2.24: Air, plate and PCM temperatures of a single plate for a full cycle, Dolado et al. proposed
unit

Characterization results were used to develop rules of thumb and an empirical model that can be used
for LHS designing purposes.
Parallel to that, the experimentally obtained results also served for the validation of a numerical model
aiming to be used as a designing tool for different applications of LHS systems.

[Link] Lin et al. system

Lin et al. (2007) developed a PCM based floor heating system with fan assistance for heating and peak
power reduction purposes. Their aim was to surpass the limitations of a passive under-floor heating
system by adding thermal storage to it.
The system was integrated in an experimental house and consists of an insulation layer
(polystyrene120mm), electric heaters, PCM (15mm), an air layer (50mm), floor cover material (40 mm),
cylindrical supports, air inlets and air outlets with fans (figure 2.25).

67
 Figure 2.25: Lin et al. experimental house and enhanced floor heating system.

The PCM used consisted of 75% paraffin and 25% polyethylene; the fusion point of paraffin is 52°C and
its heat of fusion is approximately 200 kJ/kg.
Concerning instrumentation, 18 thermocouples were used in order to measure air temperature at
different heights and places of the house and a data logger was used for their recording. The tracer gas
technique was used to calculate the air exchanges per hour and solar radiation was measured with a
thermal radiometer.
The thermal behavior of the heating system and the instantaneous indoor temperature of the room
were investigated. The operating principle was as follows: for a period of five days the electrical heating
system was turned on and for the next five days they were turned off. For the first five days the electrical
heaters worked from 23:00 to 07:00 and were shut down for temperatures higher than 65°C and for the
period 07:00 to 23:00. Measurements were made for the indoor and outdoor temperature, solar
radiation and electricity consumption every 5 minutes. Air outlets and inlets were opened and the air
supply fans were operated from 09:00 to 16:00.
Results demonstrated that indoor temperature can be effectively increased with the use of the
enhanced air supply during the working hour period. Total electrical energy consumption was shifted
from peak to off-peak period, resulting in important economic benefits due to lower night electricity
cost.
A previous numerical model was modified for this case study. With the use of experimental data
investigations were made concerning the influence of various factors such as air supply velocity,
thickness and fusion temperature of the PCM, thermal conductivity coefficient, heat of fusion as well as
floor and air layer thickness. Indoor air temperature was calculated while changing one of these factors
at a time, the final goal being to optimize the system’s design for different conditions.
Figure 2.26 illustrates the comparison between experimental and simulation results of the indoor air
temperature with (b) and without (a) air supply, and of the PCM surface temperature with (d) and
without (c) air supply.

68
 Figure 2.26: Lin et al. experimental vs simulation results for different configurations (a) indoor
temperature without air supply, (b) indoor temperature with air supply, (c) PCM surface temperature
without air supply and (d) PCM surface temperature with air supply

In the end, Lin et al. stated that their study showed the promising perspectives of the proposed system
for different climate regions while providing simple and effective design and control methods.

[Link] Turnpenny et al. system

Turnpenny et al. (2001) proposed a cooling ventilation system aiming to reduce energy and
economiccosts. The prototype consists of a number of PCM embedded heat pipes, circularly installed
and mounted on a support; a fan is also installed in the center of the circle, just above the heat pipes
(figure 2.27).

69
 Figure 2.27: Turnpenny et al. LHS unit, overall view (left) and heat exchange pipes (right)

The PCM used is the Na2SO4.10H2O salt with a melting temperature of 21°C and a latent heat fusion of
198KJ/kg.
The operating principle is as follows: during the night cool air was directed towards the upper part of the
unit with the use of a movable board. The ceiling fan blew this air downwards so as to pass over the heat
PCM pipes and cool them. During the day, the ceiling fan was also used to blow air downwards, cooling it
while it passed through the exchanger unit and absorbing internal and external heat gains.
Two PRT sensors on each side (unit and control) were used for temperature measurement and another
one was placed outside of the vent. Three stainless steel thermistors were placed in the heat pipe-PCM
unit in order to supply spot PCM temperatures at three different distances from the container wall. In
this way, it was possible to estimate the phase change process. Air speed was measured by a movable
hot-wire anemometer.
Figure 2.28 illustrates the results obtained over a 32 hour period for average PCM temperature, average
room temperature and external temperature. Figure 2.29 shows upstream-downstream and upstream-
PCM temperature difference. In both cases, the first cycle was performed with a 1 kW heat input and the
second with a 2 kW one. Complete melting is assumed to occur when the mean PCM temperature rises
above 23°C; with a 2°C difference between inlet and PCM temperature complete melting occurs after an
8 hour period whereas with a 3.5°C difference the time needed is 3 hours.
Studies were made to assess the overall cooling performance of the unit during the summer, including
temperature monitoring, melting behavior of the PCM, cooling potential, tilting and covering of the heat
pipes, different air passage configurations and heat transfer rates measurements.
Results showed that the system managed to ensure an adequate rate of heat storage and prevented
overheating phenomena for a typical UK summer climate.

70
Figure 2.28: Air, PCM and external temperature for 1 and 2 kW heat input, Turnpenny et al. LHS unit

Figure 2.29: Upstream-downstream and upstream-PCM temperature difference for 1 and 2 kW heat
input, Turnpenny et al. LHS unit

71
2.4.2 Comparative table of existing studies

Compared Parameters
Heat
exchanger
system System
PCM (Choice and
Objective integration in Model
encapsulation)
the building
Energain®, (60% Implicit finite
Borderon microencapsulated paraffin
Free cooling False ceiling difference
et al. in an ethylene based method
polymer) / Macro
encapsulated
Molecular alloy (C16 and 66%
Zalba et al. Free cooling C18) and RT25 Rubitherm - Empirical
paraffin /
Macro encapsulation
Mechanical Finite difference
Arkar et al. Free cooling RT20 paraffin Rubitherm ventilation method
system

PCM performance Paraffin GR25 Rubitherm and Enthalpy

Nagano et al. and numerical model porous media / Macro -
method
validation encapsulation

Inorganic PCM / Macro

Lazaro et al. Free cooling - -
encapsulation

Characterization of
the unit and Organic RT27 Rubitherm /
Dolado et al. - Empirical
empirical designing Macro encapsulation
tools elaboration
PCM enhanced floor 75% paraffin and 25% Enthalpy
Lin et al. Floor
heating system polyethylene method

One-
Turnpenny dimensional
Free cooling Na2SO4.10H2O salt Ceiling
et al. mathematical
model
Table 2.4: Comparative table of the different LHS units

72
2.5 Design and operation of active LHS ventilation systems
The presentation of eight active LHS examples was performed in order to study existing experimental
and numerical studies and draw interesting remarks and conclusions, useful for the elaboration of this
thesis. In this section we try to summarize all the necessary information concerning the design and
operation of such systems.

2.5.1 Main elements of an active latent heat storage system

LHS systems are usually composed of four basic elements:
 The heat exchanger that contains the macroencapsulated or microencapsulated PCM; it is the
main component of the system and the place where heat/cold is stored in latent (and sensible)
form.
 A circuit that allows the flow of the heat transferring fluid (mostly air, sometimes water).
 An electrical or solar heat source that is used to heat the circulating fluid.
 A fan or a pump that regulates the flow of the circulating fluid in the system.
In addition to these elements, various sensors are usually installed in different part of the system in
order to investigate the inlet/outlet air temperature, PCM temperature, airflow etc.

2.5.2 Operation principles

LHS applications are usually used to enhance the cooling or heating performance of mechanical
ventilation systems and for peak power reduction needs.

Cooling of the building

During daytime, hot air flows through the heat exchanger at a higher temperature than the PCM’s
melting point. The PCM, initially in its solid form, absorbs heat from the air while obtaining its liquid
form. The cooled air is then reintroduced into the building. In order to solidify the PCM, during the night
exterior air is circulated at a temperature lower than the PCM’s melting point.

Heating of the building

The principle is the same but the process is reversed. During daytime, hot air originating from the
interior of the building or produced from a solar or electrical power system, flows through the heat
exchanger of the LHS system and provides heat by melting the PCM. When heat is needed, cold or
interior air is forced through the exchanger, capturing heat from the PCM, while solidifying it. The heated
air is then reinjected into the building.

Peak power reduction

For both cooling and heating applications, the active thermal energy storage systems can be operated in
view of reducing peak power consumption. Heat or cold is stored in the developed system during an off-
peak period, preferably during the night (for lower electrical cost) or during the time when the COP of a
heat pump is optimum (for example, when indoor and outdoor temperature difference is minimal).

73
Control strategies can be developed in order to predict the cooling/heating needs of the building and
adjust the usage of the system accordingly.
To sum up, the latent heat storage systems can be integrated into any mechanical ventilation system.
The benefits of such technology can be summarized as follows:
 Storing larger amount of energy per unit mass which allows a smaller temperature swing (better
thermal comfort)
 Utilizing storage energy produced from passive or active solar heating/cooling and waste heat
which comes from exothermic processes;
 Enabling the use of energy at lower cost during off-peak periods for storage and discharge at
times when full rate would otherwise be charged.
 Allowing the use of a smaller HVAC system size and more efficient operation
 Reducing furnace/heater cycling

2.5.3 Design philosophy of a LHS system

This section tries to summarize the various parameters that influence the design philosophy of a TES
active system.
One of the most important factors when designing such a unit is that the developed system must ensure
a maximal heat exchange rate between the heat transfer fluid (air) and the latent heat storage medium
(PCM). In general, one must bear in mind that the available periods of time for the solidification/fusion
of the PCM can be limited. The geometry of the heat exchanger as well as the choice for the PCM storage
(microencapsulation or macroencapsulation) play a significant role in this matter. Materials with high
conductivity are preferred, as long as their price is not too high. In some of the macroencapsulated
cases, fins have been used to enhance heat exchange between the air and the PCM container. The
geometry of the heat exchanger also affects the pressure drop of the LHS system and as a result, the air
pumping requirements and the electrical energy consumption.
The design of a LHS system also highly depends on the heat storage needs of the place where it is
installed. Different PCMs with higher or lower latent heat values can offer solutions in this matter.
PCM thickness and airflow rates are also important aspects to take into account. Researchers often
realize experimental and numerical studies to determine the optimized values for these parameters.
Another important factor is the size of the designed unit; smaller units are preferred as they can be
implemented more easily in the building (in ventilation shafts, false ceilings, etc). A compromise must be
found between heat storage capacity and the size of the proposed unit.
Attention must be given to the viability of the system, knowing that some PCMs present corrosive
properties. In addition to that, the proper containment of the PCM can be a difficult task as dilatation is
observed between phase changes. In some case studies PCM leakage has been observed due to this
phenomenon (Lazaro et al., 2009).

74
When storing heat in the unit, it is important to consider the outlet air temperature due to risks of
overheating the coupled building. A solution to overcome this issue can be the use of a closed loop air
circuit or specially conceived control strategies.
The implementation of an LHS system in an inhabited building raises some more aspects to consider: fan
noise, PCM odors or fire hazards will not be accepted by the occupants.
LHS system design is an elaborate task with various and controversial factors to consider. Experimental
and empirical/numerical studies are useful tools for the development and further improvement of such
units.

2.5.4 Design Methods

In the introduction to this chapter, the benefits and the increasing need of LHS systems were discussed,
either as passive components of the building or active ones. In both cases, the development of advanced
design methods and tools is needed, with the perspective of a better coordination between building and
LHS design. This will lead to a more efficient integration (hence a better exploitation) of advanced LHS
systems and can contribute towards the overall goal of net zero energy buildings.
The study of several LHS units integrated into HVAC systems which is performed in the latter part of this
chapter, shows a lack of knowledge and design tools for a global and/or mutual concept of building and
LHS system. Researchers have proposed experimental units together with empirical or numerical
models. The models are then used for further amelioration of the unit and the investigation of different
configurations. In some cases the models are coupled with building simulation programs (TRNSYS,
EnergyPlus,…) for a more global approach . However, in most case studies, the outcome of the research
focuses on the specific unit and cannot be expanded to a generalized context of integration of LHS
systems in the building. This is due to the dependence of the proposed models on the characteristics of
each study, in terms of the LHS unit (dimension, building integration, PCM properties: non-linear
behavior, moving liquid-solid fraction, ...), as well as the experimental site (heating/cooling demands,
dimension, insulation, usage, occupancy, …).
Further research could be made in this area, taking into consideration the large number of parameters
covering the integration of LHS units into the buildings. The ultimate goal is to develop new and
validated design tools that can cover the whole system Building + LHS system.

2.6 Conclusion
PCM technology has been studied and used in researches for over three decades in the built
environment for heat storage purposes. The benefits of tis implementation in the building in terms of
thermal comfort conditions, energy consumption and peak shifting have been demonstrated through a
number of publications.
This chapter, presented various substances used as PCMs , along with information concerning their
containment, enhancement techniques, limitations and implementation in the building. The interest of

75
this study lies lied in HVAC integrated heat storage systems and in that perspective several case studies
were presented. The remarks that were noted were used as a guide for the development of an
experimental PCM-Air heat exchanger and are presented in the following chapter.

76
Chapter 3
Development of an experimental PCM-Air heat exchanger

77
Chapter 3 : Development of an experimental PCM-Air heat exchanger

3.1 Introduction
The study presented in this thesis aims to provide solutions to the problem of increasing peak power
demand in France. As discussed in the previous chapter, one of the possible solutions lies in the storage
of heat into an exchanger unit and then using it to shift the consumption linked to electrical heating to
an off-peak period.
This chapter describes in detail the development of such a system: TES technology choice, design
considerations, components, instrumentation, data acquisition and assembly of the unit. In a second
phase, a series of characterization tests is illustrated in order to obtain an overview of the system’s
behavior through different conditions. Finally, the coupling of the developed system to an experimental
room is presented along with a preliminary control strategy.
The development of the experimental system is necessary for the validation of the numerical model
(described in chapter 4) and in a later phase for the implementation of control strategies in realistic
scenarios.

3.2 Experimental set up description

3.2.1 Thermal energy storage as a solution

In association with the Stock Air project a specific need was identified: developing an experimental unit
that can store a sufficient amount of heat in order to substitute the elevated electrical heating
consumption.
More specifically, the application was conceived for a low energy dwelling, in a way that it could be
easily integrated into a HVAC system and be able to provide a power supply of 2KW over a period of 2
hours, which corresponds to the late afternoon (18:00-20:00) peak power demand.
These objectives resulted in a number of design considerations, sometimes conflicting with each other.
They can be depicted as follows:
 Small volume (for HVAC integration)
 Sufficient energy storage to satisfy the 4KWh need
 Specific operation temperature range
 Elevated heat transfer for a rapid stocking/destocking process
 Rigid structure that will not affect or degrade the occupants’ comfort/health
Among the available energy storage technologies presented in chapter 2, thermal storage was chosen as
the one that best fits these considerations. Furthermore, PCM technology was decided to constitute the
storage medium as it offers an increased heat storage to volume ratio.

78
3.2.2 Elaboration of a PCM-Air heat exchanger
The above considerations led to the need for the development of a PCM – Air heat exchanger that will
constitute the central element of the experimental setup and the place where the heat storage will
occur.
An analysis was performed by EDF laboratory Matériaux et Mécanique des Composants (partner of Stock
Air project), comparing the properties of various materials that could be suitable for the proposed unit.
The paraffin Microtek 37D was selected due to its advantageous properties (table 3.1). In particular it
presents a melting/solidification temperature range around 37°C, a dilatation of 10% between phases
and a latent heat of fusion of 230 KJ/kg. In its solid form it is a white powder (figure 3.1) whereas when
melted it becomes a colorless liquid. It presents a mild wax odor. It is fabricated by Microtek
Laboratories, Inc. based in Ohio, USA.

Melting heat of Melting/

Thermal
Density Heat capacity fusion / crystallization
State conductivity
(kg.m-3) ([Link]-1K-1) crystallization temperature
(W.m-1K-1)
([Link]-1) range (°C)

Solid 870 0.24 2.36 230 33 - 39,6

Liquid 779 0.18 2.45 226.8 35,5 - 30,3

Table 3.1: Properties of the Microtek 37D paraffin

Figure 3.1: Microtek 37D in solid form

79
The heat exchanger was designed so that it could be integrated into a HVAC system, for example in a
ventilation plenum, a false ceiling or the ceiling over a corridor. In regard to this, its size had to be
modest and was established as follows: 1,05m in length, 0,80m in width and 0,25m in height.
The paraffin is macro-encapsulated in 34 aluminum plates, typical screed boards used in construction
and masonry, with a (1x0,10x0,018)m dimension and openings at the two extremities (figure 3.2). They
were selected because of the high conductivity value of aluminum on one hand and its relatively low
price and ease of access on the other hand. The paraffin was inserted in liquid form and the plates were
sealed using wooden caps and an elastic glue. The plates were assembled by two and then placed
parallel one to another (figures 3.4 and 3.5).

Figure 3.2: Aluminum plates used for PCM containment

Weight measurements were made before and after the encapsulation of the paraffin, in order to obtain
the exact PCM mass. The result was a total of 31,44 kg of PCM (table 3.2).

Plate N° Tare weight (g) Laden weight (g) PCM weight (g)
1 1572,5 3402 1829,5
2 1452,5 3282 1829,5
3 1452,5 3316 1863,5
4 1452,5 3284 1831,5
5 1452,5 3248 1795,5
6 1452,5 3377 1924,5
7 1452,5 3300 1847,5
8 1452,5 3322 1869,5
9 1642,5 3437 1794,5

80
 10 1452,5 3325 1872,5
11 1452,5 3329 1876,5
12 1452,5 3377 1924,5
13 1452,5 3306 1853,5
14 1452,5 3295 1842,5
15 1452,5 3271 1818,5
16 1452,5 3322 1869,5
17 1572,5 3373 1800,5
Total PCM mass: 31443,5 g
Table 3.2: PCM mass calculation

Aluminum fins, 0,019m thick, were introduced between the plates to increase the heat exchange (figures
3.3, 3.4 and 3.5). Three of the plates were equipped with additional pierced components that were used
to introduce temperature sensors at variable depths inside the PCM, in three different parts of the
plates: inlet, middle and outlet areas (figures 3.2 and 3.5). Polystyrene panels enclosed the PCM
containers in order to thermally insulate it from the surrounding environment. Finally, wooden boards
were used to assemble and ensure the stability of the structure (figures 3.5 and 3.6).

Figure 3.3: Aluminum fin

81
 Figure 3.4: Fins between the PCM plates

Instrumented plates

Insulation panel Fins Plates Wooden board

Figure 3.5: The heat exchanger during construction phase

82
Figure 3.6: The heat exchanger through the assembly process

83
3.2.3 Experimental platform
Furthermore, the utilization of the heat exchanger necessitated the construction of an experimental
platform (figure 3.8), composed of the following components:
- Two centrifugal fans (controllable 0-10 Volts, maximum consumed power 121 W)
- Two electrical duct heaters (1KW each) with thermal protection (figure 3.9)
- An air diffuser and an air collector (fabricated out of wood) (figure 3.9)
- Several ventilation hoses and fittings to interconnect the previous elements
- An additional fan heater (2 KW) (occasional use for supplementary power)
The developed experimental platform ensures the circulation and heating of the air through the heat
exchanger under specific conditions (temperature and airflow rate) and is used to couple the exchanger
with a room or a building.

Figure 3.7: Electrical duct heater (left) and air diffuser (right)

84
 Figure 3.8: Experimental platform

3.2.4 Instrumentation, data acquisition and control-command interface

The use of the experimental platform necessitated the elaboration of a monitoring-control system. This
system is composed of a chain of components that are used for the observation of thermal properties,
data acquisition and control-command.

[Link] Instrumentation
Several sensors were installed in different parts of the platform and the heat exchanger, allowing to
monitor the condition of the heat transferring fluid (air) and the heat storage material (PCM-paraffin).
Concerning the experimental platform, temperature and relative humidity sensors were installed before
and after the heat exchanger in order to measure the inlet and outlet properties of the air. For the
temperature measurement, platinum resistance temperature PT100 class A sensors were used (figure
3.9). They were selected due to their small cylindrical shape (1,6mm) and their precision (±0,15°C at 0°C,
±0,35 °C at 100°C). Relative humidity sensors offer a precision of ±2%.

Figure 3.9: PT100 temperature sensor

85
The airflow rate measurement was performed at the outlet part of the exchanger using a Nozzle Pitot
Flow Sensor, with a ±0,5% reading accuracy (figure 3.10). A Pitot Tube measures the fluid flow velocity
based on the Bernoulli Equation, considering that no change in hydrostatic pressure (i.e. no elevation)
occurs. The instrument performs pressure measurements at two points: one pointing at the moving fluid
(static pressure, ps) and one pointing at a position where the fluid is resting (stagnation pressure, pt).
Bernoulli’s equation states that stagnation pressure is equal to the sum of the static pressure and the
dynamic pressure (linked to the kinetic energy of the moving fluid) (3.1):
𝜌𝑉 2
𝑝𝑡 = 𝑝𝑠 + ( ) equation 3.1
2

Where V is the fluid velocity, pt is the stagnation pressure, ps is the static pressure ρ is the fluid density.
The fluid flow velocity can then be calculated with the following equation :

2∙(𝑝𝑡 −𝑝𝑠 )
𝑉=√ 𝜌
equation 3.2

Figure 3.10: Nozzle Pitot Flow Sensor

Concerning the heat exchanger, 3 plates were instrumented with temperature sensors at three specific
points: inlet, middle and outlet part of the plate. Two sensors were installed for each point: one inside
the PCM, allowing a direct measurement of the PCM temperature and one on the exterior plate surface.
Regarding the PCM inserted sensors, small stainless steel tubes were used to avoid damaging risks and
ensure proper and lasting positioning (figures 3.11). The three instrumented plates were placed on the
middle and the two extremities of the heat exchanger (figure 3.5). Finally, temperature sensors were
also installed on the exterior surface of each plate. These sensors were used to verify airflow uniformity
and as a charging discharging indicator. The above mentioned temperature sensors are identical to the
ones used for the air temperature measurement (PT100 class A).

86
 Figure 3.11: Insertion of temperature sensors inside the PCM

In conclusion, the instrumentation work was performed in order to provide all the necessary information
concerning the fusion and the solidification of the PCM, the charge / discharge phases of the heat
exchanger and the available stored heat.

Physical Number of
Position Accuracy
parameter sensors

Temperature
±0,15°C at 0°C
Heat exchanger PCM 9
±0,35 °C at 100°C
Plate surface 9+17

Air velocity ±0,5% reading

1
Outlet part accuracy

Rest of the
Relative humidity 2 ±2%.
experimental
platform Temperature
±0,15°C at 0°C
Inlet and outlet 2
±0,35 °C at 100°C
part

Table 3.3: Description of the instrumentation

87
[Link] Data acquisition and command interface
LabVIEW programming language and environment was used for data acquisition and control of the
components of the experimental platform. The developed interface allows direct monitoring of air and
PCM properties (figures 3.12 and 3.13), automation of characterization tests and elaboration of control
strategies (figure 3.14).

Figure 3.12: Air and PCM properties monitoring interface

The data logger is composed of 5 National Instruments Cards that are inserted in a Compact Daq
modular system and connected to a PC through a USB port. The developed Labview interface performs
data acquisition every 15 seconds. Several tabs were created, visualizing in real time the PCM properties
at different parts of the heat exchanger (figure 3.13), as well as air properties (inlet-outlet air
temperature, relative humidity and airflow rate) (figure 3.12). Furthermore, one tab presents the real
time energy consumption in France as obtained by the RTE site (RTE) and another tab illustrates the
characteristics of the PCM.
Power and energy consumption are also directly calculated from the acquired data and displayed on the
interface. More specifically, the enthalpy of the air is calculated at the inlet and outlet parts of the
exchanger with equation (3.3) (Albright, 1990):
𝐻 = 1,006 ∙ 𝑇 + 𝑤 ∙ (2501 + 1,805 ∙ 𝑇) equation 3.3
where H is enthalpy, T is temperature and w is the mixture ratio of water vapor mass and dry air mass,
calculated with equation (3.4):
𝜑∙𝑃𝑠𝑎𝑡
𝑤 = 0,6219 ∙ 𝑃 equation 3.4
𝑎𝑡𝑚 −(𝜑∙𝑃𝑠𝑎𝑡 )

where φ is relative humidity, Psat is saturation pressure and Patm is standard atmospheric pressure
(101,325 Pa).

88
The power of the heat exchanger is then calculated with equation 3.5:
𝑃 = 𝑚̇ ∙ 𝛥𝐻 equation 3.5
where 𝑚̇ is the air mass flow rate and ΔH is the enthalpy difference of the air between the inlet and
outlet part of the exchanger.
Data are saved in a txt file; an Excel worksheet template was prepared, in which the data were copied
and graphs were created concerning all the acquired information.

Figure 3.13: PCM and surface temperature monitoring interface

Figure 3.14: Data acquisition and command implementation interface

89
3.2.5 Operation principle and associated physical phenomena
The designed system can be used to enhance the heating performance of the mechanical ventilation
system and to perform electrical load shifting.
According to the orders given by the command interface the two fans create an airflow, ranging from 0
to 550 m3/h approximately. Hence, air constitutes the heat transferring medium of the unit. The air
diffuser ensures a uniform distribution between the plates and the air collector re-concentrates the
airflow to achieve the connection with a desired room.
When heat storage is required, the electrical duct heating is activated, hot air circulates between the
plates and heats their surface by convection. The encapsulated PCM absorbs heat by conduction, at first
in sensible form and when its melting temperature range is surpassed in latent form (charging process).
As the PCM melts progressively along the plate, convection phenomena start to appear inside the plate
as well. When all the PCM is melted, heat is absorbed once again in sensible form; PCM and outlet air
tend to acquire inlet air temperature value. At that point, the stocking process is completed and the heat
exchanger is considered to be fully charged. According to the desired control strategy the unit can be
preserved in that state for a certain period (i.e. stocking period).
The opposite procedure is followed during the discharging phase: cold or interior air is forced through
the plates, the PCM is cooled and starts to solidify, releasing the previously stored heat. The heated air is
then reintroduced into the building. When the solidification process is completed, the discharging of the
unit is achieved.
The recovered heat can be used for load shifting purposes. In order to
do that, heat is stored during an off-peak period, for example
throughout the night when consumption is minimal (figure 1.7) or
during the time when the COP of a heat pump is optimum (for example,
when indoor and outdoor temperature difference is minimal). The
stored heat is then used during the peak period in order to replace the
electrical heating and reduce the consumption. Predictive tools linked to
weather forecasting and the behavior of the occupants can be
developed and integrated in the control strategies elaboration process
for optimal exploitation of the system.
Limitations appeared concerning the inlet air temperature range values.
During the charging process the maximum acquired value was 50°C, as a
thermal protection of the duct heating is activated above this value.
Concerning the discharging procedure, inlet air temperature strongly
depends on the air source. If the platform is installed in an interior
space, inlet minimum air temperature value will not be inferior to 19°C.
The developed unit could be used for cooling purposes during the
summer period (night cooling). This would require the replacement of
the selected PCM (Mikrotek37 paraffin) with one that has different
Figure 3.15: PCM during properties (for example lower melting point temperature). In such a
liquefaction
90
case, during daytime, hot air flows through the heat exchanger at a higher temperature than the PCM’s
melting point. The PCM, initially in its solid form, absorbs heat from the air while obtaining its liquid
form. The cooled air is then injected into the building. In order to solidify the PCM, during the night
exterior air is circulated at a temperature lower than the PCM’s melting point.

3.2.6 Main components of the developed experimental system

The main components of the developed experimental system can be summarized as follows:
 The heat exchanger that contains the macroencapsulated paraffin and where the heat storage
occurs
 The circuit that allows the air flow (as heat transferring fluid), consisting of the air collector, air
diffuser and several ventilation ducts and fittings.
 The electrical heat source that is used to heat the air, consisting of two electrical resistances (total
power of 2KW)
 Two fans that create and regulate the flow of the air in the system
 Various sensors installed in different parts of the system in order to monitor the inlet/outlet air
temperature, the PCM temperature, the airflow etc.

3.2.7 Coupling with an experimental test cell

The experimental platform was coupled with an experimental test cell called Hybcell situated in the
ENTPE facilities in Lyon, France (El Mankibi, 2003). Hybcell is 5.1 m long, 3.5 m wide, 2.9 m high and
represents an office or small meeting room (figure 3.16). It is located within a large hall whose
temperature can be controlled in order to create artificial environment conditions. The north-east side of
the cell is in contact with the outdoor climate and possesses six automatized window openings for hybrid
ventilation studies.
The cell is equipped with two electric heaters, a mechanical ventilation system and CO2 generation and
sensible heat supply devices (for occupancy simulation). Various sensors (temperature, pressure, relative
humidity, CO2 concentration, COV) are installed in the middle of the test cell (at one meter from the
floor), as well as in the surrounding hall and outdoors. In addition, accurate measurement of the outdoor
climatic conditions can be performed using the IDMP meteorological station which is located at the
ENTPE facilities (Dumortier, 2003). The various components and sensors are linked to an existing
software that manages data acquisition and command/control orders.
The heat exchanger was placed outside the Hybcell and was coupled to its mechanical ventilation system
with the use of two fans and two electrical heating resistances. The developed software for the heat
exchanger management was merged with the existing Hybcell software for a global and simultaneous
operation of the system.
The coupling was performed in order to evaluate the performance of the heat exchanger in real life
conditions and to elaborate control strategies.

91
 Figure 3.16: The experimental test cell Hybcell

3.3 Characterization study

Several verification and characterization tests were realized for different airflow rates, aiming to:

 Confirm the correct operation of the various components of the experimental platform
 Confirm the correct operation of the data acquisition and control command interface
 Obtain an operation profile of the heat exchanger’s behavior for various airflow rates
 Verify the airflow uniformity between the plates
 Verify the repeatability of the charging/discharging phases
 Acquire reliable experimental data for the validation of the numerical model

3.3.1 PCM temperature evolution curves

The developed system allows the monitoring of the temperature evolution in various parts of the
exchanger. The following three figures illustrate the evolution of the PCM temperature (figure 3.17),
plate surface temperature (figure 3.18) and inlet/outlet air temperature (figure 3.19) during a full
thermal cycle for an airflow rate of 500 m3/h, at three parts of the middle plate of the heat exchanger:
inlet + 10cm, middle and outlet - 10cm part. During the charging phase, the inlet air temperature was set
to 44°C and during the discharging one at 24°C, as these two temperature values are sufficiently distant
from the melting/solidifying temperature range of the PCM.

92
 45
PCM temperature, inlet +10cm
PCM temperature, middle
40 PCM temperature, outlet -10cm
Temperature [°C]

35

30

25

20
13:00

14:00

15:00

16:00

17:00

18:00

19:00

20:00

21:00

22:00

23:00

0:00
Time [h]

Figure 3.17: Downstream PCM temperature evolution

45

Surface temperature, inlet +10cm

40 Surface temperature, middle
Surface temperature, outlet -10cm
Temperature [°C]

35

30

25

20
13:00

14:00

15:00

16:00

17:00

18:00

19:00

20:00

21:00

22:00

23:00

0:00

Time [h]

Figure 3.18: Downstream surface temperature evolution

93
 45

40 Air temperature, inlet

Air temperature, outlet
Temperature [°C]

35

30

25

20
13:00

14:00

15:00

16:00

17:00

18:00

19:00

20:00

21:00

22:00

23:00

0:00
Time [h]

Figure 3.19: Inlet and outlet air temperature evolution

Two expected behaviors were verified:

- The non-linearity evolution of the temperature, due to the latent heat absorbed/released during the
phase change process.
- The temperature evolution delay along the plate, as the inlet part is heated/cooled faster than the
outlet part.
These two observations indicate that the heat stored in the PCM is not immediately available totally and
depends on air flow rate as well as convective and conductive exchanges within the exchanger. Control
strategies and improvements on the exchanger should provide solutions to this phenomenon.

3.3.4 Characterization results

As discussed earlier, the characterization study was performed in order to obtain a full operational
profile of the heat exchanger and to obtain necessary experimental data that would validate the
numerical model.
A characterization protocol was established and can be described as follows:
 Testing of two different configurations: with and without fins
 Realization of a full thermal cycle (charge, discharge) with an inlet air temperature of 44°C during
stocking and 24°C during destocking phase
 Testing for five different airflow rates, from 100 to 500 m3/h, with an increase step of 100m3/h.

94
The obtained results are presented in figures 3.20 and 3.21, for the charging and discharging processes.
45

40

35
Temperature [°C]

30 Outlet air temperature 100 m3/h [°C]

Outlet air temperature 200 m3/h [°C]
Outlet air temperature 300 m3/h [°C]
25
Outlet air temperature 400 m3/h [°C]
Outlet air temperature 500 m3/h [°C]
20
0 1 2 3 4 5 6 7 8 9 10 11 12
Time [h]

Figure 3.20: Characterization results: outlet air temperature for varying airflow rates, charging phase

45

Outlet air temperature 100 m3/h [°C]

Outlet air temperature 200 m3/h [°C]
40
Outlet air temperature 300 m3/h [°C]
Outlet air temperature 400 m3/h [°C]
Outlet air temperature 500 m3/h [°C]
35
Temperature [°C]

30

25

20
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Time [h]

Figure 3.21: Characterization results: outlet air temperature for varying airflow rates, discharging
phase

95
3.3.2 PCM leakage
Data analysis rapidly demonstrated an atypical behavior, as the outlet air temperature decreased at a
faster rate for an airflow rate of 400 m3/h than the one of an airflow rate of 500m3/h. This means that an
error occurred either in the fabrication process, or the instrumentation of the unit or the data
acquisition. The experimental platform was dismantled in order to locate the source of the problem.
Despite the efforts for airtight containment of the PCM, some leakage was observed over the wooden
caps. The leakage was probably provoked by the pressure that was created during the phase change and
the dilatation of the PCM. Liquid paraffin had penetrated the wooden caps and the glue layer (figure
3.22). The amount of the leaked PCM was not important but the issue had to be resolved as it could lead
to further deterioration of the heat exchanger and lack of knowledge of the real PCM quantity in the
plates.

Figure 3.22: PCM leakage

3.3.3 Ameliorated plates

In view of these facts, the elaboration of new improved plates was decided. The new plates were
manufactured on demand by a metallurgical factory using the extrusion process. The material used was
once again aluminum. The plates had two openings for the pouring of the PCM and the evacuation of the
pressure caused by dilation between solid and liquid phase (figures 3.23 and 3.24). Their dimensions
were intended to be the same: (1x0,10x0,018)m. Nonetheless, minor dissimilarities occurred during the
fabrication process as some small protrusions were created in some parts of the plates (figure 3.25).

96
When proceeding with the assembly of the heat exchanger and the parallel emplacement of the plates,
the accumulated protrusions led to an increased thickness of each plate. This resulted in the removal of
one plate in order to reach the former heat exchanger thickness. However, the new design of the plates
permitted a higher filling and as a result the final PCM quantity was bigger than the initial one (table 3.4).
As before, three of the plates have additional pierced components that are used to introduce
temperature sensors at variable depths inside the PCM, in three different parts of the plates: inlet,
middle and outlet areas (figure 3.15).

Figure 3.23: The improved aluminum plates used as PCM containers

PCM
insertion

Temperature
sensors
insertion

Figure 3.24: The improved aluminum plates used as PCM containers, sketch

97
PCM was recovered from the old plates and inserted into the new ones. The plates were once again
weighed before and afteer the insertion of the PCM (table 3.4). The rest of the heat exchanger’s
structure and design was kept intact: polystyrene panels for thermal insulation and wooden boards for
the assembly. The heat exchanger was reconstructed and coupled to the experimetnal platform and
Hybcell.
Plate N° Tare weight (g) Laden weight (g) PCM weight (g)
1 1806 3781 1975
2 1815 3811 1996
3 1820 3835 2015
4 1843 3884 2041
5 1793 3808 2015
6 1813 3820 2007
7 1818 3787 1969
8 1962 3987 2025
9 1811 3807 1996
10 1818 3776 1958
11 1820 3816 1996
12 1823 3791 1968
13 1835 3788 1953
14 1844 3800 1956
15 1951 3910 1959
16 1955 3920 1965
Total PCM mass: 31794 g
Table 3.4: PCM mass calculation for the new heat exchanger

3.3.3 Verification tests

The verification tests were repeated and a full series of characterization tests was launched, providing
the operation profile of the heat exchanger for various airflow rates. The tests verified the correct
functioning of the heat exchanger, as well as the uniform distribution of the airflow between the plates..

Airflow uniformity
The outlet surface temperatures at the end of each plate were used at first to verify the uniformity of the
airflow distribution between the plates. Figure 3.26 illustrates the average, maximum and minimum
outlet surface temperature of 14 plates. The two plates at the extremity of the exchanger are not
considered here, as they neighbor with the insulation panels and are submitted to different conditions.
Results showed little variations between the different plates, which is regarded as acceptable,
considering the small but existing dissimilarities in the shape of the containers that occurred during
manufacturing (figure 3.25).

98
 Figure 3.25: Deformation on containers linked to the extrusion process

Furthermore, these tests were performed during a no fins configuration. The emplacement of the fins
would further ameliorate the air distribution, by adding to the existing pressure drop. In a second phase,
these measurements were used as a global indicator of the charging/discharging state of the exchanger.

44 Average outlet surface temperature [°C]

Maximum outlet surface temperature [°C]
Minimum outlet surface temperature [°C]

39
Temperature [°C]

34

29

24
0 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
Time [h]

Figure 3.26: Average, maximum and minimum outlet surface temperatures for 14 plates (airflow rate
of 300 m3/h)

99
3.3.4 Characterization results
As discussed earlier, the characterization study was performed in order to obtain a full operational
profile of the heat exchanger and to obtain the necessary experimental data that would validate the
numerical model.
A characterization protocol was established and can be described as follows:
 Testing of two different configurations: with and without fins
 Realization of a full thermal cycle with an inlet air temperature of 44°C during stocking and 26°C
during destocking phase
 Varying airflow rate values ranging from 100 to 500 m3/h
 Airflow rate increase step of 100 m3/h

[Link] Configuration without fins

The characterization results of the no fins configuration are presented in this section by illustrating the
evolution of the outlet air temperature of the heat exchanger for the five different airflow rates, during
the charging and discharging processes (figures 3.27 and 3.28).
45

40

35
Temperature [°C]

30 Outlet air temperature 100 m3/h [°C]

Outlet air temperature 200 m3/h [°C]
Outlet air temperature 300 m3/h [°C]
25
Outlet air temperature 400 m3/h [°C]
Outlet air temperature 500 m3/h [°C]

20
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Time [h]

Figure 3.27: Characterization results without fins: outlet air temperature for varying airflow rates,
charging phase

100
 45
Outlet air temperature 100 m3/h [°C]
Outlet air temperature 200 m3/h [°C]
40
Outlet air temperature 300 m3/h [°C]
Outlet air temperature 400 m3/h [°C]
35 Outlet air temperature 500 m3/h [°C]
Temperature [°C]

30

25

20
0 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
Time [h]

Figure 3.28: Characterization results without fins: outlet air temperature for varying airflow,
discharging phase

[Link] Configuration with fins

In the second phase of the characterization results, fins were placed between the plates for augmented
heat exchange performance. Results are presented in figures 3.29 and 3.30, for the five different airflow
rates, during the charging and discharging processes.

45

40
Temperature [°C]

35
T out 100 m3/h [°C]
30 T out 200 m3/h [°C]
T out 300 m3/h [°C]
25 T out 400 m3/h [°C]
T out 500 m3/h [°C]
20
0 1 2 3 4 5 6 7 8 9 10 11 12
Time [h]

Figure 3.29: Characterization results with fins: outlet air temperature for varying airflow, charging
phase

101
 45
T out 100 m3/h [°C]
T out 200 m3/h [°C]
40
T out 300 m3/h [°C]
T out 400 m3/h [°C]
35 T out 500 m3/h [°C]
Temperature [°C]

30

25

20
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Time [h]

Figure 3.30: Characterization results with fins: outlet air temperature for varying airflow, discharging
phase

[Link] Discussion of the characterization results

Figures 3.26 to 3.30 display the results of the characterization tests performed on the heat exchanger,
under two different configurations, with and without fins. More specifically the evolution of the outlet
air temperature is presented during the charging and the discharging processes for five different airflow
rates.
Outlet air temperature evolution is presented as it constitutes the connecting element between the heat
exchanger and the room in which the control strategies are applied. The two different configurations
(with and without fins) were investigated in order to experimentally demonstrate the impact of the fins
in the heat exchange process.
As expected, the configuration with fins presents a better performance in terms of needed time for
stocking and destocking. During the stocking procedure an important additional period is observed for
the full melting of the PCM, varying from approximately 1,5 hours with an airflow rate of 500 m3/h to
almost 5 hours for an airflow rate of 100 m3/h. During the destocking process the extra time required for
the full solidification of the PCM varied from 2,5 hours for an airflow rate of 500 m3/h to approximately 6
hours for an airflow rate of 100 m3/h.
As the configuration with fins presented a better performance, we will focus on it. Even though the heat
transfer is accelerated with the presence of fins, a significant amount of time is still needed for fully
charging or discharging the heat exchanger. For lower air flow rates the necessary interval was recorded
to be 11 hours for the stocking and 18 hours for the destocking process. This can be disturbing mostly
when destocking is needed, as the two hour barrier (during which the peak power demand occurs) is

102
largely exceeded. On the other hand a high air flow rate might deteriorate the occupants’ comfort as
non-negligible drafts may occur in the dwelling. High airflow rates can also provoke overheating of the
residence during the storage phase.
Small thermal losses are also observed, as the outlet air temperature never reaches the inlet 44°C
temperature. The difference between inlet and outlet temperature increases as the airflow rate
decreases.
Characterization results gave us an overview of the LHS unit behavior for two different configurations
and for varying airflow rates, focusing on the needed time for the charging and discharging phase.
Furthermore, reliable and adequate experimental data were obtained which were used for the validation
and calibration process of the numerical model.

3.4 Coupling with an experimental room and discussion

As presented earlier, the heat exchanger was coupled with the experimental test cell (Hybcell). It was
implemented during a consecutive four day period, with similar climatic winter conditions (outdoor
temperature is illustrated in figure 3.31). The aim of the developed strategy was to maintain a
comfortable thermal environment inside the test cell throughout the testing period, using the exchanger
experimental platform as a heat source. During the peak power hours (18:00-20:00) the electrical
resistances were turned off and the heat absorbed by the exchanger was released into the cell.
The evolution of Hybcell’s air temperature is illustrated in figure 3.31, along with the consumption of the
electrical resistances (100% refers to 2000W), the airflow rate and the outdoor temperature.

T Hybcell [°C] T outdoor [°C] Heating [%] Airflow rate [m3/h]

45 300
40
250
35
Airflow rate [m3/h], Heating
30 200
25
Temperature [°C]

150
20
15 100

10 50
5
0
0
-5 -50
12:00

18:00

12:00

18:00

12:00

18:00

12:00

18:00
0:00

6:00

0:00

6:00

0:00

6:00

0:00

Time [h]

Figure 3.31: Integration of AIR-PCM exchanger and HVAC system of Hybcell

103
During the off-peak hours, part of the heat is absorbed by the heat exchanger and the rest is directed
towards the test cell. A set point value of 21°C was established for Hybcell’s temperature and the
controller responded towards that goal by altering the airflow rate and the percentage of electrical
heating used. The highlighted gray areas correspond to the 18:00-20:00 period (peak power period)
when the electrical resistances were off. In this case, the control was operated only on the airflow rate.
During this period, the indoor thermal comfort was maintained, using only the heat stored in the PCM;
the indoor air temperature was kept constant at 21°C, with small discrepancies being observed. This
means that the stored energy in the AIR-PCM exchanger is enough to provide the required thermal
comfort while being able to reduce peak demand.
During the fourth day, starting from 18:00, the heating system and the ventilation system were turned
off (i.e. no heating and no heat exchanger), resulting in a significant temperature drop down to 14°C.
This day was taken as a reference day in order to assess the potential of the heat exchanger.
This example of preliminary control strategy demonstrates the applicability of the developed system in a
realistic scenario during a consecutive 4 day winter period. Electricity heating usage is shifted to an off-
peak power period while the thermal comfort inside the text cell is not degraded, during the charging or
the discharging phase.

3.5 Conclusion
A PCM-Air heat exchanger was designed and fabricated, using paraffin as the heat storage medium. The
unit was coupled to an experimental platform and then characterized for various airflow rates. A
drawback was observed as PCM leakage occurred at the sealing points of the encapsulating plates. As a
consequence, a second exchanger was fabricated, ensuring proper containment and no leakage.
The developed system was coupled to an experimental room and a preliminary control strategy was
implemented. Results showed that the proposed system managed to perform peak load shifting while
maintaining comfortable indoor conditions.
Nevertheless, the needed charging/discharging time was superior to the peak period, this observation
leading to the need for an optimization study for the heat exchanger. For financial and time constraints,
this study was performed using a numerical model.
The experimental data obtained through the characterization tests were used to validate the model. The
development and the validation of the model are presented in the following chapter.

104
Chapter 4
Numerical model development

105
Chapter 4 : Numerical model development

4.1 Introduction
Even though an experimental unit was assembled and characterized, the development of a numerical
model was found necessary. The advantages of the numerical model lie in its ability to reproduce the
thermal performance of various configurations of the heat exchanger in a very short time period and
with little financial investment.

This chapter presents the procedure followed for the model development step by step: the heat
transferring phenomena considered, layer selection, nodal discretization, heat balance equation
formulation, matrix writing and solution. The calibration study is also detailed, focusing on the creation
of fictive heat capacity curves.

The model is validated using experimental data and the four day testing presented in the previous
chapter is numerically recreated and compared to experimental results.

4.2 State of the art of existing models

Modelling and simulations have been used in several studies with multiple objectives, including analysis
and understanding of PCM behavior and its applications, as well as optimization and design aspects. PCM
studies comprise the difficulty of nonlinear behavior and the variation of thermo-physical properties of
the studied medium during the phase change process. In order to overcome these obstacles, various
methods and techniques have been developed. Fundamentally, the developed models are based on the
formulation of a series of energy related equations and their solution using analytical and numerical
methods.
Analytical methods are utilized in one-dimensional problems with mostly simple initial and boundary
conditions and constant thermal properties. Examples of such techniques are the Neumann method
(Lamberg et al., 2003) and the Heat balance integral method (Mitchell et al., 2012).
Numerical methods are more frequently used for phase change problems, as they offer the advantage of
being able to change thermal properties values and to calculate the moving boundary of phases. Their
main disadvantages are less (but sufficient) precision and possible discontinuities linked to the phase
transition. Several reviews exist, describing the existing numerical methods for phase change modeling
(Hu et al., 1996, Voller et al., 2006, Verma et al., 2008, Dutil et al., 2011, Idelsohn et all., 1994).
This category can be generally subdivided into three main methods:
 The fixed grid method

In this method, apart from the use of temperature as a dependent variable, enthalpy is introduced as
well. As enthalpy is continuous only in time (and temperature only in space), the governing equation is
the same for the two phases and can be solved over the whole space as a fixed domain. An example of
this method can be found in Nedjar, 2002.

106
  The moving grid method
In this method, the exact location of the moving boundary is evaluated on a grid at each step, allowing to
avoid the constraints of the fixed grid formulation. Two possibilities are offered for the grid definition:
the interface-fitting grids (or variable time step method) that uses a uniform spatial grid but a non-
uniform time step and the dynamic methods (or variable space grids) where the moving boundary lies on
a particular grid point while the number of spatial intervals remains constant. An example of this method
can be found in Mackenzie et al., 2002.
 Hybrid method
Finally, hybrid methods exist, using elements from the two previous methods. Example: Voller et al.,
2006.
Some authors (Hu et al., 1996) distinguish another category, the latent heat evolution method. They
subdivide it into the apparent heat capacity method (Arkar et al., 2005), the effective capacity method
(Heim et al., 2004), the heat integration method (Rolph et al., 1982) and the source based method
(Voller et al., 1991).

4.3 Numerical model development: Heat balance approach and apparent heat capacity
method
Numerical models are powerful tools that can help overcome the limitations of experimental studies
regarding time and financial issues. In this aspect, the development of a model that would simulate the
behavior of the LHS unit was necessary. The model needs to be both fast and accurate to reproduce the
use of the heat exchanger and be coupled with an existing building model for optimization and control
strategies development.
Thus, a two dimensional numerical model was created in a Matlab/Simulink environment, using the heat
balance approach and the apparent heat capacity method.

4.3.1 Matlab/Simulink and S-function choice

The Matlab/Simulink environment was chosen for a number of reasons:
- The existing building model (Hybcell 1.2) that would later be used for the control strategies
development was also developed under the same environment; thus, the coupling between the two
models was facilitated.
- Simulink offers a variety of controller toolboxes, such as On-Off, PID and Fuzzy.
- Simulink’s graphical language is similar to LabVIEW’s, which is the interface used for controlling
the HYBCELL cell and other experimental devices. This similarity allows the transition from the numerical
to the experimental in an easier way.
- Simulink also provides a flexible and user friendly environment, compared to a procedural
programming language.

107
 - It also offers a powerful optimization toolbox.
- Nevertheless, Matlab and Simulink are compatible with other programming languages.
In fact, the developed model was programmed in MATLAB® language and was integrated into Simulink
using an S-function block (figure 4.1). S-functions (system functions) can extend the capabilities of the
Simulink environment by automatically loading and computing a subroutine written in various
programming languages (MATLAB®, C, C++, or Fortran). The S-function block interacts with the Simulink
environment by requesting input variables and supplying output results.

Figure 4.1: S-function block under Simulink environment (left) and parameters definition block (right)

4.3.2 Heat balance approach

The energy balance equation states that at any given location (or control volume or node) in a system,
the heat entering into that location is equal to the heat exiting the location, plus any heat that is stored
(heat storage is reflected in the increase of heat capacity values).

The heat balance approach is based on the division of the considered medium into a sufficient number of
volume elements and the application of the energy balance equation on each element. This is performed
by achieved through the formulation of a nodal discretization and the assumption that the temperature
value is the same for the volume that corresponds at each node. If several components or mediums are
investigated (solid, liquid, PCM), different layers can be considered in order to represent the varying
properties of each layer. All relevant energy flows can be taken into account, including conduction,
convection, advection and radiation. The heat balance equation is formed for every node and solved at
each time step to calculate the temperature evolution at the considered nodes. The finite difference
method is employed where needed to approximate the partial differential equations.
This approach offers the possibility to vary the materials’ properties with temperature and time and
provides flexibility in the choice of simulation time step.

108
Summarizing the heat balance approach, three main steps can be identified:

 Nodal discretization of the considered layer(s).

 Heat balance formulation for each node considering all the relevant energy flows.
 Transformation of the obtained equations to matrix equations and solution for each time step.
A detailed description of the followed procedure of this study is presented in subsection 4.3.4.

4.3.3 Apparent heat capacity method

This method consists in representing the phase change of the heat storage medium (paraffin here)
through an apparent increase of the storage medium’s heat capacity values for a certain temperature
range, corresponding to the latent heat absorption/release. If the process is performed uniformly, the
apparent heat capacity can be defined as:
𝐶𝑝,𝑠 , 𝑇 < 𝑇𝑠 (𝑠𝑜𝑙𝑖𝑑 𝑝ℎ𝑎𝑠𝑒)
𝐶𝑝,𝑎𝑝𝑝 = {𝐶𝑝 (𝑇), 𝑇𝑠 < 𝑇 < 𝑇𝑙 (𝑝ℎ𝑎𝑠𝑒 𝑐ℎ𝑎𝑛𝑔𝑒), [J/(kg·K)] equation 4.1
𝐶𝑝,𝑙 , 𝑇 > 𝑇𝑙 (𝑙𝑖𝑞𝑢𝑖𝑑 𝑝ℎ𝑎𝑠𝑒)

Where 𝐶𝑝 (𝑇) represents the specific heat capacity between phase changes; it is dependent on the
temperature value and is represented by a curve.

4.3.4 Nodal discretization and application of the heat balance approach

The main component of this study is the PCM-Air heat exchanger. For the numerical representation of
this unit, there are three principle mediums were heat transfer occurs: air, container (aluminum) and
PCM. Consequently, three layers were considered for nodal discretization: air, aluminum and PCM.
Half the thickness of the experimental air layer and the PCM layer thickness is considered in the
numerical study, considering that by symmetry the results can be extended to the other half (where
similar conditions apply).
Each layer was then discretized into n equal regions (nodes) of length dx=L/n, where L is the total length
of the heat exchanger (figure 4.2).

Figure 4.2: Schematic representation of the considered layers for temperature evolution calculation

109
The PCM layer was further discretized crosswise in m nodes. The discretization into multiple nodes
offers the possibility of a more thorough treatment of the medium where the phase change occurs.
Figure 4.4 illustrates in a more detailed form the nodal discretization of the three layers: one crosswise
node for the air and aluminum layer (denoted as Ta and Tp respectively) and m nodes for the PCM layer
(denoted as Tm). Lengthwise n nodes are considered for each layer. Subscript letter i signifies the nodal
lengthwise position for each layer (i=1 to n) and the subscript j signifies the nodal crosswise position in
the PCM layer (j=1 to m).
The thermal resistances are formulated between nodes (figure 4.4), with different coefficients being
used for inlet and outlet nodes, where necessary (see next section). Inlet air temperature and airflow
rate are introduced for every time step (denoted as Tin and ṁ respectively).

4.3.5 Energy balance equation leading to matrix formulation

At a random node i and depending on the considered layer, several heat transfers occur, as described in
subsection ‘3.2.5 Operation principle and associated physical phenomena’. Considering that all possible
transfer phenomena take place at that node (apart from those stated in later subsection 4.3.6 ‘Model
assumptions’) the energy balance equation can be written as follows:

𝑇𝑒𝑚𝑝𝑒𝑟𝑎𝑡𝑢𝑟𝑒 𝐶𝑜𝑛𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝐶𝑜𝑛𝑣𝑒𝑐𝑡𝑖𝑣𝑒 𝐴𝑑𝑣𝑒𝑐𝑡𝑖𝑜𝑛

= + +
{ 𝑐ℎ𝑎𝑛𝑔𝑒 𝑟𝑎𝑡𝑒 } { 𝑡𝑟𝑎𝑛𝑠𝑓𝑒𝑟𝑠 } { 𝑡𝑟𝑎𝑛𝑠𝑓𝑒𝑟𝑠 } { 𝑡𝑟𝑎𝑛𝑠𝑓𝑒𝑟𝑠 } equation 4.2
𝑖𝑛 𝑛𝑜𝑑𝑒 𝑖𝑛 𝑛𝑜𝑑𝑒 𝑖𝑛 𝑛𝑜𝑑𝑒 𝑖𝑛 𝑛𝑜𝑑𝑒

The temperature change rate (or storage of heat in the node) can be expressed as follows:

𝑇𝑒𝑚𝑝𝑒𝑟𝑎𝑡𝑢𝑟𝑒
𝑑𝑇
{ 𝑐ℎ𝑎𝑛𝑔𝑒 𝑟𝑎𝑡𝑒 } = 𝜌𝐶𝑝 𝑉 𝑑𝑡 equation 4.3
𝑖𝑛 𝑛𝑜𝑑𝑒

Where

ρ ∶ Density of the node’s medium [kg/m3]

Cp ∶ Specific heat capacity of the node’s medium [J/(kg·K)]
For the PCM layer Cp = Capp as described in equation 4.1
Va = dy ∙ dx ∙ dz ∶ Volume of the node [m3]
dy Width of air node [m]
dx Length of node [m]
dz Height of node [m]
T∶ Temperature of the node [°C]
t ∶ Time [s]

110
 Air layer

= ̇ ∙ , , ,

1
1 = + 1 1
∙ℎ 2 ∙
1 Aluminum layer
= + 2 =
∙ℎ 2 ∙ , ∙ , ,

2
= +
2 ∙ ∙

1 PCM layer
1 = + =
1∙ℎ 2 1∙ , 1∙
, ,

1 =
= + ∙
2
∙ℎ 2 ∙ =
2 2
, 2 ∙ , ,

1
= + =
∙ℎ 2 ∙ ∙ ,
, ,

Figure 4.3: Upper view of the nodal representation of the layers: air (a), aluminum (p) and PCM (m)

111
The conduction term concerns the aluminum and PCM. It is given by the following equation:

𝐶𝑜𝑛𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑇 −2 𝑇𝑖 +𝑇𝑖+1
{ 𝑡𝑟𝑎𝑛𝑠𝑓𝑒𝑟𝑠 } = 𝑖−1 equation 4.4
𝑅
𝑖𝑛 𝑛𝑜𝑑𝑒

Where

Ti ∶ Node temperature at node i [°C]

Ti−1 ∶ Node temperature at node i-1 [°C]
Ti+1 ∶ Node temperature at node i+1 [°C]
dx
R= ∶ Thermal resistance between neighboring nodes [W/K]
S∙λ
dx: Length of node [m]
S: Contact surface between neighboring aluminum nodes [m2]
λ Conductivity of the node’s material [W/(m·K)]

The advection term concerns the air layer and represents the heat transported into an air node i from a
neighboring air node i-1. In its explicit form, the advection term is given by the following equation:

𝐴𝑑𝑣𝑒𝑐𝑡𝑖𝑜𝑛
{ 𝑡𝑟𝑎𝑛𝑠𝑓𝑒𝑟𝑠 } = 𝑚̇ 𝐶𝑝𝑎 (𝑇𝑎,𝑖−1 − 𝑇𝑎,𝑖 ) equation 4.5
𝑖𝑛 𝑛𝑜𝑑𝑒

Where

Cpa ∶ Specific heat capacity of the air [J/(kg·K)]

ṁ ∶ Airflow rate [kg/s]
Ta,i ∶ Air temperature at node i [°C]
Ta,i−1 ∶ Air temperature at node i-1 [°C]

The energy balance equation is written for each node, taking into account that the heat entering into
that node is equal to the heat exiting that position.

𝑇 𝑡 +𝑇 𝑡−𝑑𝑡
The average temperature 𝑇̅ = is introduced as the temperature variation is considered to be
2
linear and the heat balance equation is formulated for each layer. The finite difference method is used to
approximate the temperature derivative terms and the final form of the equations leads to a matrix
formulation for each layer. The following section presents this analysis for the air, aluminum and PCM
layers. Regarding the PCM layer, m was considered equal to three, meaning that a three node
discretization was performed for this layer crosswise.

112
[Link] Air layer
The application of the heat balance equation for the ith air node results in the following equation:

𝑑𝑇𝑎,𝑖
𝜌𝑎 𝐶𝑝𝑎 𝑉𝑎 = 𝑚̇𝐶𝑝𝑎 (𝑇𝑎,𝑖−1 − 𝑇𝑎,𝑖 ) + ℎ𝑆(𝑇𝑝,𝑖 − 𝑇𝑎,𝑖 ) equation 4.6
𝑑𝑡

Where

ρa ∶ Air density [kg/m3]

Cpa ∶ Specific heat capacity of the air [J/(kg·K)]
Va = dya ∙ dx ∙ dz ∶ Volume of the air node [m3]
dya Width of air node [m]
dx Length of node [m]
dz Height of node [m]
Ta,i ∶ Air temperature at node i [°C]
ṁ ∶ Airflow rate [kg/s]
Ta,i−1 ∶ Air temperature at node i-1 [°C]
h∶ Convection coefficient [W/m2·K]
S = dx dz ∶ Contact surface between air and plate surface [m2]
Tp,i Surface temperature at node i [°C]

By introducing the average temperature term, equation 4.6 becomes:

2 𝜌𝑎 𝐶𝑝𝑎 𝑉𝑎
̅̅̅̅
(𝑇 𝑡−𝑑𝑡 ̅̅̅̅̅̅̅ ̅̅̅̅ ̅̅̅̅ ̅̅̅̅
𝑎,𝑖 − 𝑇𝑎,𝑖 ) = 𝑚̇ 𝐶𝑝𝑎 (𝑇𝑎,𝑖−1 − 𝑇𝑎,𝑖 ) + ℎ 𝑆 (𝑇𝑝,𝑖 − 𝑇𝑎,𝑖 )
𝑑𝑡

𝑡−𝑑𝑡
Out of which 𝑇𝑎,𝑖 can be expressed as a function of ̅̅̅̅
𝑇𝑎,𝑖 , ̅̅̅̅̅̅̅
𝑇𝑎,𝑖−1 and ̅̅̅̅
𝑇𝑝,𝑖 :

𝑚̇ 𝑑𝑡 ℎ 𝑆 𝑑𝑡
̅̅̅̅ 𝑡−𝑑𝑡
𝑇𝑎,𝑖 − 𝑇𝑎,𝑖 = ̅̅̅̅̅̅̅
(𝑇 ̅̅̅̅
𝑎,𝑖−1 − 𝑇𝑎,𝑖 ) +
̅̅̅̅
(𝑇 ̅̅̅̅
𝑝,𝑖 − 𝑇𝑎,𝑖 )
2 𝜌𝑎 𝑉𝑎 2 𝜌𝑎 𝐶𝑝𝑎 𝑉𝑎
𝑚̇ 𝑑𝑡 ℎ 𝑆 𝑑𝑡 𝑚̇ 𝑑𝑡 ℎ 𝑆 𝑑𝑡
𝑡−𝑑𝑡
𝑇𝑎,𝑖 = (1 + + ) ̅̅̅̅
𝑇𝑎,𝑖 − ̅̅̅̅̅̅̅
𝑇𝑎,𝑖−1 − ̅̅̅̅
𝑇𝑝,𝑖 ,
2 𝜌𝑎 𝑉𝑎 2 𝜌𝑎 𝐶𝑝𝑎 𝑉𝑎 2 𝜌𝑎 𝑉𝑎 2 𝜌𝑎 𝐶𝑝𝑎 𝑉𝑎

Setting

𝑎𝑎 = 1 + 𝑏𝑎 + 𝑐𝑎 ,
𝑚̇ 𝑑𝑡
𝑏𝑎 = 2 𝜌 and
𝑎 𝑉𝑎

ℎ 𝑆 𝑑𝑡
𝑐𝑎 = 2 𝜌
𝑎 𝐶𝑝𝑎 𝑉𝑎

The final equation 4.7 is obtained

113
 𝑡−𝑑𝑡
𝑇𝑎,𝑖 = 𝑎𝑎 ̅̅̅̅
𝑇𝑎,𝑖 − 𝑏𝑎 ̅̅̅̅̅̅̅
𝑇𝑎,𝑖−1 − 𝑐𝑎 ̅̅̅̅
𝑇𝑝,𝑖 equation 4.7

For node i=1:

𝑡−𝑑𝑡
𝑇𝑎,1 = 𝑎𝑎 ̅̅̅̅̅
𝑇𝑎,1 − 𝑏𝑎 ̅̅̅̅
𝑇𝑖𝑛 − 𝑐𝑎 ̅̅̅̅̅
𝑇𝑝,1 equation 4.8

Equation 4.7 and 4.8 can then be written in a matrix equation, representing nodes 1 to n:
𝑡−𝑑𝑡
𝑇𝑎,1 𝑎𝑎 0 ⋯ ⋯ 0 ̅̅̅̅̅
𝑇𝑎,1 −𝑐𝑎 0 ⋯ ⋯ 0 ̅̅̅̅̅
𝑇𝑝,1 −𝑏𝑎 𝑇𝑖𝑛
⋮ −𝑏𝑎 ⋱ ⋱ ⋮ ⋮ 0 ⋱ ⋱ ⋮ ⋮ 0
𝑡−𝑑𝑡
𝑇𝑎,𝑖 = 0 ⋱ 𝑎𝑎 ⋱ ⋮ × ̅̅̅̅
𝑇𝑎,𝑖 + ⋮ ⋱ −𝑐𝑎 ⋱ ⋮ × ̅̅̅̅
𝑇𝑝,𝑖 + ⋮
⋮ ⋮ ⋱ ⋱ 0 ⋮ ⋮ ⋱ ⋱ 0 ⋮ ⋮
𝑇 𝑡−𝑑𝑡
[ 𝑎,𝑛 ] [ 0 ⋯ 0 −𝑏𝑎 𝑎𝑎 ] [̅̅̅̅̅
𝑇𝑎,𝑛 ] [ 0 ⋯ ⋯ 0 −𝑐𝑎 ] [̅̅̅̅̅
𝑇𝑝,𝑛 ] [ 0 ]

[Link] Aluminum layer

The application of the heat balance equation for the ith aluminum node results to the following equation:

𝑑𝑇𝑝,𝑖 𝑇𝑎,𝑖 −𝑇𝑝,𝑖 𝑇𝑝,𝑖−1 −2 𝑇𝑝,𝑖 +𝑇𝑝,𝑖+1 𝑇𝑚1,𝑖 −𝑇𝑝,𝑖

𝜌𝑝 𝐶𝑝𝑝 𝑉𝑝 = + + equation 4.9
𝑑𝑡 𝑅𝑝 1 𝑅𝑝 2 𝑅𝑝 3

Where

ρp ∶ Aluminum density [kg/m3]

Cpp ∶ Specific heat capacity of aluminum [J/(kg·K)]
Vp = dyp ∙ dx ∙ dz : Volume of the aluminum node [m3]
dyp Width of aluminum node [m]
Ta,i ∶ Air temperature at node i [°C]
Tp,i ∶ Aluminum temperature at node i [°C]
Tp,i−1 ∶ Aluminum temperature at node i-1 [°C]
Tp,i+1 ∶ Aluminum temperature at node i+1 [°C]
Tm1,i ∶ First PCM temperature at node i [°C]
1 dyp
Rp 1 = + : Thermal resistance between air and aluminum nodes [W/K]
S ∙ h 2 S ∙ λp
dx
Rp 2 = ∶ Thermal resistance between neighboring aluminum nodes [W/K]
S p ∙ λp
Sp = dyp ∙ dz : Contact surface between neighboring aluminum nodes [m2]
dyp dym
Rp = + ∶ Thermal resistance between aluminum and PCM nodes [W/K]
2 S ∙ λp S ∙ λm

114
λp : Aluminum conductivity [W/(m·K)]
λm : PCM conductivity [W/(m·K)]
dym : Width of PCM node [m]

By introducing the average temperature term, equation 4.9 becomes:

2 𝜌𝑝 𝐶𝑝𝑝 𝑉𝑝 ̅̅̅̅
𝑇𝑎,𝑖 − ̅̅̅̅
𝑇𝑝,𝑖 ̅̅̅̅̅̅̅
𝑇𝑝,𝑖−1 − 2 ̅̅̅̅
𝑇𝑝,𝑖 + ̅̅̅̅̅̅̅
𝑇𝑝,𝑖+1 ̅̅̅̅̅̅
𝑇𝑚1,𝑖 − ̅̅̅̅
𝑇𝑝,𝑖
̅̅̅̅
(𝑇𝑝,𝑖 − 𝑇 𝑡−𝑑𝑡
𝑝,𝑖 ) = + +
𝑑𝑡 𝑅𝑝 1 𝑅𝑝 2 𝑅𝑝

𝑡−𝑑𝑡
Out of which, 𝑇𝑝,𝑖 can be expressed as a function of ̅̅̅̅
𝑇𝑎,𝑖 , ̅̅̅̅
𝑇𝑝,𝑖 , ̅̅̅̅̅̅̅
𝑇𝑝,𝑖−1 , ̅̅̅̅̅̅̅
𝑇𝑝,𝑖+1 and ̅̅̅̅̅
𝑇𝑚,𝑖 :

𝑑𝑡 1 2 1 𝑑𝑡
𝑡−𝑑𝑡
𝑇𝑝,𝑖 = [1 + (2 𝜌 ) (𝑅 +𝑅 + 𝑅 )] ̅̅̅̅
𝑇𝑝,𝑖 − 2 𝜌 ̅̅̅̅̅̅̅
(𝑇 ̅̅̅̅̅̅̅
𝑝,𝑖−1 + 𝑇𝑝,𝑖+1 ) −
𝑝 𝐶𝑝𝑝 𝑉𝑝 𝑝1 𝑝2 𝑝3 𝑝 𝐶𝑝𝑝 𝑉𝑝 𝑅𝑝 2
𝑑𝑡 𝑑𝑡
̅̅̅̅
𝑇𝑎,𝑖 − ̅̅̅̅̅̅
𝑇𝑚1,𝑖 ,
2 𝜌𝑝 𝐶𝑝𝑝 𝑉𝑝 𝑅𝑝 1 2 𝜌𝑝 𝐶𝑝𝑝 𝑉𝑝 𝑅𝑝 3

Setting
𝑑𝑡 𝑑𝑡 𝑑𝑡
𝑎𝑝 = 1 + 2 𝑏𝑝 + 𝑐𝑝 + 𝑑𝑝 , 𝑏𝑝 = , 𝑐𝑝 = 𝑎𝑛𝑑 𝑑𝑝 =
2 𝜌𝑝 𝐶𝑝𝑝 𝑉𝑝 𝑅𝑝 2 2 𝜌𝑝 𝐶𝑝𝑝 𝑉𝑝 𝑅𝑝 1 2 𝜌𝑝 𝐶𝑝𝑝 𝑉𝑝 𝑅𝑝 3

The final equation 4.10 is obtained

𝑡−𝑑𝑡
𝑇𝑝,𝑖 = 𝑎𝑝 ̅̅̅̅ ̅̅̅̅̅̅̅
𝑇𝑝,𝑖 − 𝑏𝑝 (𝑇 ̅̅̅̅̅̅̅ ̅̅̅̅ ̅̅̅̅̅̅
𝑝,𝑖−1 + 𝑇𝑝,𝑖+1 ) − 𝑐𝑝 𝑇𝑎,𝑖 − 𝑑𝑝 𝑇𝑚1,𝑖 equation 4.10

For node i=1:

𝑑𝑡
𝑎𝑝 1 = 1 + 𝑏𝑝 + 𝑐𝑝 + 𝑑𝑝 + 𝑒𝑝 𝑎𝑛𝑑 𝑒𝑝 = 2 𝜌 𝐶𝑝
𝑝 𝑝 𝑉𝑝 𝑅𝑝 4

Obtaining:
𝑡−𝑑𝑡
𝑇𝑝,1 = 𝑎𝑝 1 ̅̅̅̅̅
𝑇𝑝,1 − 𝑏𝑝 ̅̅̅̅̅
𝑇𝑝,2 − 𝑐𝑝 ̅̅̅̅̅
𝑇𝑎,1 − 𝑑𝑝 ̅̅̅̅̅̅̅
𝑇𝑚1,1 − 𝑒𝑝 ̅̅̅̅
𝑇𝑖𝑛 equation 4.11

For node i=n

𝑎𝑝 𝑛 = 1 + 𝑏𝑝 + 𝑐𝑝 + 𝑑𝑝

Obtaining:
𝑡−𝑑𝑡
𝑇𝑝,𝑛 ̅̅̅̅̅
= 𝑎𝑝 𝑛 𝑇 ̅̅̅̅̅̅̅̅ ̅̅̅̅̅ ̅̅̅̅̅̅̅
𝑝,𝑛 − 𝑏𝑝 𝑇𝑝,𝑛−1 − 𝑐𝑝 𝑇𝑎,𝑛 − 𝑑𝑝 𝑇𝑚1,𝑛 equation 4.12

Equation 4.10, 4.11 and 4.12 can then be written in a matrix equation, representing nodes 1 to n:

115
 𝑡−𝑑𝑡
𝑇𝑝,1 𝑎𝑝 1 −𝑏𝑝 ⋯ ⋯ 0 ̅̅̅̅̅
𝑇𝑝,1 −𝑐𝑝 0 ⋯ ⋯ 0 ̅̅̅̅̅
𝑇𝑎,1
⋮ −𝑏𝑝 𝑎𝑝 ⋱ ⋮ ⋮ 0 ⋱ ⋱ ⋮ ⋮
𝑡−𝑑𝑡
𝑇𝑝,𝑖 = 0 ⋱ 𝑎𝑝 ⋱ ⋮ × ̅̅̅̅ 𝑇𝑝,𝑖 + ⋮ ⋱ −𝑐𝑝 ⋱ ⋮ × ̅̅̅̅
𝑇𝑎,𝑖
⋮ ⋮ ⋱ 𝑎𝑝 −𝑏𝑝 ⋮ ⋮ ⋱ ⋱ 0 ⋮
𝑇 𝑡−𝑑𝑡
[ 𝑝,𝑛 ] [ 0 ⋯ 0 −𝑏𝑝 𝑎𝑝 𝑛 ] [𝑇̅̅̅̅̅
𝑝,𝑛 ] [ 0 ⋯ ⋯ 0 −𝑐𝑝 ] [̅̅̅̅̅
𝑇𝑎,𝑛 ]
−𝑑𝑝 0 ⋯ ⋯ 0 ̅̅̅̅̅̅̅
𝑇𝑚1,1 −𝑒𝑝 𝑇𝑖𝑛
0 ⋱ ⋱ ⋮ ⋮ 0
+ ⋮ ⋱ −𝑑𝑝 ⋱ ⋮ × ̅̅̅̅̅̅ 𝑇𝑚1,𝑖 + ⋮
⋮ ⋱ ⋱ 0 ⋮ ⋮
[ 0 ⋯ ⋯ 0 −𝑑𝑝 ] [̅̅̅̅̅̅̅ 𝑇𝑚1,𝑛 ] [ 0 ]

[Link] PCM layers

As discussed previously, the PCM layer was further discretized in m nodes crosswise. In this section (for
illustration purposes), m was considered equal to 3, meaning that a three node discretization was
performed for this layer.

[Link].1 First PCM node (PCM volume in contact with aluminum)

The application of the heat balance equation for the ith and first crosswise (j=1) PCM node results in the
following equation:

𝑑𝑇𝑚1,𝑖 𝑇𝑝,𝑖 −𝑇𝑚1,𝑖 𝑇𝑚1,𝑖−1 −2 𝑇𝑚1,𝑖 +𝑇𝑚1,𝑖+1 𝑇𝑚2,𝑖 −𝑇𝑚1,𝑖

𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 = + + equation 4.13
𝑑𝑡 𝑅𝑝 3 𝑅𝑚 4 𝑅𝑚 7

Where

ρm ∶ PCM density [kg/m3]

Cpm ∶ Specific heat capacity of the PCM [J/(kg·K)]
Vm1 = Sm1 ∙ dx ∶ Volume of the first PCM node [m3]
dym dym
Sm1 = ( + ) ∙ dz ∶ Contact surface between neighboring PCM nodes [m2]

Tm2,i ∶ Second PCM temperature at node i [°C]

Tm1,i−1 ∶ First PCM temperature at node i-1 [°C]
Tm1,i+1 ∶ First PCM temperature at node i+1 [°C]
dx
Rm = ∶ Thermal resistance between neighboring PCM nodes [W/K]
Sm1 ∙ λm
dym
Rm = ∶ Thermal resistance between first and second PCM nodes [W/K]
S ∙ λm

116
By introducing the average temperature term, equation 4.13 becomes:

2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 ̅̅̅̅

𝑇𝑝,𝑖 − 𝑇̅̅̅̅̅̅
𝑚1,𝑖
̅̅̅̅̅̅̅̅̅
𝑇𝑚1,𝑖−1 − 2 ̅̅̅̅̅̅
𝑇𝑚1,𝑖 + ̅̅̅̅̅̅̅̅̅
𝑇𝑚1,𝑖+1 ̅̅̅̅̅̅
𝑇𝑚2,𝑖 − ̅̅̅̅̅̅
𝑇𝑚1,𝑖
̅̅̅̅̅̅
(𝑇 𝑡−𝑑𝑡
𝑚1,𝑖 − 𝑇𝑚1,𝑖 ) = + +
𝑑𝑡 𝑅𝑝 𝑅𝑚 𝑅𝑚

𝑡−𝑑𝑡
Out of which, 𝑇𝑚1,𝑖 can be expressed as a function of ̅̅̅̅
𝑇𝑝,𝑖 , ̅̅̅̅̅̅
𝑇𝑚1,𝑖 , ̅̅̅̅̅̅̅̅̅
𝑇𝑚1,𝑖−1 , ̅̅̅̅̅̅̅̅̅
𝑇𝑚1,𝑖+1 and ̅̅̅̅̅̅
𝑇𝑚2,𝑖 :

𝑑𝑡 1 2 1 𝑑𝑡
𝑡−𝑑𝑡
𝑇𝑚1,𝑖 = [1 + ( ) ( + + )] ̅̅̅̅̅̅
𝑇𝑚1,𝑖 − ̅̅̅̅̅̅̅̅̅
(𝑇 ̅̅̅̅̅̅̅̅̅
𝑚1,𝑖−1 + 𝑇𝑚1,𝑖+1 ) −
2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑝 3 𝑅𝑚 4 𝑅𝑚 7 2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑚 4
𝑑𝑡 𝑑𝑡
̅̅̅̅
𝑇𝑝,𝑖 − ̅̅̅̅̅̅
𝑇 ,
2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑝 3 2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑚 7 𝑚2,𝑖

Setting:
𝑑𝑡 𝑑𝑡
𝑎𝑚1 = 1 + 2 𝑏𝑚1 + 𝑐𝑚1 + 𝑑𝑚1 , 𝑏𝑚1 = 2 𝜌 𝐶𝑝
, 𝑐𝑚1 = 2𝜌 𝑎𝑛𝑑 𝑑𝑚1 =
𝑚 𝑚 𝑉𝑚1 𝑅𝑚 4 𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑝 3
𝑑𝑡
2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑚 7

The final equation 4.14 is obtained

𝑡−𝑑𝑡
𝑇𝑚1,𝑖 = 𝑎𝑚1 ̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅̅
𝑇𝑚1,𝑖 − 𝑏𝑚1 (𝑇 ̅̅̅̅̅̅̅̅̅ ̅̅̅̅ ̅̅̅̅̅̅
𝑚1,𝑖−1 + 𝑇𝑚1,𝑖+1 ) − 𝑐𝑚1 𝑇𝑝,𝑖 − 𝑑𝑚1 𝑇𝑚2,𝑖 equation 4.14

For node i=1:

2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 ̅̅̅̅̅ ̅̅̅̅̅̅̅̅

𝑇𝑝,1 −𝑇𝑚1,1 𝑇 −𝑇𝑚1,1 ̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅
𝑇 −𝑇𝑚1,1 𝑇 ̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅
−𝑇𝑚1,1 ̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅
̅̅̅̅̅̅̅
(𝑇 𝑡−𝑑𝑡
𝑚1,1 − 𝑇𝑚1,1 ) = + 𝑚1,0 + 𝑚1,2 + 𝑚2,1
𝑑𝑡 𝑅 𝑝3 𝑅 𝑅 𝑚1 𝑅 𝑚4 𝑚7

𝑑𝑡 1 1 1 1 𝑑𝑡
𝑡−𝑑𝑡
𝑇𝑚1,1 = [1 + ( ) ( + + + )] ̅̅̅̅̅̅̅
𝑇𝑚1,1 − ̅̅̅̅̅̅̅
(𝑇 𝑚1,0 ) −
2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑝 3 𝑅𝑚 1 𝑅𝑚 4 𝑅𝑚 7 2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑚 1
𝑑𝑡 𝑑𝑡 𝑑𝑡
̅̅̅̅̅̅̅
(𝑇 𝑚1,2 ) − 2 𝜌
̅̅̅̅̅
𝑇𝑝,1 − 2 𝜌 ̅̅̅̅̅̅̅
𝑇𝑚2,1 ,
2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑚 4 𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑝 3 𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑚 7

Setting:
1
𝑎𝑚1 1 = 1 + 𝑏𝑚1 + 𝑐𝑚1 + 𝑑𝑚1 + 𝑒𝑚1 𝑤𝑖𝑡ℎ 𝑒𝑚1 = 2 𝜌 𝐶𝑝
𝑚 𝑚 𝑉𝑚1 𝑅𝑚 1

𝑡−𝑑𝑡
𝑇𝑚1,1 = 𝑎𝑚1 1 ̅̅̅̅̅̅̅
𝑇𝑚1,1 − 𝑏𝑚1 ̅̅̅̅̅̅̅
𝑇𝑚1,2 − 𝑐𝑚1 ̅̅̅̅̅
𝑇𝑝,1 − 𝑑𝑚1 ̅̅̅̅̅̅̅
𝑇𝑚2,1 − 𝑒𝑚1 ̅̅̅̅
𝑇𝑖𝑛 equation 4.15

For node i=n:

2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 ̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅

𝑇𝑝,𝑛 −𝑇𝑚1,𝑛 𝑇 −𝑇 ̅̅̅̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅
𝑇 −𝑇 ̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅
̅̅̅̅̅̅̅
(𝑇 𝑡−𝑑𝑡
𝑚1,𝑛 − 𝑇𝑚1,𝑛 ) = + 𝑚1,𝑛−1 𝑚1,𝑛 + 𝑚2,𝑛 𝑚1,𝑛
𝑑𝑡 𝑅 𝑝3 𝑅 𝑅 𝑚4 𝑚7

𝑑𝑡 1 1 1 𝑑𝑡
𝑡−𝑑𝑡
𝑇𝑚1,𝑛 = [1 + (2 𝜌 𝐶𝑝
) (𝑅 +𝑅 +𝑅 )] ̅̅̅̅̅̅̅
𝑇𝑚1,𝑛 − 2 𝜌 ̅̅̅̅̅̅̅̅̅̅
(𝑇 𝑚1,𝑛−1 ) −
𝑚 𝑚 𝑉𝑚1 𝑝3 𝑚4 𝑚7 𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑚 4
𝑑𝑡 𝑑𝑡
̅̅̅̅̅
𝑇𝑝,𝑛 − 2 𝜌 ̅̅̅̅̅̅̅
𝑇𝑚2,𝑛
2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑝 3 𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑚 7

117
Setting:

𝑎𝑚1 𝑛 = 1 + 𝑏𝑚1 + 𝑐𝑚1 + 𝑑𝑚1

𝑡−𝑑𝑡
𝑇𝑚1,𝑛 = 𝑎𝑚1 𝑛 ̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅̅̅
𝑇𝑚1,𝑛 − 𝑏𝑚1 𝑇 ̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
𝑚1,𝑛−1 − 𝑐𝑚1 𝑇𝑝1,𝑛 − 𝑑𝑚1 𝑇𝑚2,𝑛 equation 4.16

Equation 4.14, 4.15 and 4.16 can then be written in a matrix equation, representing nodes 1 to n:
𝑡−𝑑𝑡
𝑇𝑚1,1 𝑎𝑚1 1 −𝑏𝑚1 ⋯ ⋯ 0 ̅̅̅̅̅̅̅
𝑇𝑚1,1 −𝑐𝑚1 0 ⋯ ⋯ 0 ̅̅̅̅̅
𝑇𝑝,1
⋮ −𝑏𝑚1 𝑎𝑚1 ⋱ ⋮ ⋮ 0 ⋱ ⋱ ⋮ ⋮
𝑡−𝑑𝑡
𝑇𝑚1,𝑖 = 0 ⋱ 𝑎𝑚1 ⋱ ⋮ × ̅̅̅̅̅̅
𝑇𝑚1,𝑖 + ⋮ ⋱ −𝑐𝑚1 ⋱ ⋮ × ̅̅̅̅
𝑇𝑝,𝑖
⋮ ⋮ ⋱ 𝑎𝑚1 −𝑏𝑚1 ⋮ ⋮ ⋱ ⋱ 0 ⋮
𝑡−𝑑𝑡
[𝑇𝑚1,𝑛 ] [ 0 ⋯ 0 ̅̅̅̅̅̅̅
−𝑏𝑚1 𝑎𝑚1 𝑛 ] [𝑇𝑚1,𝑛 ] [ 0 ⋯ ⋯ 0 ̅̅̅̅̅
−𝑐𝑚1 ] [𝑇 𝑝,𝑛 ]
−𝑑𝑚1 0 ⋯ ⋯ 0 ̅̅̅̅̅̅̅
𝑇𝑚2,1 −𝑒𝑚1 𝑇𝑖𝑛
0 ⋱ ⋱ ⋮ ⋮ 0
+ ⋮ ⋱ −𝑑𝑚1 ⋱ ⋮ × ̅̅̅̅̅̅
𝑇𝑚2,𝑖 + ⋮
⋮ ⋱ ⋱ 0 ⋮ ⋮
[ 0 ⋯ ⋯ 0 −𝑑𝑚1 ] [̅̅̅̅̅̅̅
𝑇𝑚2,𝑛 ] [ 0 ]

[Link].2 Second PCM node (internal PCM volume)

The application of the heat balance equation for the ith and first crosswise (j=2) PCM node results in the
following equation:

𝑑𝑇𝑚2,𝑖 𝑇𝑚1,𝑖 −2 𝑇𝑚2,𝑖 +𝑇𝑚3,𝑖 𝑇𝑚2,𝑖−1 −2 𝑇𝑚2,𝑖 +𝑇𝑚2,𝑖+1

𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚2 = + equation 4.17
𝑑𝑡 𝑅𝑚 7 𝑅𝑚 5

Where

Vm2 = Sm2 ∙ dx ∶ Volume of the second PCM node [m3]

dym
Sm2 = 2
∙ dz ∶ Contact surface between neighboring PCM nodes [m2]

Tm ,i ∶ Third PCM temperature at node i [°C]

Tm2,i−1 ∶ First PCM temperature at node i-1 [°C]
Tm2,i+1 ∶ First PCM temperature at node i+1 [°C]
dx
Rm = S ∶ : Thermal resistance between neighboring PCM nodes [W/K]
m2 ∙λm

By introducing the average temperature term, equation 4.17 becomes:

2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚2 ̅̅̅̅̅̅

𝑇 ̅̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅̅
𝑚1,𝑖 − 2 𝑇𝑚2,𝑖 + 𝑇𝑚 ,𝑖 𝑇𝑚2,𝑖−1 − 2 ̅̅̅̅̅̅
𝑇𝑚2,𝑖 + ̅̅̅̅̅̅̅̅̅
𝑇𝑚2,𝑖+1
̅̅̅̅̅̅
(𝑇 𝑚2,𝑖 − 𝑇 𝑡−𝑑𝑡
𝑚2,𝑖 ) = +
𝑑𝑡 𝑅𝑚 𝑅𝑚

118
 𝑡−𝑑𝑡
Out of which, 𝑇𝑚2,𝑖 can be expressed as a function of ̅̅̅̅̅̅
𝑇𝑚1,𝑖 , ̅̅̅̅̅̅̅̅̅
𝑇𝑚2,𝑖−1 , ̅̅̅̅̅̅̅̅̅
𝑇𝑚2,𝑖+1 and ̅̅̅̅̅̅
𝑇𝑚 ,𝑖 :

𝑑𝑡 2 2 𝑑𝑡
𝑡−𝑑𝑡
𝑇𝑚2,𝑖 = [1 + ( ) ( + )] ̅̅̅̅̅̅
𝑇𝑚1,𝑖 − ̅̅̅̅̅̅̅̅̅
(𝑇 ̅̅̅̅̅̅̅̅̅
𝑚2,𝑖−1 + 𝑇𝑚2,𝑖+1 )
2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚2 𝑅𝑚 𝑅𝑚 2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚2 𝑅𝑚
𝑑𝑡
− (𝑇̅̅̅̅̅̅ ̅̅̅̅̅̅
𝑚1,𝑖 + 𝑇𝑚 ,𝑖 )
2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚2 𝑅𝑚

Setting:
𝑑𝑡 𝑑𝑡
𝑎𝑚2 = 1 + 2 𝑏𝑚2 + 2 𝑑𝑚2 , 𝑏𝑚2 = 2 𝜌 𝐶𝑝𝑚 𝑉𝑚2 𝑅𝑚 5
𝑎𝑛𝑑 𝑑𝑚2 = 2 𝜌 𝐶𝑝𝑚 𝑉𝑚2 𝑅𝑚 7
𝑚 𝑚

The final equation 4.18 is obtained:

𝑡−𝑑𝑡
𝑇𝑚2,𝑖 = 𝑎𝑚2 ̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅̅
𝑇𝑚2,𝑖 − 𝑏𝑚2 (𝑇 ̅̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅ ̅̅̅̅̅̅
𝑚2,𝑖−1 + 𝑇𝑚2,𝑖+1 ) − 𝑑𝑚2 𝑇𝑚1,𝑖 − 𝑑𝑚2 𝑇𝑚 ,𝑖 equation 4.18

For node i=1

1
𝑎𝑚2 1 = 1 + 𝑏𝑚2 + 2 𝑑𝑚2 + 𝑒𝑚2 𝑤𝑖𝑡ℎ 𝑒𝑚2 = 2 𝜌 𝐶𝑝𝑚 𝑉𝑚2 𝑅𝑚 2
𝑚

Obtaining:
𝑡−𝑑𝑡
𝑇𝑚2,1 = 𝑎𝑚2 1 ̅̅̅̅̅̅̅
𝑇𝑚2,1 − 𝑏𝑚2 ̅̅̅̅̅̅̅
𝑇𝑚2,2 − 𝑑𝑚2 ̅̅̅̅̅̅̅
𝑇𝑚1,1 − 𝑑𝑚2 ̅̅̅̅̅̅̅
𝑇𝑚 ,1 − 𝑒𝑚2 ̅̅̅̅
𝑇𝑖𝑛 equation 4.19

For node i=n:

𝑎𝑚2 𝑛 = 1 + 𝑏𝑚2 + 2 𝑑𝑚2

Obtaining:
𝑡−𝑑𝑡
𝑇𝑚2,𝑛 = 𝑎𝑚2 𝑛 ̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅̅̅
𝑇𝑚2,𝑛 − 𝑏𝑚2 𝑇 ̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅
𝑚2,𝑛−1 − 𝑑𝑚2 𝑇𝑚1,𝑛 − 𝑑𝑚2 𝑇𝑚 ,𝑛 equation 4.20

Equation 4.18, 4.19 and 4.20 can then be written in a matrix equation, representing nodes 1 to n:
𝑡−𝑑𝑡
𝑇𝑚2,1 𝑎𝑚2 1 −𝑏𝑚2 ⋯ ⋯ 0 ̅̅̅̅̅̅̅
𝑇𝑚2,1 −𝑑𝑚2 0 ⋯ ⋯ 0
⋮ −𝑏𝑚2 𝑎𝑚2 ⋱ ⋮ ⋮ 0 ⋱ ⋱ ⋮
𝑡−𝑑𝑡
𝑇𝑚2,𝑖 = 0 ⋱ 𝑎𝑚2 ⋱ ⋮ × ̅̅̅̅̅̅
𝑇𝑚2,𝑖 + ⋮ ⋱ −𝑑𝑚2 ⋱ ⋮
⋮ ⋮ ⋱ 𝑎𝑚2 −𝑏𝑚2 ⋮ ⋮ ⋱ ⋱ 0
𝑡−𝑑𝑡
[𝑇𝑚2,𝑛 ] [ 0 ⋯ 0 −𝑏𝑚2 𝑎𝑚2 𝑛 ] [̅̅̅̅̅̅̅
𝑇𝑚2,𝑛 ] [ 0 ⋯ ⋯ 0 −𝑑𝑚2 ]
̅̅̅̅̅̅̅
𝑇𝑚1,1 −𝑑𝑚2 0 ⋯ ⋯ 0 ̅̅̅̅̅̅̅
𝑇𝑚 ,1 −𝑒𝑚2 𝑇𝑖𝑛
⋮ 0 ⋱ ⋱ ⋮ ⋮ 0
× ̅̅̅̅̅̅
𝑇𝑚1,𝑖 + ⋮ ⋱ −𝑑𝑚2 ⋱ ⋮ × 𝑇̅̅̅̅̅̅
𝑚 ,𝑖 + ⋮
⋮ ⋮ ⋱ ⋱ 0 ⋮ ⋮
[̅̅̅̅̅̅̅
𝑇𝑚1,𝑛 ] [ 0 ⋯ ⋯ 0 −𝑑𝑚2 ] [̅̅̅̅̅̅̅
𝑇𝑚 ,𝑛 ] [ 0 ]

119
[Link].3 Third PCM node (final PCM volume located at the middle of the PCM layer)
The application of the heat balance equation for the ith and first crosswise (j=3) PCM node results in the
following equation:

𝑑𝑇𝑚3,𝑖 𝑇𝑚2,𝑖 −𝑇𝑚3,𝑖 𝑇𝑚3,𝑖−1 −2 𝑇𝑚3,𝑖 +𝑇𝑚3,𝑖+1

𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚 = + equation 4.21
𝑑𝑡 𝑅𝑚 7 𝑅𝑚 6

Where

Vm = Sm ∙ dx ∶ Volume of the third PCM node [m3]

dym
Sm2 = ∙ dz ∶ Contact surface between neighboring PCM nodes [m2]

Tm ,i−1 ∶ Third PCM temperature at node i-1 [°C]

Tm ,i+1 ∶ Third PCM temperature at node i+1 [°C]

By introducing the average temperature term, equation 4.21 becomes:

2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚 ̅̅̅̅̅̅
𝑇 ̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅̅
𝑚2,𝑖 − 𝑇𝑚 ,𝑖
̅̅̅̅̅̅
𝑇𝑚 ,𝑖−1 − 2 𝑇 ̅̅̅̅̅̅̅̅̅
𝑚 ,𝑖 + 𝑇𝑚 ,𝑖+1
̅̅̅̅̅̅
(𝑇 𝑡−𝑑𝑡
𝑚 ,𝑖 − 𝑇𝑚 ,𝑖 ) = +
𝑑𝑡 𝑅𝑚, 𝑅𝑚

Setting:
𝑑𝑡 𝑑𝑡
𝑎𝑚 = 1 + 2 𝑏𝑚 + 𝑑𝑚 , 𝑏𝑚 = 2 𝜌 𝐶𝑝𝑚 𝑉𝑚3 𝑅𝑚 6
𝑎𝑛𝑑 𝑑𝑚 = 2 𝜌 𝐶𝑝𝑚 𝑉𝑚3 𝑅𝑚,7
𝑚 𝑚

The final equation 4.22 is obtained:

𝑡−𝑑𝑡
𝑇𝑚 ̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅
,𝑖 = 𝑎𝑚 𝑇𝑚 ,𝑖 − 𝑏𝑚 (𝑇𝑚 ,𝑖−1 + 𝑇𝑚 ,𝑖+1 ) − 𝑑𝑚 𝑇𝑚2,𝑖 equation 4.22

For nodes i=1 and i=n

1
𝑎𝑚 1 = 1 + 𝑏𝑚 + 𝑑𝑚 + 𝑒𝑚 , 𝑎𝑚 𝑛 = 1 + 𝑏𝑚 + 𝑑𝑚 𝑤𝑖𝑡ℎ 𝑒𝑚 =
2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚3 𝑅𝑚 3

Obtaining:
𝑡−𝑑𝑡
𝑇𝑚 ,1 = 𝑎𝑚
̅̅̅̅̅̅̅
𝑇𝑚 ,1 − 𝑏𝑚 ̅̅̅̅̅̅̅
𝑇𝑚 ,2 − 𝑑𝑚 ̅̅̅̅̅̅̅
𝑇𝑚2,1 − 𝑒𝑚 ̅̅̅̅
𝑇𝑖𝑛 equation 4.23
1

and
𝑡−𝑑𝑡
𝑇𝑚 ,𝑛 = 𝑎𝑚 𝑛
̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅̅̅
𝑇𝑚 ,𝑛 − 𝑏𝑚 𝑇 ̅̅̅̅̅̅̅
𝑚 ,𝑛−1 − 𝑑𝑚 𝑇𝑚2,𝑛 equation 4.24

Equations 4.22, 4.23 and 4.24 can then be written in a matrix equation, representing nodes 1 to n:

120
 𝑡−𝑑𝑡
𝑇𝑚 ,1 𝑎𝑚 1 −𝑏𝑚 ⋯ ⋯ 0 ̅̅̅̅̅̅̅
𝑇𝑚 ,1 −𝑑𝑚 0 ⋯ ⋯ 0
⋮ −𝑏𝑚 𝑎𝑚 ⋱ ⋮ ⋮ 0 ⋱ ⋱ ⋮
𝑡−𝑑𝑡
𝑇𝑚 ,𝑖 = 0 ⋱ 𝑎𝑚 ⋱ ⋮ × ̅̅̅̅̅̅
𝑇𝑚 ,𝑖 + ⋮ ⋱ −𝑑𝑚 ⋱ ⋮
⋮ ⋮ ⋱ 𝑎𝑚 −𝑏𝑚 ⋮ ⋮ ⋱ ⋱ 0
𝑡−𝑑𝑡
[𝑇𝑚 ,𝑛 ] [ 0 ⋯ 0 −𝑏𝑚 ̅̅̅̅̅̅̅
𝑇
𝑎𝑚 𝑛 ] [ 𝑚 ,𝑛 ] [ 0 ⋯ ⋯ 0 −𝑑𝑚 ]
̅̅̅̅̅̅̅
𝑇𝑚2,1 −𝑒𝑚 𝑇𝑖𝑛
⋮ 0
× ̅̅̅̅̅̅
𝑇𝑚2,𝑖 + ⋮
⋮ ⋮
[̅̅̅̅̅̅̅
𝑇𝑚2,𝑛 ] [ 0 ]

[Link].4 Centralized matrix and solution

In the end, the matrix equations obtained from each layer are assembled into the following centralized
matrix equation:

𝑇̅ = 𝐴−1 (𝑇 𝑡−𝑑𝑡 − 𝐵) equation 4.25

Where
𝑇̅ is a n*(m+2) column matrix containing average temperature values for the air, aluminum and PCM
layers,
A is a n*(m+2) x n*(m+2) matrix containing thermal resistance coefficients,
𝑇 𝑡−𝑑𝑡 is a n*(m+2) column matrix containing temperature values for the three layers at the previous
time step and
B is a n*(m+2) containing initial temperature values for the three layers.
Finally, air, aluminum and PCM temperatures at each node and at each time step are calculated using
inlet air temperature, neighboring nodes temperature and temperatures calculated at the previous time
step with the final equation:

𝑇 𝑡 = 2 𝑇̅ − 𝑇 𝑡−𝑑𝑡 equation 4.26

In the end, the use of the heat balance approach and the finite difference method as presented in the
previous subsections was reduced to the solution of the matrix equation shown in figure 4.4.

121
 𝑡−𝑑𝑡
𝑇𝑎,1 𝑎𝑎 0 ⋯ ⋯ 0 −𝑐𝑎 0 ⋯ ⋯ 0 0 ⋯ 0 0 ⋯ 0 0 ⋯ 0 ̅̅̅̅̅
𝑇𝑎,1
⋮ −𝑏𝑎 ⋱ ⋱ ⋮ 0 ⋱ ⋱ ⋮ ⋮ −𝑏𝑎 𝑇𝑖𝑛
𝑡−𝑑𝑡
𝑇𝑎,𝑖 0 ⋱ 𝑎𝑎 ⋱ ⋮ ⋮ ⋱ −𝑐𝑎 ⋱ ⋮ ⋮ ⋱ ⋮ ⋮ ⋱ ⋮ ⋮ ⋱ ⋮ ̅̅̅̅
𝑇𝑎,𝑖 0
⋮ ⋮ ⋱ ⋱ 0 ⋮ ⋱ ⋱ 0 ⋮ ⋮
𝑡−𝑑𝑡
𝑇𝑎,𝑛 0 ⋯ 0 −𝑏𝑎 𝑎𝑎 0 ⋯ ⋯ 0 −𝑐𝑎 0 ⋯ 0 0 ⋯ 0 0 ⋯ 0 ̅̅̅̅̅
𝑇𝑎,𝑛 ⋮
𝑡−𝑑𝑡
𝑇𝑝,1 −𝑐𝑝 0 ⋯ ⋯ 0 𝑎𝑝1 −𝑏𝑝 ⋯ ⋯ 0 −𝑑𝑝 0 ⋯ ⋯ 0 ̅̅̅̅̅ 0
0 ⋯ 0 0 ⋯ 0 𝑇𝑝,1
⋮ 0 ⋱ ⋱ ⋮ −𝑏𝑝 𝑎𝑝 ⋱ ⋮ 0 ⋱ ⋱ ⋮ ⋮ −𝑒𝑝 𝑇𝑖𝑛
𝑡−𝑑𝑡
𝑇𝑝,𝑖 ⋮ ⋱ −𝑐𝑝 ⋱ ⋮ 0 ⋱ 𝑎𝑝 ⋱ ⋮ ⋮ ⋱ −𝑑𝑝 ⋱ ⋮ ⋮ ⋱ ⋮ ⋮ ⋱ ⋮ ̅̅̅̅
𝑇𝑝,𝑖 0
⋮
⋮ ⋮ ⋱ ⋱ 0 ⋮ ⋱ 𝑎𝑝 −𝑏𝑝 ⋮ ⋱ ⋱ 0 ⋮
𝑡−𝑑𝑡 0 ⋯ 0 0 ⋯ 0 ̅̅̅̅̅̅̅ ⋮
𝑇𝑝,𝑛 0 ⋯ ⋯ 0 −𝑐𝑝 0 ⋯ 0 −𝑏𝑝 𝑎𝑝𝑛 0 ⋯ ⋯ 0 −𝑑𝑝 𝑇𝑝,𝑛
0
𝑡−𝑑𝑡
𝑇𝑚1,1 −𝑐𝑚1 0 ⋯ ⋯ 0 𝑎𝑚1 1 −𝑏𝑚1 ⋯ ⋯ 0 −𝑑𝑚1 0 ⋯ ⋯ 0 ̅̅̅̅̅̅̅
𝑇𝑚1,1
0 ⋯ 0 0 ⋯ 0 −𝑒𝑚1 𝑇𝑖𝑛
⋮ 0 ⋱ ⋱ ⋮ −𝑏𝑚1 𝑎𝑚1 ⋱ ⋮ 0 ⋱ ⋱ ⋮ ⋮ 0
𝑡−𝑑𝑡
𝑇𝑚1,𝑖 = ⋮ ⋱ −𝑐𝑚1 ⋱ ⋮ 0 ⋱ 𝑎𝑚1 ⋱ ⋮ ⋮ ⋱ −𝑑𝑚1 ⋱ ⋮ ̅̅̅̅̅̅
× 𝑇 𝑚1,𝑖 + ⋮
⋮ ⋱ ⋮ ⋮ ⋱ ⋮
⋮ ⋮ ⋱ ⋱ 0 ⋮ ⋱ 𝑎𝑚1 −𝑏𝑚1 ⋮ ⋱ ⋱ 0 ⋮ ⋮
𝑡−𝑑𝑡
𝑇𝑚1,𝑛 0 ⋯ 0 0 ⋯ ⋯ 0 −𝑐𝑚1 0 ⋯ 0 −𝑏𝑚1 𝑎𝑚1 𝑛 0 ⋯ ⋯ 0 −𝑑𝑚1 0 ⋯ 0 ̅̅̅̅̅̅̅
𝑇𝑚1,𝑛 0
𝑡−𝑑𝑡
𝑇𝑚2,1 −𝑑𝑚2 0 ⋯ ⋯ 0 𝑎 −𝑏 ⋯ ⋯ 0 −𝑑𝑚2 0 ⋯ ⋯ 0 ̅̅̅̅̅̅̅
𝑇𝑚2,1 −𝑒𝑚2 𝑇𝑖𝑛
0 ⋯ 0 0 ⋯ 0 𝑚2 1 𝑚2
⋮ 0 ⋱ ⋱ ⋮ −𝑏𝑚2 𝑎𝑚2 ⋱ ⋮ 0 ⋱ ⋱ ⋮ ⋮ 0
𝑡−𝑑𝑡
𝑇𝑚2,𝑖 ̅̅̅̅̅̅
𝑇 ⋮
⋮ ⋱ ⋮ ⋮ ⋱ ⋮ ⋮ ⋱ −𝑑𝑚2 ⋱ ⋮ 0 ⋱ 𝑎𝑚2 ⋱ ⋮ ⋮ ⋱ −𝑑𝑚2 ⋱ ⋮ 𝑚2,𝑖
⋮ ⋮
⋮ ⋮ ⋱ ⋱ 0 ⋮ ⋱ 𝑎𝑚2 −𝑏𝑚2 ⋮ ⋱ ⋱ 0
𝑡−𝑑𝑡 ̅̅̅̅̅̅̅
𝑇 0
𝑇𝑚2,𝑛 0 ⋯ 0 0 ⋯ 0 0 ⋯ ⋯ 0 −𝑑𝑚2 0 ⋯ 0 −𝑏𝑚2 𝑎𝑚2 𝑛 0 ⋯ ⋯ 0 −𝑑𝑚2 𝑚2,𝑛
̅̅̅̅̅̅̅ −𝑒𝑚 𝑇𝑖𝑛
𝑇𝑚𝑡−𝑑𝑡 −𝑑 0 ⋯ ⋯ 0 −𝑑𝑚 0 ⋯ ⋯ 0 𝑇𝑚 ,1
,1 0 ⋯ 0 0 ⋯ 0 0 ⋯ 0 𝑚 0
⋮ 0 ⋱ ⋱ ⋮ 0 ⋱ ⋱ ⋮ ⋮ ⋮
𝑇𝑚𝑡−𝑑𝑡 ̅̅̅̅̅̅
𝑇
,𝑖 ⋮ ⋱ ⋮ ⋮ ⋱ ⋮ ⋮ ⋱ ⋮ ⋮ ⋱ −𝑑𝑚 ⋱ ⋮ ⋮ ⋱ −𝑑𝑚 ⋱ ⋮ 𝑚 ,𝑖 ⋮
⋮ ⋮ ⋱ ⋱ 0 ⋮ ⋱ ⋱ 0 ⋮ [ 0 ]
̅̅̅̅̅̅̅
𝑡−𝑑𝑡
[𝑇𝑚 ,𝑛 ] [ 0 ⋯ 0 0 ⋯ 0 0 ⋯ 0 0 ⋯ ⋯ 0 −𝑑𝑚 0 ⋯ ⋯ 0 −𝑑 ] [𝑇𝑚 ,𝑛 ]
𝑚

A B
Figure 4.4 Centralized matrix equation

122
4.3.6 Model assumptions
The main objective of the numerical study is to develop a model that uses as entries the air temperature
and airflow rate and calculates the outlet air temperature, emphasizing on the calculation speed and not
necessarily on the complexity of the phenomena taking place inside the PCM. In view of this fact, several
assumptions and simplifications were made. First of all, the plates are considered to be fully filled with
PCM, whereas in reality a small percentage was left vacant to avoid leakage risks. Airflow is considered
equal for all air layers; this is performed by simply dividing the initial total airflow by the number of air
layers. In reality and as presented in the previous chapter small variations may exist.
As seen in figure 4.2, the model calculates heat transfers in two dimensions (lengthwise and crosswise),
considering that same conditions are met height wise. The model calculates the temperature evolution
for half of the air and PCM layers, assuming that by symmetry the results can be extended to the other
half (where similar conditions apply). The model considers convection transfers between the air and the
aluminum layers and conduction transfers for aluminum and PCM layers. However, it does not take into
account convection inside the PCM during the melting and the liquid phases, as well as radiative
emissions between plates.
Concerning the volume change between phases, the model integrates a change of the PCM density value
but does not consider the increase of the PCM level inside the container.
Finally, the convective heat transfer coefficient was treated as a calibration factor and was empirically
determined at 50 W/m2K for an airflow rate of 100m3/h1, with a corresponding increase of 5 W/m2K for
every 100m3/h augmentation of the airflow rate.

4.4 Numerical model calibration, specific heat capacity importance

As previously mentioned, a partner of the Stock-Air project (EDF R&D Matériaux et Mécanique des
Composants) was assigned to determine the properties of the Microtek 37D properties, including
thermal conductivity, density, heat capacity and phase change temperature range. Table 3.1 presents
these values for the solid and liquid phases of the paraffin. The value of each property was altered at
each node, according to the PCM temperature using a Matlab interpolation function (interp1).
These values were used in the developed numerical model, providing promising but not satisfying
results, when compared to experimental ones. After various factors were investigated, the importance
of the PCM heat capacity (Cp) values was made clear.

4.4.1 DSC Method

The heat capacity values of the PCM were obtained by EDF R&D Matériaux et Mécanique des
Composants using the Differential Scanning Calorimetry (DSC) method. DSC is a measuring technique
where a sample material (in this case paraffin) is compared to a reference material (with known
properties) while undergoing a heating or cooling process. More specifically, known masses of the
sample and of the reference material are placed on two symmetrically mounted sample holders (or
pans). Two heating elements are used to heat the materials at a constant rate. A temperature control

123
system measures the temperature difference between the two pans and adjusts this difference to zero
by controlling a differential component of the total heating power. The measurement of this differential
power and the knowledge of the heat capacity value of the reference material allow the calculation of
the heat capacity of the sample material.

Figure 4.5: Representation of the DSC method (Borderon, 2012)

For the Microtek 37D paraffin, measurements were made on a 0.22623 g sample, for three different
heating/cooling rates of 0.15 °C/min, 0.5 °C/min and 1 °C/min. Results are illustrated in figure 4.6. During
the heating process, the phase change occurs between 30°C and 42 °C, shifting towards higher
temperatures as the heating rate increases. During the cooling process, the phase change occurs
between 25 and 35°C, shifting towards lower temperatures as the cooling rate increases.

Cp heating 1°C/min
120 Cp heating 0,5°C/min
Cp heating 0,15°C/min
70 Cp cooling 1°C/min
Cp cooling 0,5°C/min
Cp cooling 0,15°C/min
20
Cp [J/(g*°C]

-30

-80

-130

-180
20 25 30 35 40 45 50 55 60
Temperature [°C]

Figure 4.6: DSC obtained heat capacity curves for three different heating/cooling rates

124
The integration of the heat capacity values over temperature equals the enthalpy change for a
determined increase/decrease of temperature, for example from T1 to T2, as shown in equation 4.27:
𝑇
𝛥𝐻 = ∫𝑇 2 𝐶𝑝 𝑑𝑇 equation 4.27
1

The enthalpy change for the six heat capacity curves obtained through the DSC method was calculated in
Matlab using the trapezoidal numerical integration function (trapz). This function integrates the heat
capacity values with temperature spacing increment by breaking the area down into trapezoids with
more easily computable areas. Results are presented in table 4.1.

Heating / Cooling rate

0,15 °C/min 0,5 °C/min 1 °C/min

Enthalpy change during

364,23 361,53 356,35
the heating process (J/g)

Enthalpy change during

-362,38 -362,65 -360,59
the cooling process (J/g)
Table 4.1: Enthalpy change during heating/cooling of the PCM based on DSC results

Even though six different curves were integrated, the obtained enthalpy change values are similar for the
different heating/cooling rates. Three options were then offered as heat capacity values for the
numerical model. The ones corresponding to the 0,5°C/min heating/cooling rates were initially selected
and a full thermal cycle of charging and discharging was numerically reproduced in order to compare the
results with the experimental ones. The model uses as inputs the experimental inlet air temperature and
airflow rate values to calculate the downstream temperature evolution of the air, aluminum and PCM
layer for every time step dt=10s. An interpolation Matlab function was used to associate heat capacity
values to temperature values. Furthermore, the model switches between the heating curve and the
cooling one by comparing the PCM temperature value during two consecutive time steps.

Discussion of simulation results using DSC heat capacity values

Simulations were carried out for five airflow rates, in accordance to the experimental characterization
tests of the heat exchanger. Results for an airflow rate of 300m3/h are presented here as they are
representative of the rest. Therefore, the two following figures illustrate the comparison between
experimental and numerical results for the inlet and outlet air temperature values (figure 4.7) as well as
the PCM temperature values at the middle of the PCM layer (node n=5 lengthwise, figure 4.8).

Results show that the numerical model underestimates the temperature values, this occurring with an
increasing rate downstream. No distinct end of the phase change process is observed (end of the melting
procedure), as the temperature values increase with an almost constant rate after the beginning of the
melting phase. All this results in lower temperatures at the end of the charging phase and higher
temperatures at the end of the discharging phase, compared to experimental data. Lastly, during the
discharge phase, two peaks appear in the experimental data for the PCM temperature values, linked to

125
the intrinsic properties of the paraffin (figures 4.8 and 3.17). However, and even though the heat
capacity values used during simulation present two peaks, numerical results do not reproduce the
experimentally observed phenomenon (figure 4.8).

50
Tair,in exp 300 m3/h (°C)
45 Tair,in sim 300 m3/h (°C)
Tair,out exp 300 m3/h (°C)
40
Temperature [°C]

Tair,out sim 300 m3/h (°C)

35

30

25

20
0 5 10 15 20
Time [h]

Figure 4.7: Simulation results for inlet and outlet air temperature using DSC obtained heat capacity
curves and comparison with experimental data

50
Tpcm,5 exp 300 m3/h (°C)
45
Tpcm,5 sim 300 m3/h (°C)

40
Temperature [°C]

35

30

25

20
0 5 10 15 20
Time [h]

Figure 4.8:Simulation results for PCM temperature at node 5 using DSC obtained heat capacity curves
and comparison with experimental data

126
Albright et al. conclude that the DSC process can provide inaccurate data due to undesired thermal
gradients within the sample and that corrections should be made. In addition, as noted in the final report
of IEA Annex 17, DSC results depend on the heating rate. This dependency was also observed for the
PCM in question and the results obtained using the DSC method, where different curves are acquired for
different heating/cooling rates (figure 4.6 again).

4.4.2 Creation of adequate heat capacity values

The aforementioned observations led to the investigation of the heat capacity values and the creation of
adequate curves based on DSC results. This was performed through a several steps process, each time
observing the impact of the created heat capacity curves on the simulation results and comparing them
to experimental data. This process is presented in the following subsections.

4.4.3 First approach: isosceles triangle

The first approach consisted in representing the heat capacity values with a curve having the form of an
isosceles triangle (figure 4.9). The same curve was used for the heating and cooling processes. The
underestimation of numerical temperature values pointed to a decrease of the heat capacity values, as,
physically speaking, this quantity represents the amount of heat needed to raise a material’s
temperature the temperature of a material by one degree. Indeed, the calculated enthalpy change
(integration of heat capacity over temperature) for the created heat capacity curve was 220 J/g,
significantly lower than the DSC values (around 360 J/g, table 4.1).

25

Cp heating/cooling

20

15
Cp [J/(g*°C]

10

0
0 10 20 30 40 50 60 70 80
Temperature [°C]

Figure 4.9: Fictive Heat capacity values, first approach

127
Discussion of simulation results using isosceles triangle shaped heat capacity values
The numerical reproduction of the characterization tests demonstrated that this modification was on the
right path. Figures 4.10 and 4.11 illustrate, as before, the comparison between experimental and
numerical results for the inlet and outlet air temperature values as well as the PCM temperature values
at the middle of the PCM layer (node n=5 lengthwise). This time, the model managed to provide results
closer to the experimental ones. Nevertheless, the two peaks observation during the experimental
discharge phase was still not apparent in the numerical results.

50
Tair,in exp 300 m3/h (°C)
45 Tair,in sim 300 m3/h (°C)
Tair,out exp 300 m3/h (°C)
Temperature [°C]

40
Tair,out sim 300 m3/h (°C)
35

30

25

20
0 5 10 15 20
Time [h]

Figure 4.10: Simulation results for inlet and outlet air temperature using first approach’s heat capacity
curves and comparison with experimental data

50

45
Tpcm,5 exp 300 m3/h (°C)
Temperature [°C]

40 Tpcm,5 sim 300 m3/h (°C)

35

30

25

20
0 5 10 15 20
Time [h]

Figure 4.11: Simulation results for PCM temperature at node 5 using first approach’s heat capacity
curves and comparison with experimental data

128
4.4.4. Second approach: two peaks curve for the discharge phase
A second approach was considered, creating a right triangle for the charging phase and an irregular
polyhedron for the discharging phase. Two peaks were created for the discharging phase. An effort was
made to maintain the same enthalpy change value. After calculation ΔH,heating= 220,58 J/g and
ΔH,cooling= -234,79 J/g.

40

30 Cp heating

Cp cooling
20

10
Cp [J/(g*°C]

-10

-20

-30

-40

-50
0 10 20 30 40 50 60 70 80
Temperature [°C]

Figure 4.12: Fictive Heat capacity values, second approach

Discussion of simulation results using a two peaks curve for the discharge phase
Results of the second approach are presented in figure 7.13 and 7.14, illustrating the comparison
between experimental and numerical results for the inlet and outlet air temperature values as well as
the PCM temperature values at the middle of the PCM layer (node n=5 lengthwise). The introduction of
two peaks in the heat capacity values resulted in the apparition of two peaks in the PCM temperature
evolution curve (figure 7.14). However, these peaks were small and misplaced as they are on lower
temperature values.

129
 50
Tair,in exp 300 m3/h (°C)
Tair,in sim 300 m3/h (°C)
45
Tair,out exp 300 m3/h (°C)
Tair,out sim 300 m3/h (°C)
40
Temperature [°C]

35

30

25

20
0 5 10 15 20
Time [h]

Figure 4.13: Simulation results for inlet and outlet air temperature using second approach’s heat
capacity curves and comparison with experimental data

50

Tpcm,5 exp 300 m3/h (°C)

45

Tpcm,5 sim 300 m3/h (°C)

40
Temperature [°C]

35

30

25

20
0 5 10 15 20
Time [h]

Figure 4.14: Simulation results for PCM temperature at node 5 using second approach’s heat capacity
curves and comparison with experimental data

130
4.4.5. Third approach: 2 separate triangles for the discharge phase
Following the remarks of the last approach, a third one was tested, introducing two separate peaks for
the heat capacity values of the discharge phase. As before, the enthalpy change was calculated,
obtaining ΔH,cooling= -221,37 J/g.

40
Cp heating
30
Cp cooling
20

10

0
Cp [J/(g*°C]

-10

-20

-30

-40

-50

-60
0 10 20 30 40 50 60 70 80
-70
Temperature [°C]

Figure 4.15: Fictive Heat capacity values, third approach

Discussion of simulation results using 2 separate triangles for the discharge phase
The results of the second approach are illustrated in figure 4.16 and 4.17. Two separate and distinctive
peaks appear on the PCM temperature curve during the discharging phase, following the peaks observed
on experimental data (figure 4.17). This approach provided satisfying results but an offset was still
observed for the PCM temperature values along the plate (inlet, middle and outlet parts).

131
 50
Tair,in exp 300 m3/h (°C)
Tair,in sim 300 m3/h (°C)
45 Tair,out exp 300 m3/h (°C)
Tair,out sim 300 m3/h (°C)
40
Temperature [°C]

35

30

25

20
0 5 10 15 20
Time [h]

Figure 4.16: Simulation results for inlet and outlet air temperature using third approach’s heat capacity
curves and comparison with experimental data

50

Tpcm,5 exp 300 m3/h (°C)

45
Tpcm,5 sim 300 m3/h (°C)

40
Temperature [°C]

35

30

25

20
0 5 10 15 20
Time [h]

Figure 4.17: Simulation results for PCM temperature at node 5 using third approach’s heat capacity
curves and comparison with experimental data

132
4.4.6 Final approach: heat capacity values for different positions along the PCM layer
The observed offset and the remarks of Annex 17 report led to an investigation of experimental results
by the downstream positioning of the heat exchanger. Detailed analysis of the experimental data
resulted in two observations: heating/cooling rates of the PCM also depend on the stage of
stocking/destocking process as well as the part of the plate that we are considering (inlet, middle, and
outlet).

This can be observed on figure 4.18 where the surface temperature difference at specific points of the
middle plate is illustrated within a one minute interval. More specifically, the temperature difference per
minute is calculated at seven points, located at a 15 cm distance from each other. The data presented
here correspond to a slightly different setup, as the fins (initially installed between the containers) were
totally removed from the heat exchanger in order to install temperature sensors on the surface of the
middle plate.

1
ΔT/min, Inlet + 5 cm [°C/min]
ΔT/min, Inlet + 20 cm [°C/min]
Surface temperature difference per minute

0,9
ΔT/min, Inlet + 35 cm [°C/min]
0,8
ΔT/min, Inlet + 50 cm [°C/min]
0,7 ΔT/min, Inlet + 65 cm [°C/min]
0,6 ΔT/min, Inlet + 80 cm [°C/min]
ΔT/min, Inlet + 95 cm [°C/min]
[°C/min]

0,5
0,4
0,3
0,2
0,1
0
-0,1
0,4 0,5 0,6 0,7 0,8 0,9 1
Time [h]

Figure 4.18: Surface temperature difference per minute at different parts of the plate

Data treatment confirms the intuitive observation that the temperature difference decreases as the
distance from the inlet part increases. The temperature difference also varies with the stage of the
melting/cooling process. These two observations demonstrate that the heating or cooling rates are not
the same alongside the plates and that their exact determination can be quite complex to determine
considering the various heat transfer processes, the phase change, the eventual changes of the airflow
rate, etc. On the contrary, during the DSC measurement heating/cooling rates (ΔT/dt) are kept constant
at all times.

133
This last observation led to the separation of the considered PCM section into three parts (inlet, middle
and outlet) and the use of different heat capacity values for each part. For example, for a discretization
of n=10 nodes, one curve is used for the four first nodes, a second curve for the next three ones and a
final curve for the remaining three (figure 4.19).

Cp cooling inlet part

40 Cp cooling middle part
Cp cooling outlet part
30
Cp heating inlet part
20 Cp heating middle part
Cp heating outlet part
10

0
Cp [J/g°C]

-10

-20

-30

-40

-50
10 20 30 40 50 60 70
-60
Temperature [°C]

Figure 4.19: Calibrated cp values created for the numerical model

The curves were calibrated having experimental data as a reference and the enthalpy change was once
again calculated. The best results were obtained for the values presented in table 4.2.

Heating / Cooling rate

Node 1 to 4 Node 5 to 7 Node 8 to 10

Enthalpy change during

220,58 220,43 220,70
the heating process (J/g)

Enthalpy change during

-223,23 -196,92 -209,96
the cooling process (J/g)
Table 4.2: Enthalpy change for the three part discretization heat capacity curves (heating and cooling)

4.5 Numerical model validation

Results obtained using the last calibrated heat capacity values (figure 4.19) are demonstrated in this
section. The model’s ability to reproduce the heat exchanger’s behavior is demonstrated with a vast
comparison of numerical and experimental results.

134
More specifically, results are presented for three different airflow rates (100, 300 and 500 m 3/h) at
various parts of the heat exchanger:

- Inlet and outlet air temperature (figures 4.20 and 4.21)

- Surface temperature at nodes 1, 5 and 10 length wise (figure 4.22, 4.23 and 4.24)
- PCM temperature at nodes 1, 5 and 10 length wise (figure 4.25, 4.26 and 4.27)

The non-linearity of the temperature evolution during the phase changes is reproduced, including the
two peaks that occur during the solidification stage (linked to the intrinsic properties of the paraffin). The
small deviations observed on the outlet air temperature are due to the difficulty of the model to
integrate the thermal losses of the experimental unit.

Overall, the developed model manages to accurately simulate the heat exchanger‘s behavior as observed
through the experimental platform and is considered validated.

50 Tair,in exp 100 m3/h (°C)

Tair,in sim 100 m3/h (°C)
Tair,in exp 300 m3/h (°C)
Tair,in sim 300 m3/h (°C)
Temperature [°C]

40 Tair,in exp 500 m3/h (°C)

Tair,in sim 500 m3/h (°C)

30

20
0 5 10 15 Time [h] 20 25 30 35

Figure 4.20: Numerical and experimental results for inlet air temperature, for 100, 300 and 500 m3/h

50 Tair,out exp 100 m3/h (°C)

Tair,out sim 100 m3/h (°C)
Tair,out exp 300 m3/h (°C)
Tair,out sim 300 m3/h (°C)
Temperature [°C]

40 Tair,out exp 500 m3/h (°C)

Tair,out sim 500 m3/h (°C)

30

20
0 5 10 15 Time [h] 20 25 30 35

Figure 4.21: Numerical and experimental results for outlet air temperature, for 100, 300 and 500 m3/h

135
 50 Tsurf,1 exp 100 m3/h (°C)
Tsurf,1 sim 100 m3/h (°C)
Tsurf,1 exp 300 m3/h (°C)
Temperature [°C]

Tsurf,1 sim 300 m3/h (°C)

40
Tsurf,1 exp 500 m3/h (°C)
Tsurf,1 sim 500 m3/h (°C)

30

20
0 5 10 15 20 25 30 35
Time [h]

Figure 4.22: Numerical and experimental results for surface temperature at node 1, for 100, 300 and 500 m3/h

50
Tsurf,5 exp 100 m3/h (°C)
Tsurf,5 sim 100 m3/h (°C)
Tsurf,5 exp 300 m3/h (°C)
Temperature [°C]

Tsurf,5 sim 300 m3/h (°C)

40
Tsurf,5 exp 500 m3/h (°C)
Tsurf,5 sim 500 m3/h (°C)

30

20
0 5 10 15 20 25 30 35
Time [h]

Figure 4.23: Numerical and experimental results for surface temperature at node 5, for 100, 300 and 500 m3/h

50 Tsurf,10 exp 100 m3/h (°C)

Tsurf,10 sim 100 m3/h (°C)
Tsurf,10 exp 300 m3/h (°C)
Tsurf,10 sim 300 m3/h (°C)
Temperature [°C]

40 Tsurf,10 exp 500 m3/h (°C)

Tsurf,10 sim 500 m3/h (°C)

30

20
0 5 10 15 20 25 30 35
Time [h]

Figure 4.24: Numerical and experimental results for surface temperature at node 10, for 100, 300 and 500 m3/h

136
 50
Tpcm,1 exp 100 m3/h (°C)
Tpcm,1 sim 100 m3/h (°C)
Tpcm,1 exp 300 m3/h (°C)
Temperature [°C]

40 Tpcm,1 sim 300 m3/h (°C)

Tpcm,1 exp 500 m3/h (°C)
Tpcm,1 sim 500 m3/h (°C)

30

20
0 5 10 15 20 25 30 35
Time [h]

Figure 4.25: Numerical and experimental results for PCM temperature at node 1, for 100, 300 and 500 m3/h

50 Tpcm,5 exp 100 m3/h (°C)

Tpcm,5 sim 100 m3/h (°C)
Tpcm,5 exp 300 m3/h (°C)
Temperature [°C]

Tpcm,5 sim 300 m3/h (°C)

40
Tpcm,5 exp 500 m3/h (°C)
Tpcm,5 sim 500 m3/h (°C)

30

20
0 5 10 15 20 25 30 35
Time [h]

Figure 4.26: Numerical and experimental results for PCM temperature at node 5, for 100, 300 and 500 m3/h

50 Tpcm,10 exp 100 m3/h (°C)

Tpcm,10 sim 100 m3/h (°C)
Tpcm,10 exp 300 m3/h (°C)
Tpcm,10 sim 300 m3/h (°C)
Temperature [°C]

40 Tpcm,10 exp 500 m3/h (°C)

Tpcm,10 sim 500 m3/h (°C)

30

20
0 5 10 15 20 25 30 35
Time [h]

Figure 4.27: Numerical and experimental results for PCM temperature at node 10, for 100, 300 and 500 m3/h

137
4.6 Coupling with existing building model

4.6.1 Building’s model: Hybcell 1.2

Once validated, the numerical model was coupled to an existing building model, the HYBCELL 1.2 (El
Mankibi, 2003). HYBCELL 1.2 was developed at LGCB/ENTPE laboratory, under Matlab/Simulink
environment by coupling a thermal model and a pressure air flow model.

The thermal model can be divided into three parts: the envelope model based on the finite differences
method, the short wavelength radiation model based on an accurate description of the sunspot inside
the simulated building and the long wavelength radiation model using the radiosity method. The air flow
model is a pressure model that takes into account the effect of the wind and the thermal buoyancy of air
to calculate the airflow through the openings of the building.

Indoor air temperature is calculated taking into account several thermal evolution phenomena, such as
heat transfer through the walls, air infiltration and ventilation, internal heat gains and auxiliary heating
or cooling. Schedule and occupation patterns can also be integrated in the model (figure 4.28).

Figure 4.28: Hybcell 1.2 model for building simulation

4.6.2 Definition of the experimental test cell

The characteristics of the experimental test cell were introduced into Hybcell 1.2, including geometry,
orientation, openings and composition of the envelope (figure 4.29). The four walls, floor and ceiling
were introduced, defining their composition and properties, such as reflection, absorption and emissivity
coefficients (figures 4.30 and 4.31, north-east wall). Likewise, the six window openings were introduced
in the north-east wall, entering the exact position, materials, and properties of each one (figure 4.31).

138
Figure 4.29: Visual representation of the modeled experimental cell

Figure 4.30: Composition and properties of Hybcell’s north-east wall

139
 Figure 4.31: North-east wall definition, including openings

4.6.3 Reproduction of an experimentally tested scenario and comparison

The four day test realized at the Hybcell room (section 3.4) was reproduced with the use of the heat
exchanger model coupled to the building model (figure 4.32). The test consisted in a four consecutive
day scenario, during which load shifting was implemented for the three four days and the last day was
taken as a reference (no heating at all).

As before, the heat exchanger model uses, the experimental inlet air temperature and the airflow rate as
input; this time these values vary according to the implemented control strategy during the experimental
test. The outlet air temperature (as calculated from the exchanger’s model) and the airflow rate are then
employed as ventilation inputs for the Hybcell 1.2 model.

Figures 4.33 and 4.34 illustrate the obtained numerical results and a comparison is performed with
experimental ones. The heat exchanger model manages to reproduce the experimentally observed
behavior even for varying inlet air temperature and airflow rate (figure 4.33). Likewise, the building
model after calibration represents with precision the evolution of the indoor air temperature for the
three days of peak load shifting and the fourth reference day (figure 4.34).

140
Controller Heat exchanger
model Data display and
saving

Hybcell 1.2

Figure 4.32: Coupling of the heat exchanger’s model with the building’s model Hybcell 1.2 under Simulink environment

141
 T outlet sim [°C] T outlet exp [°C] T inlet exp [°C] Airflow rate [m3/h]
50
220
200
45
180
160

Airflow rate [m3/h]

Temperature [°C]

40
140

35 120
100

30 80
60
25 40
20
20 0
0:00

6:00

0:00

6:00

0:00

6:00

0:00
12:00

18:00

12:00

18:00

12:00

18:00

12:00

18:00
Time [hour]

Figure 4.33: Numerical and experimental results for the heat exchanger unit and the Hybcell room
coupling during a 4 days test

T Hybcell sim [°C] T Hybcell exp [°C] Airflow rate [m3/h]

40
220
200
35
180
160

Airflow rate [m3/h]

30
Temperature [°C]

140

25 120
100

20 80
60
15 40
20
10 0
6:00
0:00

6:00

0:00

6:00

0:00

0:00
12:00

18:00

12:00

18:00

12:00

18:00

12:00

18:00

Time [hour]

Figure 4.34: Numerical and experimental results for the heat exchanger unit and the Hybcell room
coupling during a 4 days test

142
4.7 Conclusion
A numerical model was developed, based on the formulation of heat balance equations and the
apparent heat capacity method. It considers three heat transferring mediums, air, aluminum surface and
PCM, and is able to reproduce the downstream temperature evolution for each of them. A significant
part of the study focused on the specific heat capacity values through the phase change and the creation
of suitable fictive curves, based on DSC results.

It was validated using experimental data and used to reproduce a preliminary control strategy
(presented in the previous chapter). The model manages to accurately and quickly reproduce the
performance of the heat exchanger unit under different conditions, such as varying airflow rate and inlet
air temperature.

The model will then be used for an optimization study of the heat exchanger and the development of
control strategies when coupled to a building.

143
Chapter 5
Heat exchanger optimization and control strategies
implementation

144
Chapter 5 : Heat exchanger optimization and control strategies implementation
5.1 Introduction
So far, we discussed the French peak power problem, along with its strong ties with electrical space
heating. The advantages of latent heat storage technology were presented, along with the benefits of
HVAC integrated PCM heat exchangers able to provide load shifting solutions. Following this, the
development of a PCM-Air heat exchanger prototype and of a numerical model that simulates its
behavior was described in detail. Nevertheless, the characterization results demonstrated that there
were some limitations concerning the time needed to charge and discharge the proposed unit as well as
the final amount of stored heat.
This chapter presents at first the verification of such reasoning as the exchanger is numerically coupled
to an 80m2 house and tested under different configurations (inlet exchanger air temperature and
discharging period). As the necessity of an optimization study became clear, two approaches were
considered. The first one focused on the modification of the heat exchanger’s dimensions, by altering the
height of the unit, the thickness of the PCM and the air layer, and investigating the impact on energy
consumption and thermal comfort when the unit was coupled to a room. The second one concentrated
on both the dimensions of the heat exchanger and the properties of the PCM and targeted the
configurations with minimum time for stocking and maximum heat storage.

5.2 Conception of a low energy house

5.2.1 Materials and properties of the house
In order to evaluate the performance of the heat exchanger and test it under different operation
scenarios, a building was designed numerically. The Hybcell 1.2 model was once again used for this
purpose (see description in subsection 4.6.2).
The designed building is an 80m2 single zone dwelling having a rectangular shape (10m long, 8m wide
and 3m high). South, east and west walls have two window openings each with a 2.5mx1.5 dimension
(figure 5.1).
The properties of the walls and of the windows were defined in a way that corresponds to a low energy
house, as the heat exchanger is conceived for such a case. The properties of the used materials are
presented in table 5.1.

Density(kg/m3) Specific heat (J/kg.K) Conductivity (W/m.k)

Concrete 2300 880 2.3
Rockwool 100 900 0,036
Plaster 680 960 0,25
Parquet flooring 500 1600 0,13

Table 5.1: Properties of the materials used for the definition of the building

145
 Figure 5.1: The conceived mono zone house, as presented in Hybcell 1.2 model
The composition and the thickness of the different surfaces are presented in table 5.2.

Internal Middle External

Floor Parquet flooring 0,01m Rockwool 0,12m Concrete 0,20m
Ceiling Plaster 0,0125m Rockwool 0,18m Concrete 0,20m
Wall Plaster 0,0125m Concrete 0,20m Rockwool 0,18m

Table 5.2: Composition of the building's surfaces

The U value obtained for each element is as follows: Uceiling=0,188W/m2.K, Ufloor=0,187W/m2.K and
Uwall=0,188W/m2.K, confirming efficient thermal insulation performance (IEA, 2008). Windows are
double glazed, with a 1.2 U value that corresponds to a 4/16/4 type window with an argon blade and a
low emissivity layer on the internal face of the glazing.
5.2.2 Climate definition
Meteorological data were acquired using the IDMP station located on the ENTPE campus (Dumortier,
2003) and correspond to the weather condition in Vaulx-en-Velin (Lyon, France) for the year 2013. Figure
5.2 illustrates the outdoor temperature for January 2013, as this month constitutes the selected testing
period for most of the simulations.

146
 Figure 5.2: Outdoor temperature for the tested month (January 2013)

5.2.3 Heating needs of the house

In order to verify that the building satisfies the low energy consumption criteria (50 KWH/m 2/year), a
simulation was performed for the month of January 2013, using a conventional heating mode (i.e.
without heat exchanger). A PID controller was linked to the heating component of the model in order to
maintain an indoor air temperature of 20°C (figure 5.3). After calibration of several factors, the monthly
heating needs of the building were calculated at 652,23 KWh, or 48,9 KWH/m 2/year; a value lower than
the initial target of 50 KWH/m2/year.
However, the indoor temperature presents peaks for some of the tested days, during the afternoon
hours, linked to horizontal radiation and the fact that the house is well insulated. Furthermore, indoor
temperature reaches lower than the set point values, usually after the afternoon temperature peaks, as
the PID controller encounters difficulties to deal with sharp changes. These peaks and troughs appear in
the simulations performed with the heat exchanger as well.

147
 Figure 5.3: Indoor air temperature using conventional heating, setpoint of 20°C
The building model was then used to determine the heating needs of the house during the peak power
hours (18:00-20:00) over the tested period, using a conventional heating mode. Results are presented in
figure 5.4. Heating needs vary from 1.37 to 2.77 KWh for the period in question, depending on weather
conditions.

Heating consumption without occupants, 18:00-20:00 (kWh)

3

2,5

2
Energy [KWh]

1,5

0,5

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 [d]
Time 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

Figure 5.4: Heating needs during 18:00-20:00 for the conceived house

148
5.3 Testing of the initial exchanger
As presented in the previous subsection, a house was designed in order to test the performance of the
heat exchanger for several scenarios over a one month period. The heat exchanger model was coupled
to the building model Hybcell 1.2 in its initial form (dimension, PCM quantity and properties), as
described in section 4.6. The conventional heating module was disconnected and the house was only
heated using the heat exchanger.
A first control strategy was tested, aiming to investigate whether the exchanger would manage to shift
the electrical heating consumption while maintaining thermal comfort. The heat exchanger was
discharged during the peak power period (18:00-20:00) and charged during the off peak period. A PID
controller regulated the airflow rate according to the thermal needs of the building; an upper limit of
720m3/h was established (3 volumes per hour). The inlet air temperature of the exchanger was initially
set at 38°C during the charging phase, as this value surpasses the PCM’s melting temperature. During the
peak power period, the inlet air temperature was taken equal to the indoor air temperature, meaning
that no heating takes place before the heat exchanger.
For a better demonstration of the obtained results, figure 5.5 illustrates days 3 to 8, as a presentation of
31 days offers limited detail concerning the evolution of the represented parameters.
These five days presented increased heating needs (figure 5.4) and for that reason they will be used for
demonstration purposes for each test from now on. During the charging phase the inlet exchanger air
temperature is 38°C and the controller regulates the airflow rate in order to maintain the house’s air
temperature at 20°C. During the afternoon hours, some peaks are observed in the house’s temperature,
linked as already mentioned to solar radiation. During these peaks, no heating is required and the airflow
rate diminishes or is set to zero. During the discharging phase, the inlet exchanger air temperature is
equal to the house’s air temperature and the controller significantly increases the airflow rate to provide
more power to the house (while destocking the heat stored to the exchanger).

149
Figure 5.5: Performance of the initial exchanger, destocking 18:00-20:00, inlet exchanger temperature
38°C

Towards the end of the discharging phase the airflow rate reaches its maximum value, but even at this
value the indoor air temperature presents a significant drop, almost reaching 18°C.
As the current configuration failed to produce the desired temperature conditions, the inlet exchanger
air temperature was augmented, and tests were performed for 40, 42 and 44°C. Results are presented in
figure 5.6.
Even when the inlet exchanger air temperature was augmented, the use of the heat exchanger failed to
maintain comfortable conditions, the indoor temperature showing values near 18°C.
We were then interested to investigate the performance of the heat exchanger for a longer load shifting
period, first with a two hour increase (17:00-21:00) and then with a four hour increase (16:00-22:00).
Two inlet exchanger air temperature values were tested for each case: 38°C and 44°C. Results are
presented in figure 5.7.

150
 Figure 5.6: Indoor temperature for different inlet exchanger air temperatures

Figure 5.7: Indoor temperature for different load shifting periods, initial exchanger

151
In all four cases a serious degradation of thermal comfort was observed, as indoor temperatures
decrease to values near 15°C.
As a final step of the testing of the initial exchanger performance, the price rate of each case was
calculated, using the values proposed by the French Electricity power company (EDF). In fact, EDF has
established a cheaper night tariff, starting from 22:00 and ending at 06:00, in an effort to motivate
clients to use high energy consuming appliances at night (water heater, washing machine, etc). The night
and day tariffs are presented in table 5.3.

Electricity price rates (EDF, France)

Day rates 06:00-22:00 0,1510 €/KWh

Night rates 22:00-06:00 0,1044 €/KWh

Table 5.3: French electricity price rates during the day and during the night
These tariffs were used to calculate the monthly energy cost for space heating of the house for the cases
presented so far. Results are presented in the following table.

Inlet exchanger Monthly energy Monthly

Shifting period Load Thermal
air temperature consumption electricity bill
(h) shifting comfort
(C°) (KWh) (€)

Conventional
- 652,23 86,10  
heating

18:00-20:00 38 768,26 100,36  

18:00-20:00 40 773,44 101,20  

18:00-20:00 42 777,448 101,91  

18:00-20:00 44 780,77 102,51  

17:00-21:00 38 758,77 97,83  

17:00-21:00 44 771,21 100,04  

16:00-20:00 38 740,85 90,08  

16:00-20:00 44 751,64 91,16  

Table 5.4: Monthly energy consumption and cost for different inlet air temperature values and load
shifting periods for the initial exchanger
Several cases were investigated, altering the inlet exchanger air temperature and the load shifting
period, leading to the conclusion that the initial configuration of the heat exchanger was not efficient

152
enough to guarantee acceptable indoor thermal conditions. Initial assumptions regarding the need for an
optimization study were verified.

5.4 Optimization first approach

5.4.1 Definition of the optimization parameters

The numerical coupling and usage of the heat exchanger confirmed the experimental observations
considering the limited energy storage and the important time needed for this storage to occur. The first
optimization approach of the experimental unit consisted in the usage of the developed model to modify
several parameters of the heat exchanger.
More specifically, the investigated parameters concerned the dimensions of the unit and were identified
as follows:
 The height of the three considered layers, PCM, air and aluminum (dz)
 The thickness of the PCM layer (dym)
 The thickness of the air layer (dya)
The parameters are noted in the figure 5.8, representing the front view of the heat exchanger.

Aluminum layer Air layer PCM layer

dz

dyp dya dym

Figure 5.8: Representation of the modified parameters (PCM and air layers, height of the unit)

Three parameters were introduced in the numerical model as multiplying factors of the dimensions in
question. The values of these parameters were defined as follows:

 The height was multiplied by 1, 1,5 and 2

 The PCM layer thickness was multiplied by 1, 2, 3 and 4

153
  The air layer thickness was multiplied by 1, 2, 3 and 4

5.4.2 Testing protocol

Considering the combination of all the possibilities, a total of 48 configurations occurred. In order to test
these configurations, it was decided to perform the simulations for a 3,5 day period. This period
corresponds to the test presented in subsection 3.4 (4 to 8 February 2013), only this time the building on
which the simulation were tested was the 80m2 low energy one. We decided to run the simulations at
this period for rapidity reasons and because the meteorological data were already acquired and available
in the building model.
For this series of tests we decided to increase the discharging period from two hours to six hours (16:00-
22:00) and set the inlet exchanger air temperature during the charging phase to 38°C. During the
discharging phase the inlet exchanger temperature was taken equal to the building’s indoor
temperature. The airflow rate was regulated using a PID controller, with a maximum value of 720 m 3/h.
Three criteria were established: load shifting realization, minimum monthly consumption and no
degradation of thermal comfort. A decrease of 0,5°C of the house’s air temperature was tolerated.

5.4.3 Selection of three configurations

Results showed that the configuration presenting minimum consumption is the ‘1/4/1’ (75,6 KWh), but
with a significant impact on thermal comfort, as the indoor temperature reaches 18,5°C. After
investigation of all the 48 configurations three of them were selected as they fulfilled the established
criteria. They are presented in the following table:

Description Configuration 1 Configuration 2 Configuration 3

Multiplying factors (dz/dym/dya) 1,5/4/1 1/3/2 1,5/2/2
Final exchanger dimensions (m) 1x1,04x0,27 1x1,56x0,18 1x0,85x0,27
PCM mass (kg) 132 99 66
3,5 days consumption (KWh) 76,05 77,68 81,79
2
Annual consumption (kWh/m ) 48,89 49,94 53,3

Table 5.5: Characteristics of the three selected configurations

All three configurations manage to maintain acceptable indoor conditions (with a 0,5°C temperature
decrease tolerance), as illustrated in figures 5.9, 5.10 and 5.11. The highlighted green areas correspond
to the load shifting period (16:00-22:00). The first configuration uses more PCM, thus having a bigger
volume, but consumes less energy. The second and third ones present decreasing PCM mass and volume
but increasing energy consumption. All three, offer different points of interest and were chosen for the
testing of the coupled system ‘heat exchanger + building’ through different scenarios.

154
Figure 5.9: Indoor air temperature, 3,5 days testing, optimized exchanger 1

Figure 5.10: Indoor air temperature, 3,5 days testing, optimized exchanger 2

155
 Figure 5.11: Indoor air temperature, 3,5 days testing, optimized exchanger 3

5.4.5 First scenario: simple load shifting

In order to evaluate the performance of the heat exchanger, the selected configurations were tested for
a larger time scale. The month of January 2013 was once again selected for this purpose, as it represents
a winter climate and offers a sufficient amount of time for thermal inertia phenomena to take place.
The first scenario was established as before: charging of the heat exchanger between 22:00 and 16:00
and discharging during a six hour period (16:00-22:00). The inlet exchanger air temperature was set at
38°C and the airflow was regulated from a PID controller. Table 5.6 summarizes the operation conditions
of the first scenario.
The aim of this test was to verify the efficient performance of the heat exchanger (load shifting and
thermal comfort) for a longer time period and to calculate the energy consumption and electricity bill at
the end of the month. Figures 5.12, 5.13 and 5.14 present the results of the simulation for the three
configurations. For a better demonstration of the obtained results, these figures zoom on days 3 to 8, as
higher heating power is needed during these days.
For all three configurations, the heat exchanger managed to maintain the indoor air temperature at 20°C
during the shifting and the non-shifting period. As before, the highlighted green areas represent the
shifting period (16:00-22:00).

156
 Time

00:00-16:00 16:00-22:00 22:00-00:00

Inlet exchanger air Indoor air

38°C 38°C
temperature temperature

Airflow rate PID on PID on PID on

exchanger temperature temperature temperature

Inlet house air

Exchanger outlet Exchanger outlet Exchanger outlet
temperature

Airflow rate house Same as exchanger Same as exchanger Same as exchanger

Conventional
No No No
heating
Table 5.6: Simple load shifting operation description, first optimization approach

Figure 5.12: Performance of the heat exchanger, days 3-8, configuration 1, simple load shifting scenario, first
optimization approach

157
Figure 5.13 Performance of the heat exchanger, days 3-8, configuration 2, simple load shifting scenario, first
optimization approach

Figure 5.14: Performance of the heat exchanger, days 3-8, configuration3, simple load shifting scenario, first
optimization approach

158
Energy consumption and the electricity bill corresponding to the three configurations are presented in
table 5.7. The previous observation concerning a lower consumption when more PCM is used is
confirmed. All three configurations consume more energy when compared to conventional heating
appliance.

Monthly energy Monthly

Load Thermal
Configuration consumption electricity bill
shifting comfort
(KWh) (€)

Conventional heating 652,23 86,10  

1 (1,5/4/1) 669,43 82,83  

2 (1/3/2) 687,30 85,20  

3 (1,5/2/2) 720,28 88,79  

Table 5.7: Monthly performance of the three tested configurations, simple load shifting scenario, first
optimization approach

5.4.6 Second scenario: charging of the heat exchanger during the night
A detailed observation of the heating needs of the house indicates that in some cases no heating is
required during the day. Figure 5.15 demonstrates the heating power provided to the house using a
conventional heating system for 3 consecutive days, from January 11 to January 13. Depending on
external weather conditions (mostly solar gains) the power needed decreases or even becomes zero
from 10:00 to 16:00.

Heating power [W]

1200

1000

800
Power [W]

600

400

200

0
12:00

16:00

20:00

12:00

16:00

20:00

12:00

16:00

20:00
0:00

4:00

8:00

0:00

4:00

8:00

0:00

4:00

8:00

0:00

Time [h]

Figure 5.15: Heating needs from 11 to 13 January

159
This observation led us to the creation of a night charging scenario. In this case, the heat exchanger is
charged during the night, starting at 22:00 and finishing at 06:00 in the morning. Until the start of the
load shifting period (16:00) conventional heating is used, if necessary. This means that the heat
exchanger is charged during an 8 hour period and is kept in that state during the day and until the
initiation of the shifting process. Table 5.8 summarizes the operation of the system for the night charging
scenario.

Time

00:00-6:00 06:00-16:00 16:00-22:00 22:00-00:00

Inlet exchanger air Indoor air Indoor air

38°C, 44°C 38°C
temperature temperature temperature

Airflow rate PID on PID on PID on

0
exchanger temperature temperature temperature

Inlet house air

Exchanger outlet Exchanger outlet Exchanger outlet Exchanger outlet
temperature

Airflow rate house Same as exchanger 0 Same as exchanger Same as exchanger

Conventional
No Yes, PID No No
heating
Table 5.8: Night charging scenario operation description, first optimization approach

Figure 5.16 presents the obtained results for configuration 1, with an inlet exchanger air temperature of
38°C. During the discharging phase, the indoor temperature decreases to 17.5 °C, meaning that thermal
comfort is deteriorated and that the current configuration failed to provide efficient control of the
house. Further tests were then performed, raising the inlet exchanger air temperature at 40, 42 and
44°C. Better results were acquired with the last temperature value and are presented in figure 5.17.
Configurations 2 and 3 were also tested for different inlet air exchanger temperatures. Comfortable
indoor air conditions were obtained with 44°C and are presented in graphs 5.18 and 5.19.
The charging of the heat exchanger strictly during the night presents the advantage of moving the load
shifting process to a period with lower energy consumption and a cheaper electricity price.
Monthly energy consumption and electricity bill are calculated for the four presented cases and are
presented in table 5.9. Once again more PCM led to less energy consumption. All three configurations
managed to provide good results, but with an increased inlet exchanger air temperature.
Even though the energy consumption was increased using the heat exchanger, a decrease of the final
cost is observed as the exchanger is charged during the night, when lower tariffs apply. This observation
is valid for the first two configurations as the last one led to an important increase of energy that was
also reflected on the final electricity bill.

160
Figure 5.16: Performance of the heat exchanger, days 3-8, configuration 1, inlet exchanger air temperature=38°C,
night charging scenario, first optimization approach

Figure 5.17: Performance of the heat exchanger, days 3-8, configuration 1, inlet exchanger air temperature=44°C,
night charging scenario, first optimization approach

161
Figure 5.18: Performance of the heat exchanger, days 3-8, configuration 2, inlet exchanger air temperature=44°C,
night charging scenario, first optimization approach

Figure 5.19: Performance of the heat exchanger, days 3-8, configuration 3, inlet exchanger air temperature=44°C,
night charging scenario, first optimization approach

162
 Inlet exchanger Monthly energy Monthly
Load Thermal
Configuration air temperature consumption electricity bill
shifting comfort
(°C) (KWh) (€)

Conventional
- 652,23 86,10  
heating

1 (1,5/4/1) 38 676,29 80,19  

1 (1,5/4/1) 44 686,40 81,15  

2 (1/3/2) 44 686,49 81,17  

3 (1,5/2/2) 44 747,34 87,53  

Table 5.9: Monthly performance of the three tested configurations, night charging scenario, first
optimization approach

5.4.7 Third scenario: occupancy and indoor air quality

[Link] Occupancy pattern definition

Lastly, a third and last scenario was tested, introducing the presence of a four person family to the
building model. The occupancy simulation causes two significant changes compared to the two previous
scenarios: carbon dioxide (CO2) concentration increase and heat release from each occupant. The
occupants are considered to exhale CO2 when present in the building, augmenting the initial non
polluted concentration of 400 parts per million (ppm). Augmented CO2 concentration has been linked to
health problems (drowsiness, headaches, lethargy) and lower productivity levels (Siskos et al., 2001).
Extremely high CO2 concentration during long exposure periods can be lethal. As humans are the main
indoor source of CO2, the concentration of this compound was used in this study as an indicator of the
adequacy of the ventilation of the house.
For the integration of the four person family into the model a weekly occupancy program was
established. The presence of the occupants varies according to the day and time; in general all four
persons are present during the night and mostly absent during the day. The weekly occupancy program
is illustrated in figure 5.20. According to ASHRAE (2013), an indoor CO2 concentration not exceeding
1000 to 1200 ppm is an indicator of a good indoor air quality (IAQ). With respect to that, a value of 1000
ppm was set in this study, as the upper limit of CO2 concentration.
Concerning the heat gains from the human presence, the model considers that each occupant
constitutes a heat source equivalent to 100W.

163
 Number of occupants
4

3,5

3
Occupants [#]

2,5

1,5

0,5

0
12:00
18:00

12:00
18:00

12:00
18:00

12:00
18:00

12:00
18:00

12:00
18:00

12:00
18:00

12:00
18:00
0:00
6:00

0:00
6:00

0:00
6:00

0:00
6:00

0:00
6:00

0:00
6:00

0:00
6:00

0:00
6:00

0:00
Time [h]

Figure 5.20: Weakly occupancy program

[Link] System operation description

The presence of the occupants introduces some changes in the operation of the system, compared to
the first two scenarios. First of all, the indoor air temperature can no longer be taken equal to the
house’s air temperature, as this would mean that indoor polluted air would be recycled in the house. A
more advanced ventilation system is conceived for this scenario. More specifically an energy recovery
ventilation (or double flow ventilation) system is considered to use the heat from stale indoor air before
expelling it from the house to heat the incoming fresh air. In that way, the air used to ventilate the house
is warmer than the outside air and not CO2 polluted. A relationship was used to calculate the
temperature of incoming air, considering a 90% efficient energy recovery ventilation system:
𝑇𝑎𝑖𝑟,𝑖𝑛𝑐𝑜𝑚𝑖𝑛𝑔 = 𝑇𝑎𝑖𝑟,𝑒𝑥𝑡𝑒𝑟𝑖𝑜𝑟 + (𝑇𝑎𝑖𝑟,𝑖𝑛𝑡𝑒𝑟𝑖𝑜𝑟 − 𝑇𝑎𝑖𝑟,𝑒𝑥𝑡𝑒𝑟𝑖𝑜𝑟 ) ∙ 𝜀 equation 5.1

Where ε the efficiency of the heat recovery system; here ε=90%.

Figure 5.21 illustrates the obtained incoming air temperature for an indoor air temperature of 20°C and
18°C and the outside temperature values for the month of January 2013.
Furthermore, a bypass system was also considered, meaning that the incoming air can circulate through
the PCM-Air heat exchanger or be injected directly into the house. This allows the renewal of the indoor
air without affecting the heat exchanger’s state.
All the above considered, the operation of the system is as follows:
- From 00:00 to 06:00 the inlet air exchanger temperature is heated and used to charge the unit
(night charging scenario still active).

164
 - From 06:00 to 16:00 the heat exchanger is bypassed and the double flow ventilation system is
used to ventilate the house. A PID controller ensures that the limit of 1000 ppm is not surpassed.
Conventional heating is used to heat the house during this period, if needed.
- From 16:00 to 22:00 the PCM-Air heat exchanger is used to heat the house; its inlet air is the
outlet double flow ventilation air. A second airflow rate PID controller is introduced for the IAQ
management, along with the one that regulates the indoor air temperature. The maximum value
of these two controllers is finally retained as the airflow rate used in the model.
- From 16:00 to 00:00 the same conditions apply as for the first period (00:00-06:00).
The operation of the system for the occupancy scenario is summarized in table 5.10.
22
T inlet house if Tindoor=20°C [°C]
T inlet house if Tindoor=18°C [°C]
21

20
Temperature [°C]

19

18

17

16

15

14
0 5 10 15 20 25 30
Time [d]

Figure 5.21: Incoming air temperature for an indoor temperature of 18 and 20 °C

Time

00:00-6:00 06:00-16:00 16:00-22:00 22:00-00:00

Inlet exchanger air Indoor air Indoor air

38°C 38°C
temperature temperature temperature

PID on PID on
Airflow rate PID on
temperature and 0 temperature and
exchanger temperature
IAQ IAQ

Inlet house air Exchanger outlet Indoor air Exchanger outlet Exchanger outlet

165
 temperature temperature

Airflow rate house Same as exchanger PID on IAQ Same as exchanger Same as exchanger

Conventional
No Yes, PID No No
heating
Table 5.10: Occupancy scenario operation description, first optimization approach

[Link] Occupancy scenario results

As with the first two scenarios, the three selected configurations were tested for the occupancy scenario.
This time configuration 3 was initially tested with an inlet exchanger temperature of 38°C (figure 5.22).
The proposed control strategy managed to maintain CO2 levels below the 1000 ppm limit but failed to
provide acceptable thermal comfort conditions.

Figure 5.22: Performance of the heat exchanger, days 3-8, configuration 3, inlet exchanger air temperature=38°C,
occupancy scenario, first optimization approach
A second test was then launched, increasing the inlet exchanger air temperature to 44°C (figure 5.23).
This change issued better results as the indoor air temperature was stabilized at 20°C during the load
shifting period. The test was reproduced for configurations 1 and 2; results are presented in figures 5.24
and 5.25.

166
Figure 5.23: Performance of the heat exchanger, days 3-8, configuration 3, inlet exchanger air temperature=44°C,
occupancy scenario, first optimization approach

Figure 5.24: Performance of the heat exchanger, days 3-8, configuration 1, inlet exchanger air temperature=44°C,
occupancy scenario, first optimization approach

167
Figure 5.25: Performance of the heat exchanger, days 3-8, configuration 2, inlet exchanger air temperature=44°C,
occupancy scenario, first optimization approach

The proposed optimized system managed to provide acceptable indoor conditions (thermal comfort and
IAQ) for all three configurations.

The exchanger airflow rate was mostly defined by the IAQ controller but this did not interfere with the
charging of the heat exchanger or the indoor air temperature. Had the need for fresh air supply been
stronger, the predominance of the CO2 controller might have deteriorated the performance of the
system in terms of thermal comfort.

Heating needs were lowered when the occupants were present in the house. This was also confirmed
with the testing of the house using a conventional heating mode (no heat exchanger) and a CO2
controller using the aforementioned double flow ventilation system (469,86 KWh instead of 652,23
KWh). Monthly energy consumption and electricity bill for this case are presented in table 5.11 along
with the tested heat exchanger configurations results. Once again, energy consumption increases but the
electricity bill decreases as the heat exchanger is charged during the night.

Inlet air
Monthly energy Monthly
exchanger Load Thermal Indoor air
Configuration consumption electricity
temperature shifting comfort quality
(KWh) bill (€)
(°C)

Conventional
- 469,86 62,44   
heating

168
 1 (1,5/4/1) 44 490,07 58,14   

2 (1/3/2) 44 491,87 58,40   

3 (1,5/2/2), 38 38 509,93 60,45   

3 (1,5/2/2), 38 44 529,30 62,37   

Table 5.11: Monthly performance of the three tested configurations, occupancy scenario, first
optimization approach

5.4.8 Conclusions of the first approach

The first optimization approach was based on a study of different dimensions of the heat exchanger and
the performance of the unit when coupled to a house, regarding load shifting, thermal comfort and
acceptable IAQ.
We were able to demonstrate the benefits of the proposed optimized unit under different scenarios.
However, the first optimization approach presented some limits:
- The PCM quantity and the volume of the heat exchanger were increased compared to the initial
experimental unit. This increase leads to a higher manufacturing cost of the unit and complicates
its integration into the ventilation system of the building.
- The increased PCM mass also resulted in a longer load shifting period (16:00-22:00 instead of
18:00-20:00).
As a result, a second optimization approach was found necessary in order to overcome these limitations.

5.5 Optimization second approach

5.5.1 Introduction
The limitations of the first optimization approach demonstrated the need for a second optimization
study of the heat exchanger. The second approach was aimed to focus on the performance of the heat
exchanger by itself (i.e. without investigation at a first stage of the impact on the building).

This time, apart from the modification of the dimensions of the unit, the PCM properties were also
altered. Two assessment parameters were identified as optimization criteria:

 The time needed to fully charge the heat exchanger;

 The energy that is stocked to the unit during this period.

The aim was to find the configuration that would charge the sufficient quantity of energy during the
smallest amount of time.

169
 5.5.2 Definition of the modified parameters
The parameters of the heat exchanger that were investigated during the second optimization approach
are those concerning the heat exchanger dimensions and those concerning the PCM properties.

The dimension parameters were modified in a way that does not change the total PCM quantity. For
example, a decrease of the thickness of the PCM layer resulted in a proportionate increase of the height
or the length of the same layer (and of the exchanger respectively).

Concerning the PCM properties, the modified parameters were conductivity, density and specific heat.
Lastly, the container conductivity was also investigated. Different available PCMs were reviewed and a
range of values was proposed for each property. The same process was performed for the container, as
the investigation of the conductivity of different materials (copper, steel, etc) provided a range of
conductivity values. As the PCM properties change values between solid and liquid phases, a coefficient
was used to modify (multiply) the initial paraffin values. Three new values for each property were tested
according to the desired range established.

Concerning the heat exchanger dimension treatment, the following modifications were tested:

- Modification of the PCM layer thickness (dym) while inversely modifying its height (dz)
- Modification of the PCM layer thickness (dym) while inversely modifying its length (dx)
- Modification of the PCM layer thickness (dym) while inversely modifying both its height (dz) and
length (dx)

Table 5.12 summarizes the modified parameters, presenting the initial paraffin values, the range
according to the properties of other PCMs, the three multiplying coefficients and the three new values,
used for the second optimization study.

For each of the three different dimension configurations, the combination of all the possible PCM
properties values led to a total of 243 cases.

Modified parameters
Initial
Parameters Range New New New
values Coef. Coef. Coef.
values values values
PCM
conductivity 0.24 / 0.18 0.18-0.7 2 0.48-0.36 3 0.72-0.54 4 0.96-0.72
(W/(m.K))
PCM
1220- 1600-
density 870 / 780 700-1600 0.9 783-700 1.4 1.84
1090 1435
(kg/m3)
PCM
specific
2.4 / 2 1.4-3.6 0.7 1.7-1.4 1.25 3- 2.5 1.5 3.6-3
heat
(kJ/(kg.K))
Container
conductivity 250 20 -400 0.5 125 1.3 325 1.6 400
(W/(m.K))

170
 PCM layer
thickness 0.09 0.006-0.015 0.66 0.006 1.33 0.012 1.66 0.015
vs length 0.10 0.06-0.15 0.66-1 0.15 1.33-1 0.075 1.66-1 0.06
(m)
PCM layer
thickness 0.09 0.006-0.015 0.66 0.006 1.33 0.012 1.66 0.015
vs height 0.18 0.108-0.272 0.66-1 0.272 1.33-1 0.135 1.66-1 0.108
(m)
PCM layer
0.09 0.006-0.015 0.66 0.006 1.33 0.012 1.66 0.015
thickness vs
0.10 0.077-0.123 0.66-1 /2 0.123 1.33-1 /2 0.086 1.66-1/2 0.077
height and
0.18 0.139-0.221 0.66-1 /2 0.221 1.33-1 /2 0.156 1.66-1/2 0.139
length (m)
Table 5.12: Modified parameters for the second optimization approach

5.5.3 Testing of the modified parameters

The optimization tests were performed in Matlab environment for a full charging and a full discharging
process. For the charging cycle the inlet air temperature was set at 44°C and for the discharging one at
20°C. The PCM temperature before the starting of the test was set at 26°C for the first case and 44°C for
the second one. A criterion had to be established in order to define the end of the charging/discharging
phase. The difference of the outlet exchanger temperature of the heat exchanger was calculated for a
two minute interval. The mean value of this calculation was determined taking into account the last six
values. If this number was inferior to 0,005 °C then the charging or discharging of the unit was
considered finished. The previous statement can be represented as follows in the following relation:

∑𝑡𝑡−𝑛∙𝑑𝑡(𝑇𝑜𝑢𝑡
𝑡 𝑡−2𝑚𝑖𝑛
− 𝑇𝑜𝑢𝑡 ) < 0,005 => 𝑒𝑛𝑑 𝑠𝑖𝑚𝑢𝑙𝑎𝑡𝑖𝑜𝑛 equation 5.2

The stocked or destocked energy until the end of the simulation was calculated as follows:

𝐸 = ∑𝑡=𝑒𝑛𝑑
𝑡=0 [ṁ ∙ Cpa ∙ (𝑇𝑎𝑖𝑟,𝑖𝑛𝑙𝑒𝑡 − 𝑇𝑎𝑖𝑟,𝑜𝑢𝑡𝑙𝑒𝑡 ) ∙ 𝑑𝑡/ 600] equation 5.3

Where

E Stocked or destocked energy [KWh]

Cpa ∶ Specific heat capacity of the air [J/(kg·K)]
ṁ ∶ Airflow rate [kg/s]
Tair,inlet ∶ Inlet exchanger air temperature [°C]
Tair,outlet ∶ Outlet exchanger air temperature [°C]
dt Time step [s]

Results were classified according to the minimum time for energy storage or release. Simulations were
performed for the stocking process and the destocking one; they were repeated for three different
airflow rates: 100, 300 and 500 m3/h. Lastly, simulations were also carried out for the initial dimension of

171
 the heat exchanger, while modifying the values of the PCM properties. All configurations combined a
total of 4860 simulations was performed.

A drawback appeared in the 100 m3/h tests, as the slow heat exchange rates led to the ending of the
simulation before the full charge or discharge of the exchanger for some of the tests. Consequently,
these tests were not considered reliable and were not taken into account.

A Matlab script was used to automatize this procedure and classify the configurations according to the
selected criteria. Part of the results is demonstrated in table 5.13. These results correspond to the first
seventy configurations for charging the unit with an airflow rate of 300m3/h. Column 1 states the
classification of the configurations according to the time needed for the charging to take place, columns
2 to 5 illustrate the multiplying coefficients for the PCM and container properties, columns 6 to 8 present
the tested dimensions; the last two columns present the selected criteria for the second optimization
approach: the time needed for charging and the stocked energy during that time. Some of these
configurations were used for the coupling with the building and are investigated in the following parts of
this chapter.

More results are presented in appendix A.

PCM
PCM PCM Container Thickness Length Height Time Energy
Classification Specific
Conductivity Density conductivity (m) (m) (m) (h) (KWh)
heat
1 4 0,9 0,7 0,5 0,006 0,1231 0,2216 3,35 2721,2
2 3 0,9 0,7 0,5 0,006 0,1231 0,2216 3,37 2725,62
3 2 0,9 0,7 0,5 0,006 0,1231 0,2216 3,4 2729,83
4 4 0,9 0,7 1,3 0,006 0,1231 0,2216 3,5 2751,41
5 3 0,9 0,7 1,3 0,006 0,1231 0,2216 3,52 2753,57
6 4 0,9 0,7 1,6 0,006 0,1231 0,2216 3,54 2758,54
7 2 0,9 0,7 1,3 0,006 0,1231 0,2216 3,55 2756,05
8 3 0,9 0,7 1,6 0,006 0,1231 0,2216 3,56 2760,62
9 2 0,9 0,7 1,6 0,006 0,1231 0,2216 3,59 2762,72
10 4 0,9 0,7 0,5 0,012 0,0868 0,1561 3,94 2387,77
11 3 0,9 0,7 0,5 0,012 0,0868 0,1561 4,01 2394,55
12 4 0,9 0,7 1,3 0,012 0,0868 0,1561 4,03 2397
13 4 0,9 0,7 1,6 0,012 0,0868 0,1561 4,05 2398,52
14 3 0,9 0,7 1,3 0,012 0,0868 0,1561 4,09 2402,43
15 3 0,9 0,7 1,6 0,012 0,0868 0,1561 4,12 2404,6
16 2 0,9 0,7 0,5 0,012 0,0868 0,1561 4,14 2400,49
17 2 0,9 0,7 1,3 0,012 0,0868 0,1561 4,22 2406,53
18 2 0,9 0,7 1,6 0,012 0,0868 0,1561 4,24 2408,48
19 4 0,9 0,7 0,5 0,015 0,0777 0,1398 4,27 2311,24
20 4 0,9 0,7 1,3 0,015 0,0777 0,1398 4,34 2316,97
21 4 0,9 0,7 1,6 0,015 0,0777 0,1398 4,35 2317,88

172
22 3 0,9 0,7 0,5 0,015 0,0777 0,1398 4,37 2321,58
23 3 0,9 0,7 1,3 0,015 0,0777 0,1398 4,43 2325,74
24 3 0,9 0,7 1,6 0,015 0,0777 0,1398 4,45 2326,66
25 2 0,9 0,7 0,5 0,015 0,0777 0,1398 4,57 2329,98
26 2 0,9 0,7 1,3 0,015 0,0777 0,1398 4,62 2332,91
27 2 0,9 0,7 1,6 0,015 0,0777 0,1398 4,64 2333,55
28 4 1,4 0,7 0,5 0,006 0,1231 0,2216 4,86 4050,71
29 3 1,4 0,7 0,5 0,006 0,1231 0,2216 4,88 4056,92
30 2 1,4 0,7 0,5 0,006 0,1231 0,2216 4,94 4064,84
31 4 1,4 0,7 1,3 0,006 0,1231 0,2216 5,04 4087,51
32 3 1,4 0,7 1,3 0,006 0,1231 0,2216 5,06 4091,68
33 4 1,4 0,7 1,6 0,006 0,1231 0,2216 5,09 4096,06
34 2 1,4 0,7 1,3 0,006 0,1231 0,2216 5,11 4095
35 3 1,4 0,7 1,6 0,006 0,1231 0,2216 5,12 4100,17
36 2 1,4 0,7 1,6 0,006 0,1231 0,2216 5,16 4102,46
37 4 0,9 1,25 0,5 0,006 0,1231 0,2216 5,47 4598,93
38 3 0,9 1,25 0,5 0,006 0,1231 0,2216 5,5 4606,99
39 2 0,9 1,25 0,5 0,006 0,1231 0,2216 5,56 4615,19
40 4 0,9 1,25 1,3 0,006 0,1231 0,2216 5,66 4638,49
41 3 0,9 1,25 1,3 0,006 0,1231 0,2216 5,69 4642,1
42 4 0,9 1,25 1,6 0,006 0,1231 0,2216 5,72 4647,35
43 2 0,9 1,25 1,3 0,006 0,1231 0,2216 5,74 4645,71
44 3 0,9 1,25 1,6 0,006 0,1231 0,2216 5,74 4650,97
45 4 1,4 0,7 0,5 0,012 0,0868 0,1561 5,74 3607,86
46 2 0,9 1,25 1,6 0,006 0,1231 0,2216 5,8 4654,41
47 3 1,4 0,7 0,5 0,012 0,0868 0,1561 5,84 3617,59
48 4 1,4 0,7 1,3 0,012 0,0868 0,1561 5,85 3618,16
49 4 1,4 0,7 1,6 0,012 0,0868 0,1561 5,88 3620,31
50 3 1,4 0,7 1,3 0,012 0,0868 0,1561 5,94 3626,07
51 3 1,4 0,7 1,6 0,012 0,0868 0,1561 5,97 3627,56
52 2 1,4 0,7 0,5 0,012 0,0868 0,1561 6,04 3626,19
53 2 1,4 0,7 1,3 0,012 0,0868 0,1561 6,12 3631,33
54 4 1,84 0,7 0,5 0,006 0,1231 0,2216 6,15 5213,87
55 2 1,4 0,7 1,6 0,012 0,0868 0,1561 6,15 3632,9
56 3 1,84 0,7 0,5 0,006 0,1231 0,2216 6,19 5222,88
57 4 1,4 0,7 0,5 0,015 0,0777 0,1398 6,22 3502,55
58 2 1,84 0,7 0,5 0,006 0,1231 0,2216 6,25 5232,21
59 4 1,4 0,7 1,3 0,015 0,0777 0,1398 6,29 3508,13
60 4 1,4 0,7 1,6 0,015 0,0777 0,1398 6,31 3509,24
61 4 1,84 0,7 1,3 0,006 0,1231 0,2216 6,35 5254,49
62 3 1,4 0,7 0,5 0,015 0,0777 0,1398 6,36 3516,69

173
 63 3 1,84 0,7 1,3 0,006 0,1231 0,2216 6,38 5259,19
64 4 0,9 1,5 0,5 0,006 0,1231 0,2216 6,4 5447
65 4 1,84 0,7 1,6 0,006 0,1231 0,2216 6,41 5264,21
66 3 1,4 0,7 1,3 0,015 0,0777 0,1398 6,43 3520,35
67 3 1,84 0,7 1,6 0,006 0,1231 0,2216 6,44 5268,14
68 2 1,84 0,7 1,3 0,006 0,1231 0,2216 6,44 5263,56
69 3 0,9 1,5 0,5 0,006 0,1231 0,2216 6,44 5456,69
70 3 1,4 0,7 1,6 0,015 0,0777 0,1398 6,45 3521,21
Table 5.13: Classification of the obtained results for the charging of the unit according to faster
charging time, airflow rate of 300m3/h

5.5.4 Results of the second optimization approach

The great number of simulations and the observation of the classification of the configurations led to the
verification of our initial thoughts concerning the impact of the investigated parameters and the
formulation of a pattern concerning the configurations with the minimum time needed for
charging/discharging.

These observations can be summarized as follows:

- A thinner PCM layer led to a faster charging of the unit.

- Smaller values of the PCM specific heat and density values led to faster charging but with less
stored energy.
- Larger values of the PCM conductivity led to faster charging, even though the differences were
not very important.
- Oddly, the opposite effect is observed for the container conductivity values, as the smaller the
value of this property, the shorter the time needed to charge. Nevertheless, the impact of the
change of this value is very small, linked probably to the limited thickness of the layer (2mm).
- Concerning the comparison between the configurations involving a thinner PCM layer with
increased length versus increased height, no clear conclusions could be drawn. The impact of the
modification of the height or the length of the exchanger was similar for the two cases.

These observations are reflected in the classification of the tested configurations (table 5.13).
Configurations positioned at the first places were those with smaller specific heat and density values,
thinner PCM layer and larger conductivity values. A similar pattern was detected for all different testing
conditions (charging/discharging, different airflow rates), even though small variations existed.

The following table presents the time needed for a full charge of the initial version of the heat exchanger
and compares this value with optimized versions of the unit for a similar amount of energy storage.

174
 Configuration Air flow rate (m3/h) Energy stored (KWh) Time (h)

Initial 300 3,872 5,97

Case 28 (dymvs dx vs dz) 300 4,050 4,86
Case 28 (dymvs dz) 300 4,055 4,91
Case 28 (dymvs dx) 300 4,045 4,80
Initial 500 3,677 4,34
Case 27 (dymvs dx vs dz) 500 3,804 3,34
Case 25 (dymvs dz) 500 3,798 3,29
Case 22 (dymvs dx) 500 3,791 3,23
Table 5.14: Comparison between initial and optimized version of the heat exchanger

The comparison between the initial and the optimized versions of the heat exchanger shows a quicker
charge of the unit for the same (or even slightly increased) amount of heat storage. The optimized
versions present a one hour decrease for the required charging time.

Furthermore, the values presented in this table verify that the way that is used to compensate for the
modifications of the PCM layer thickness is indifferent, as whether it concerns the height, the length or
both, results are very similar.

The results of the second optimization study provided us with a large list of configurations, each
presenting advantages and disadvantages for our case study. The scenarios that were tested with the
results of the first optimization approach were then reproduced using the newly obtained
configurations.

5.5.3 Simple scenario, second approach

The simple scenario of the first optimization approach involved the testing of three selected optimized
versions of the heat exchanger, without occupants, in order to shift the electrical heating load during the
period 16:00-22:00 for the month of January 2013.

One of the objectives of the second approach was to reduce the load shifting period in order to better
comply with the peak power problem observed in France. The same scenario was therefore reproduced,
modifying nevertheless the load shifting period to 18:00-20:00 and testing the configurations obtained
with the second approach.

Observation of the second study results showed that the same or similar configurations were classified in
the same order for the charging and discharging process and for different airflow rates. Following this,

175
configurations that correspond to the 300 m3/h case (table 5.13) will be used for the assessment of
testing of the previously established scenarios.

Table 5.15 is a reminder of the operation conditions of the simple load shifting scenario.

Time

00:00-6:00 06:00-18:00 18:00-20:00 20:00-22:00 22:00-00:00

Inlet exchanger Indoor air

38°C 38°C 38°C 38°C
air temperature temperature

Airflow rate PID on PID on PID on PID on PID on

exchanger temperature temperature temperature temperature temperature

Inlet house air Exchanger Indoor air Exchanger Exchanger Exchanger

temperature outlet temperature outlet outlet outlet

Airflow rate Same as Same as Same as Same as

PID on CO2
house exchanger exchanger exchanger exchanger

Conventional
No No No No No
heating

Table 5.15: Simple load shifting operation description, second optimization approach

The first tested configuration was the one that presented the quickest charge (case 1, table 5.13). This
configuration was tested for various values of the inlet exchanger air temperature. Figure 5.26 illustrates
the performance of the heat exchanger for this configuration with the last tested inlet air value, 44°C.
The highlighted green areas represent the shifting period (18:00-20:00).
The configuration failed to provide satisfying results, as the indoor air temperature decreased during the
load shifting period. The airflow rate significantly increases in an effort to provide more power but the
heat exchanger cannot provide the needed power.

A second test was performed, using the 28th configuration (table 5.13 again) as it is the configuration
presenting a significant increase in the amount of energy stored in the heat exchanger (compared to the
previous ones). This configuration was also tested for increasing inlet exchanger air temperature but
none of them managed to maintain a stable indoor air temperature during the load shifting period.
Figure 5.27 presents the results obtained for this case with an inlet air temperature of 44°C.

176
Figure 5.26: Performance of the heat exchanger, days 3-8, case 1, inlet exchanger air temperature=44°C, simple
load shifting scenario, second optimization approach

Figure 5.27: Performance of the heat exchanger, days 3-8, case 28, inlet exchanger air temperature=44°C, simple
load shifting scenario, second optimization approach

177
After the inadequacy of the first two cases, a third configuration was tested. The configuration that was
tested was the configuration 38 which corresponds to the configuration presenting the same time for a
full charging as with the initial version of the exchanger for an airflow rate of 300m 3/h. The augmented
stored energy of this configuration permitted an efficient performance of the system, as can be observed
in figure 5.28.

Figure 5.28: Performance of the heat exchanger, days 3-8, case 38, inlet exchanger air temperature=38°C, simple
load shifting scenario, second optimization approach
We then wanted to test the performance of this configuration with a more difficult scenario. As
described in the occupancy scenario of the first optimization study, in case of human presence indoor air
can no longer be reinjected into the house, as it is considered to be polluted. In order to overcome this
obstacle a heat recovery system was used, providing air at a lower temperature than the indoor one
(equation 5.1, figure 5.21).
This time we tested this option but with a more challenging condition: the inlet exchanger air
temperature was given the value of indoor air temperature minus 4°C, during the load shifting period,
exceeding the lower values obtained by the heat recovery system.
The testing of the 38th configuration under such conditions showed that an increase of the inlet
exchanger air temperature during the off peak period was needed in order to obtain stable thermal
comfort conditions. Figure 5.29 presents the performance of the 38th configuration with a 44°C air
temperature during the charging of the unit. A minor decrease of the indoor air temperature is observed
during the load shifting period. Lastly, a configuration with even more increased energy storage was
tested (case 64, table 5.13). This configuration managed to provide an efficient operation of the system

178
with a 38°C inlet exchanger air temperature during the charging process and an indoor minus 4°C during
the discharging one (figure 5.30).

Figure 5.29: Performance of the heat exchanger, days 3-8, case 38, inlet exchanger air temperature=44°C, inlet
house air temperature= outlet exchanger –4°C, simple load shifting scenario, second optimization approach

Figure 5.30: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air temperature=38°C, inlet
house air temperature= outlet exchanger –4°C, simple load shifting scenario, second optimization approach

179
 As with the first optimization approach, the monthly energy consumption and electricity bill are
calculated for each of the presented cases; they are demonstrated in table 5.16.

Monthly
Inlet air Discharge Inlet air Monthly
energy Load Thermal
Configuration exchanger exchanger electricity
consumption shifting comfort
temperature (°C) temperature (°C) bill (€)
(KWh)

Conventional
- - 652,23 86,10  
heating

1 44 Indoor 832,06 110,27  

29 44 Indoor 814,98 107,03  

38 38 Indoor 820,79 106,19  

38 44 Indoor -4 858,71 111,22  

64 38 Indoor -4 829,19 108,39  

Table 5.16: Monthly performance of the tested configurations, simple load shifting scenario, second
optimization approach

5.5.4 Night charging scenario

The tests presented in the previous subsection provided us with two configurations able to satisfy the
criteria of load shifting and acceptable indoor air conditions. These two configurations were used to
reproduce the night charging scenario and the occupancy scenario, as presented in the first optimization
approach.
Concerning the night charging scenario, two modifications were made in the operation of the system:
the load shifting period was reduced to two hours (18:00-20:00) and conventional heating was used
between 20:00 and 22:00. Table 5.17 summarizes the operation of the system for the configurations of
the second optimization approach.

Time

00:00-6:00 06:00-18:00 18:00-20:00 20:00-22:00 22:00-00:00

Inlet exchanger Indoor air Indoor air Indoor air

38°C 38°C
air temperature temperature temperature temperature

Airflow rate PID on PID on PID on

0 0
exchanger temperature temperature temperature

Inlet house air Exchanger Indoor air Exchanger Exchanger Exchanger

180
 temperature outlet temperature outlet outlet outlet

Airflow rate Same as Same as Same as

0 0
house exchanger exchanger exchanger

Conventional
No Yes No Yes No
heating
Table 5.17: Night charging operation description, second optimization approach

Case 38 was initially tested with an inlet exchanger air temperature of 44°C during the charging of the
exchanger. When the unit was not charged, a heat recovery system was considered to provide the inlet
exchanger air at a temperature according to equation 5.1. Results are presented in figure 5.31. The
decreased PCM quantity and the longer conservation period (from the end of charging until the
beginning of discharging) led to an important involuntary discharge of the unit. As a result, indoor air
temperature decreased during the load shifting period.

In view of this fact, we decided to reduce the heat losses of the heat exchanger to the surrounding
environment by a factor of 30%. In fact the heat exchanger model includes a heat loss factor that
depends on the exterior structure and surface of the unit. Even though special caution was given for an
efficient insulation of the heat exchanger, the experimental nature of its conception (openings for
temperature sensors, coupling with the experimental platform etc) resulted in non-negligible heat
losses. These losses seem to play an important role when the exchanger must be maintained at a
charged state for a long period as they lead to involuntary discharge of the unit.

The last test was then repeated for a better insulated version of the heat exchanger. Figure 5.32
illustrates the performance of the unit for the night charging scenario under such conditions. The outlet
exchanger air temperature decreases at a lower rate during the conservation period and the indoor
temperature presents smaller variations.

Furthermore, the 64th case was tested under the same scenario and conditions. As for the 34th case, the
performance of the unit with the initial insulation was not sufficient for a stable indoor air temperature
(figure 5.33). The modification of this parameter demonstrated a better and adequate performance, as
can be seen in figure 5.34.

181
Figure 5.31: Performance of the heat exchanger, days 3-8, case 38, inlet exchanger air temperature=44°C, inlet
house air temperature= outlet exchanger –4°C, simple load shifting scenario, second optimization approach

Figure 5.32: Performance of the heat exchanger, days 3-8, case 38, inlet exchanger air temperature=44°C,
ameliorated insulation, night charging scenario, second optimization approach

182
Figure 5.33: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air temperature=38°C, initial
insulation, night charging scenario, second optimization approach

Figure 5.34: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air temperature=38°C,
ameliorated insulation, night charging scenario, second optimization approach

183
Table 5.18 illustrates the energy consumption and subsequent electricity bill over the tested period. As
expected, configurations presenting better insulation led to lower consumption and prices.

Inlet air
Monthly energy Monthly
exchanger Load Thermal
Configuration Insulation consumption electricity
temperature shifting comfort
(KWh) bill (€)
(°C)

Conventional
- - 652,23 86,10  
heating

38 44 Initial 821,10 101,16  

38 44 Ameliorated 747,88 93,09  

64 38 Initial 793,32 98,27  ~

64 38 Ameliorated 731,71 91,51  

Table 5.18: Monthly performance of the tested configurations, night charging scenario, second
optimization approach

5.5.5 Occupancy scenario

The occupancy scenario was reproduced in order to test the performance of configurations 34 and 64.
The occupants presence was simulated according to the weekly program as presented in figure 5.20. The
system is considered to operate similarly to the first optimization approach (subsection [Link]), with the
difference of a shorter load shifting period. Table 5.19 summarizes this operation. As with the first
optimization approach, the presence of occupants necessitates a heat recovery system and a heat
exchanger bypass unit; it also involves heat gains from each person.

Time

00:00-6:00 06:00-18:00 18:00-20:00 20:00-22:00 22:00-00:00

Inlet exchanger Indoor air Indoor air Indoor air

38°C 38°C
air temperature temperature temperature temperature

Airflow rate PID on PID on PID on

0 0
exchanger temperature temperature temperature

Inlet house air Exchanger Indoor air Exchanger Indoor air Exchanger
temperature outlet temperature outlet temperature outlet

184
 Airflow rate Same as Same as Same as
PID on CO2 PID on CO2
house exchanger exchanger exchanger

Conventional
No Yes No Yes No
heating
Table 5.19: Occupancy scenario operation description, second optimization approach

Configuration 38 was tested for initial (figure 5.35) and ameliorated (figure 5.36) thermal insulation
conditions. Both presented an adequate performance, even though the former led to a small indoor
temperature decrease at the end of the load shifting period.

Figure 5.35: Performance of the heat exchanger, days 3-8, case 38, inlet exchanger air temperature=44°C, initial
insulation, occupancy scenario, second optimization approach

185
 Figure 5.36: Performance of the heat exchanger, days 3-8, case 38, inlet exchanger air temperature=44°C,
ameliorated insulation, occupancy scenario, second optimization approach

Lastly, configuration 64 was also tested for the occupancy scenario, under different conditions. Results
are presented I in the following figures:
- Figure 5.37, inlet exchanger air temperature during charging period 38°C and initial insulation
- Figure 5.38, inlet exchanger air temperature during charging period 38°C and ameliorated
insulation
- Figure 5.39, inlet exchanger air temperature during charging period 40°C and initial insulation
- Figure 5.40, inlet exchanger air temperature during charging period 40°C and ameliorated
insulation
Apart from the first case, the tested configurations managed to maintain a stable indoor temperature
during the discharge of the heat exchanger.
Table 5.20 presents the monthly electricity consumption and price for the tested period. Lower inlet
exchanger air temperature values and better insulation led to lower consumption (and a lower electricity
bill) while fulfilling the goals of the operation of the system.

186
Figure 5.37: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air temperature=38°C, initial
insulation, occupancy scenario, second optimization approach

Figure 5.38: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air temperature=38°C,
ameliorated insulation, occupancy scenario, second optimization approach

187
Figure 5.39: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air temperature=40°C, initial
insulation, occupancy scenario, second optimization approach

Figure 5.40: Performance of the heat exchanger, days 3-8, case 64, inlet exchanger air temperature=40°C,
ameliorated insulation, occupancy scenario, second optimization approach

188
 Inlet air Monthly
Monthly Indoor
exchanger energy Load Thermal
Configuration Insulation electricity air
temperature consumption shifting comfort
bill (€) quality
(°C) (KWh)

Conventional -
469,86 62,44   
heating

38 44 Initial 601,79 74,05  ~ 

38 44 Ameliorated 539,23 67,10   

64 38 Initial 571,31 71,00  ~ 

64 38 Ameliorated 514,72 64,88   

64 40 Initial 578,80 71,69  ~ 

64 40 Ameliorated 527,28 66,12   

Table 5.20: Monthly performance of the tested configurations, occupancy scenario, second
optimization approach

5.6 Conclusions of the second approach

The second optimization approach permitted us to overcome the limitations of the first one and to
demonstrate the efficient operation of the system for the selected scenarios. A smaller and lighter
exchanger was able to provide good indoor thermal and air quality conditions.

Even though the second approach’s optimized configurations are manually defined (in fact, we didn’t use
metaheuristic methods such as genetic algorithms), their performances are sufficient. Indeed, as an
example configuration 64 and 38 meet all indoor air conditions criteria while providing necessary load
shifting with up to 10% energy consumption compared to the initial optimization approach. It is also
important to stress the fact that those performances are not airflow rate and inlet temperature
dependent.

In addition to the demonstration of the performances and capacities of the optimized configurations, the
performed simulations also lead to several conclusions related to PCM thickness, PCM thermal
proprieties and exchanger geometry that can be capitalized on when establishing design
recommendations.

189
Conclusive summary and future work

190
Conclusive summary and future work

Synopsis
The research performed in this PhD dissertation was mainly motivated by the peak power problem
observed in France during the late afternoon winter period. As these peaks are highly related to
electrical heating, the objective of our work was initially to conceive and develop reliable numerical and
experimental tools based on thermal storage technology and secondly to use these tools in order to
develop efficient control strategies based on load shifting solutions.

The experimental part of the study consisted in the design and construction of a PCM-Air heat exchanger
as the main element of the experimental platform and the place where thermal storage occurs in latent
form. The heat storage system is composed of a set of PCM (paraffin) plates, embedded in the exchanger
and integrated in a ventilation system. The initially observed drawback of PCM leakage was overcome
with the construction of a second exchanger possessing a more rigorous structure. The unit was
characterized and coupled to an experimental test cell (Hybcell), in which preliminary strategies were
implemented, focusing on load shifting and maintenance of comfortable indoor conditions. The
characterization results as well as the Hybcell tests which were carried out demonstrated the
applicability of the proposed system but limitations were revealed, mostly concerning the time needed
for the charging and discharging of the exchanger.

This observation along with the advantages of numerical studies (less time and money consuming) led to
the development of a bidimensional numerical model with the purpose of reproducing the behavior of
the heat exchanger. The model uses several methods and techniques, including the heat balance
approach, the apparent heat capacity and the finite difference method in order to successfully represent
the phase change process and the general operation of the system. The importance of the heat capacity
values was revealed and further investigation of this parameter led to the creation of adequate curves
according to the considered position of the PCM plate. Numerical and experimental results were
confronted in order to calibrate and validate the model. Following this, the model was coupled to an
existing building model in order to test the performance of the heat exchanger concerning load shifting,
thermal comfort and indoor air quality through different scenarios.

Two optimization approaches were considered and numerically explored to overcome the limitations of
the initial performance of the heat exchanger. The first one investigated the impact of the modification
of the geometry of the unit when coupled to the building. Three configurations were selected and tested
on a low energy building for a one month period. The benefits of the proposed system were
demonstrated for scenarios presenting increasing difficulty: simple load shifting, night charging of the
unit and lastly operation during the presence of the occupants. Electrical consumption was shifted from
peak to off peak hours without affecting the thermal comfort and the air quality. Even though the final
monthly power consumption was increased, the night charging of the unit allowed the operation of the
system during hours with advantageous tariffs and the final electricity cost was reduced. Nevertheless,
this approach demonstrated some disadvantages concerning the required PCM mass, the volume of the
exchanger and the increased time which was needed for the charging of the unit.

191
This led to the second optimization approach which in addition to the geometry of the heat exchanger
also investigated the PCM properties. Density, specific heat and conductivity values were modified,
seeking the configurations with minimum charging time and maximum energy storage. The optimized
versions of the unit were classified according to charging time and a one hour gain was observed for the
same energy storage. In the end, the scenarios tested with the results of the first optimization approach
were reproduced and we were able to obtain an efficient system with the initial PCM mass which
performs load shifting while maintaining comfortable thermal and air quality conditions.

Perspectives
In this dissertation, the applicability and the benefits of latent heat storage technology were
demonstrated through the numerical and experimental study of a PCM-Air heat exchanger, integrated in
the ventilation system. Several fields can be investigated as a continuity of this work.

First of all, the controllers used for the operation of the system can be upgraded to more advanced ones,
such as those using fuzzy logic. These controllers can offer a more adapted approach to the needs of the
study and present smaller variations and more efficient results. Additionally, predictive tools can be
integrated in the system’s control/command interface. These tools may focus on the weather conditions
or the presence of occupants in order to anticipate the heating needs of the building and adjust the
operation of the system accordingly. Depending on the predicted needs, different energy storage levels
can be decided, leading to a low, moderate or full charge of the heat exchanger. Metaheuristic methods
such as genetic algorithms can also be used in the optimization study providing a different but more
efficient approach.

Furthermore, the efficiency of the proposed system can be tested for different building properties and
structure. The modification of the inertia or thermal insulation of the building may constitute examples
of such a study. A more detailed design of the building with the separation of zones according to the
occupants’ usage and needs can be also studied.

The robustness of the proposed numerical solutions can be tested in an in-situ experimentation, whether
in an experimental house equipped with all the necessary instrumentation or a real life dwelling with
occupants who would understand the operation of the system. The latter case would necessitate a wider
investigation of the impact of the integration of the system in an inhabited environment, focusing on fire
hazards and potential PCM odor diffusion.

The study of the system can be extended to summer conditions in order to take advantage of night
cooling concept technology, either experimentally or numerically. In the first case, the
macroencapsulation choice of containment and the rigid structure of the second heat exchanger would
allow a relatively easy replacement of the current paraffin with one more adapted to summer
conditions (mostly one with a different melting range, maybe a different energy storage capacity). In
the second case, the adaption to summer conditions would be performed through the modification of
the PCM properties in the numerical model. Osterman et al. are currently studying a similar heat
exchanger configuration with a PCM which may satisfy both winter and summer climates. It would be

192
interesting to see whether their results can contribute to this field and verify the efficiency of the
proposed unit with the tools we developed.

Lastly, even though this study focused on the peak power issue, thermal storage technology can be
coupled to other existing innovative technologies such as renewable energy resources, smart grids, living
walls, etc. In that way, the benefits of each domain can be combined and offer wider and more efficient
solutions to the operation of the built environment.

193
194
Texte de synthèse (français)

195
Texte de synthèse (français)

Optimisation numérique et expérimentale de stratégies d’effacement

énergétique

Nikolaos Stathopoulos, 2015

Mots clés:
Bâtiment, Energie, Effacement énergétique, Stratégies de contrôle, Matériaux à changement de phase,
Stockage thermique, Échangeur Thermique, Simulation thermique, Confort thermique, Qualité de l’air
intérieur

Contexte
Le changement climatique dû à l’activité humaine, la raréfaction des ressources fossiles, les chocs
pétroliers, l'augmentation de la population mondiale et urbaine constituent une liste alarmante de
questions qui nécessitent des débats sérieux, de prise de décisions rigoureuses et des solutions
innovantes. Plusieurs accords politiques ont déjà été mis en pratique aux niveaux nationaux et
internationaux et d'autres initiatives sont en cours de préparation pour l'avenir, ce qui reflète une
préoccupation mondiale ainsi qu’une nécessité d'agir croissante.
Prenant en compte ces considérations, le secteur du bâtiment joue un rôle clé, car il constitue à travers
le monde l'un des principaux contributeurs en ce qui concerne la consommation d'énergie et les
émissions de gaz à effet de serre. Compte tenu de ces faits, les politiques européennes et françaises
envisagent d'imposer des réglementations des bâtiments à énergie zéro ou même à énergie positive
dans un avenir proche. Deux problèmes principaux doivent être dépassés en ce qui concerne ces
objectifs: l'intermittence des énergies renouvelables et les variations quotidiennes ou saisonnières de la
consommation d'énergie. Plusieurs recherches ont été effectuées concernant le stockage de l'énergie, ce
qui prouve que ce domaine peut fournir des applications pertinentes en réponse à ces inconvénients.
En général, la France fait face aux défis énoncés ci-dessus, mais elle présente également une particularité
spécifique: des demandes de puissance de pointe élevées. Ces pics sont principalement rencontrés en fin
d'après-midi en période d'hiver et sont fortement liés au chauffage électrique. Un projet a été élaboré
sous l'égide et le financement de l'Agence Nationale de la Recherche, visant à développer un échangeur
de chaleur MCP-Air couplé à une pompe à chaleur, comme une solution technique à la demande
d'électricité augmentée au cours du créneau 18: 00-20: 00. Les partenaires du projet comprennent
quatre unités de recherche (LGCB, CETHIL, ARMINES - CEP et EDF R & D) et un partenaire industriel
(RIBO).

196
Objectifs et approche
C’est dans ce contexte que cette thèse vise à étudier le potentiel d'un échangeur de chaleur MCP-Air et
propose des stratégies d’effacement de consommation électrique sans dégrader le confort thermique et
la qualité de l'air intérieur.
L'idée générale est de développer un échangeur de chaleur intégré dans un système de ventilation, dans
lequel une quantité suffisante de chaleur est stockée afin de remplacer l'utilisation du chauffage
électrique pendant la période de de pointe. Trois objectifs majeurs ont été identifiés:
• Le développement des outils expérimentaux et numériques qui permettront l'élaboration, l'essai et
l'évaluation des stratégies de contrôle;
• L'optimisation de la performance de l'échangeur de chaleur;
• L'élaboration des stratégies d’effacement pour différents scénarios.
L'approche qui a été suivie pour atteindre ces objectifs est un processus rétroactif sur la base de deux
outils: un expérimental et un numérique. Un échangeur de chaleur MCP-Air a été conçu, construit,
instrumenté, caractérisé et couplé à une cellule de test expérimental. En parallèle, un modèle numérique
a été développé en utilisant l'approche de capacité thermique apparente et la méthode des différences
finies. Le modèle a été calé et validé à l’aide des données expérimentales et suite à cela il a été couplé à
un modèle de bâtiment. Une étude d'optimisation a été réalisée se focalisant sur la géométrie de
l'échangeur et les propriétés du MCP pour obtenir des périodes de charge / décharge plus rapides avec
un effet minimal sur l'énergie stockée. Enfin, la version optimisée de l'échangeur a été utilisé pour
l'élaboration de stratégies de contrôle.

Chapitre 1 : Contexte énergétique et bâtiment

Pendant près d'un demi-siècle les questions liées à l'énergie ont acquis un intérêt croissant au sein des
sociétés du monde entier en raison de leur importance économique et environnementale. Cette la prise
de conscience mondiale est reflétée sur un nombre important d’accords signés entre plusieurs pays à
travers le monde, visant à réduire la consommation d'énergie et de gaz à effet de serre (GES). La pierre
angulaire de ces efforts a été le Protocole de Kyoto (Nations Unies, 1998). En ce qui concerne l'Union
Européenne, le «paquet climat et énergie» (Commission européenne, 2008) a été adoptée en Décembre
2008, mettant l'accent sur les réductions d'émissions, augmentation de la production d'énergie
renouvelable et l'efficacité énergétique.
En France, la nouvelle Loi sur la réglementation thermique (RT 2020, actuellement en discussion) va
introduire le concept de bâtiments à énergie positive (BEPOS). Elle visera à diviser la consommation
d'énergie par quatre et obliger tous les nouveaux bâtiments à atteindre les normes BEPOS.
Milieu bâti: un gros consommateur d'énergie
Un des facteurs les plus influençant quand il s’agit du domaine de l'énergie est le Bâtiment: selon
l'Agence Internationale de l'Energie les bâtiments représentent environ 40% de la consommation
d'énergie primaire parmi ses 29 pays membres. En outre, il est estimé (UNEP SBCI, 2009) que les

197
bâtiments contribuent de 33% aux émissions mondiales de GES, principalement en raison de l'utilisation
de combustibles fossiles au cours de la phase opérationnelle. Ces chiffres montrent clairement l'impact
environnemental significatif des bâtiments et le rôle clé qu'ils peuvent jouer dans la réduction de la
consommation d'énergie et les émissions de GES.
La France n’est pas une exception par rapport à la situation européenne et internationale car ce secteur
représente une part importante de la consommation d'énergie et des émissions de GES (ADEME, 2013, a,
b). Plus précisément, le secteur du logement et du tertiaire sont responsable du 44% de la
consommation finale d'énergie, 20% des émissions de GES et 22% des émissions de CO2. L'électricité est
la principale source d'énergie des bâtiments français représentant 37% de la consommation finale, suivis
par le gaz (32%), l'huile (16%), le renouvelable / déchets (15%) et le charbon (0,4%).
En 2012, le secteur résidentiel compte 33,4 millions de maisons, d'où 27,8 millions sont des résidences
principales. L’électricité et le gaz sont les deux principales sources d'énergie du secteur résidentiel. Près
des 2/3 de la consommation d'énergie est utilisée pour les besoins de chauffage, suivie par l'usage
spécifique (PC, de séchage, de la télévision et d'autres appareils) (19,5%), l'eau chaude (12,1 %) et la
cuisson (7%).
Le pourcentage d'utilisation du chauffage électrique dans le logement neuf atteint près de 75% des cas
en 2008. La dernière. En outre, selon certaines estimations Négawatt (2009), le chauffage électrique
atteint une consommation annuelle de 60 TWh, un nombre qui représente 27% de la consommation
totale d'électricité du secteur résidentiel et de 12,3% de la consommation annuelle totale d'électricité en
France.
Pour les 15 dernières années, la consommation électrique en France a présenté des pics de
consommation électrique croissants. Pour la première fois, le seuil symbolique de 100.000 MW a été
dépassé au cours de la vague de froid de Février 2012. En outre, selon RTE (RTE, 2012, a), l'augmentation
de ces pics est plus rapide et plus dynamique que l'augmentation annuelle de la consommation
électrique. Les valeurs maximales sont principalement observées pendant les heures de fin d'après-midi
de la période d'hiver (19:00). Ces pics constituent le résultat de la fin des activités tertiaires, des besoins
accrus de transport public et du démarrage des activités domestiques du soir. Un des facteurs les plus
influents de ce phénomène est le chauffage électrique. Par ailleurs, RTE souligne la forte dépendance de
la consommation électrique sur les valeurs de la température extérieure, en particulier pendant l'hiver.
Le terme thermo-sensibilité est appliqué pour décrire ce phénomène, avec une hausse de 2300 MW de
la consommation par augmentation d’un degré Celsius de la température pendant l'hiver 2012. Cette
valeur présente une croissance de 35% par rapport à celle de 2001/2002. Ce phénomène est beaucoup
plus visible en France que dans les autres pays européens, car il représente près de la moitié de la
thermo-sensibilité totale observée en Europe.
L'ampleur de la demande de puissance en heures de pointe a des conséquences économiques,
environnementales et sociales importantes. Le système de gestion de l'énergie française est confronté à
des difficultés croissantes pour répondre à de tels événements, car il repose en grande partie sur les
centrales nucléaires, présentant des taux de production essentiellement constants de 63,130 MW,
couvrant 77% de la consommation totale d'énergie (RTE, 2014). En conséquence, la France a à faire

198
appel à d'autres moyens de production d'énergie et / ou à s’adresser à des pays voisins pour importer.
Dans les deux cas l'utilisation de combustibles fossiles est généralement impliquée, conduisant à une
augmentation des émissions de CO2.
De plus, la période de pic de consommation électrique représente un pourcentage relativement faible de
la consommation annuelle totale. Néanmoins, des investissements surdimensionnés importants doivent
être réalisés sont pratiquement pas utilisés pour la plus grande partie de l'année. Ces coûts sont
également susceptibles d'être répercutés sur les consommateurs, et plus particulièrement à la partie fixe
de la facture électrique qui correspond au développement des infrastructures. Ce résultat pénalise
surtout les petits consommateurs individuels, c’est-à-dire les ménages précaires. Les factures croissantes
peuvent contribuer à la précarité énergétique des ménages à faible revenu qui ne peuvent pas se
permettre un logement bien isolé ou d'un budget énergétique plus élevé.

Chapitre 2 Stockage d’énergie, MCP et application intégrées dans le système de

ventilation
Comme indiqué précédemment, le bâtiment est l'un des principaux contributeurs en ce qui concerne la
consommation d'énergie et les émissions de GES. En même temps, la population mondiale augmente et
devrait se élever à 9,6 milliards d'ici 2050 (ONU, 2013). Un pourcentage croissant de la population vivra
dans les villes, ce qui indique que les paysages urbains de haute densité seront également augmentés
dans l'avenir. Une des conséquences de ce fait sera l'intensification de l'effet et de l'Îlot de chaleur et de
pollution urbaine qui peut influencer négativement les taux de mortalité et les maladies liées à la chaleur
(Haja et al., 2002). En effet, la vague de chaleur de l'été 2003 a été directement liée à 70,000 décès dans
les pays européens (15.000 décès en France uniquement) (Robine et al., 2008).
Compte tenu de ces faits, la nécessité pour une gestion énergétique du bâtiment plus efficace semble
essentielle, le remplacement des sources d'énergie conventionnelles par des ressources renouvelables
étant une perspective très prometteuse. Néanmoins, l’intégration des ressources renouvelables dans le
bâtiment présente quelques inconvénients, la difficulté principale étant la variabilité et l'intermittence
de leur disponibilité (notamment concernant l'énergie éolienne et solaire).
En outre, comme présentés dans la partie précédente, les variations journalières et saisonnières de la
consommation énergétique conduisent à des pics de consommations inquiétants et ont tendance à
devenir un problème important pour les compagnies d'énergie, la France étant l'un des pays les plus
touchés en Europe.
Dans ces conditions, le stockage d'énergie (energy storage - ES) peut jouer un rôle clé dans les bâtiments
à efficacité énergétique en fournissant de nombreuses solutions concernant l’intermittentes des
ressources renouvelables, tout en conservant un environnement intérieur confortable. En outre, il peut
contribuer à la sécurité de l'approvisionnement en énergie, à l’élaboration des stratégies de contrôle
avancées (associé à des smart grids) et l'augmentation de l'inertie. Le stockage thermique peut être
effectué par divers procédés : thermiques, mécaniques, électromagnétiques, chimiques et biologiques.

199
Certains d'entre eux offrent un stockage d'énergie seulement à court terme et d'autres peuvent fournir
un stockage à long terme.

Stockage thermique
Le stockage d'énergie thermique permet l’utilisation d’une certaine quantité de chaleur à une période
ultérieure avec un décalage horaire, quotidienne ou même saisonnier. Ce stockage peut être atteint
grâce à trois méthodes principales: le stockage de chaleur sensible, stockage de chaleur latente et
stockage de chaleur chimique.
Le stockage de chaleur sensible se produit dans un milieu où le changement de température reste dans
une plage fixe et sans changement de phase. Au contraire le stockage de chaleur sous fore latente se
produit lorsque le support de stockage subit un changement de phase, de l'état solide à l'état liquide
(généralement), ou de liquide à gaz ou de solide à gaz, et vice versa. La chaleur stockée est utilisé pour le
changement de phase, ce qui permet d'obtenir une capacité de stockage beaucoup plus élevé et une
plage de température de fonctionnement plus étroite par rapport aux systèmes de chaleur sensible.
Enfin, le stockage de chaleur chimique est basé sur l'énergie libérée et absorbée au cours de la rupture et
la formation de liaisons moléculaires pour des réactions chimiques complètement réversibles.

Matériaux à changement de phase

Les matériaux à changement de phase (MCP) constituent l'un des principaux éléments des systèmes de
stockage de chaleur latente. Le processus de changement de phase et le stockage de l'énergie latente
subséquente leur permet de stocker des quantités importantes de chaleur dans un petit volume et dans
une plage de température spécifique. Dans la plupart des applications à base de MCP du bâtiment, le
changement de phase solide-liquide est préféré en raison de la faible variation de volume par rapport
aux transformations gaz-liquide ou gaz-solide.
Plusieurs critères existent pour la classification des MCP en fonction de leur composition chimique, leur
température de fusion, leur chaleur latente de fusion, etc. La catégorisation la plus fréquente les divise
en trois groupes: organiques, inorganiques et eutectiques. Un grand nombre de matériaux ont été
utilisées comme MCP et étudié dans des applications de stockage thermique, y compris les sels hydratés,
les paraffine, les acides gras et les eutectiques de composés organiques et non-organiques.
Le confinement des MCP est un sujet très important dans dispositif de stockage thermique car elle est
liée au processus de transfert de chaleur et la rigidité générale de la structure de l'application. Comme
les MCP présentent une faible conductivité thermique, un confinement adéquat peut aider surmonter
cet inconvénient, par exemple en fournissant une surface d'échange élevée par rapport au volume ou en
ayant une faible épaisseur de stockage. Comme indiqué précédemment, des solutions de confinement
mal-adaptés peuvent conduire à la corrosion et / ou des fuites du MCP, en rendant l'équipement non
opérationnel. Plusieurs techniques sont proposées pour le confinement adéquat des MCP, y compris la
macroencapsulation, la microencapsulation, l'incorporation directe et l’immersion.

200
Applications de MCP dans le bâtiment
Plusieurs applications utilisant des MCP ont déjà été étudiées dans le domaine du bâtiment ; elles
peuvent être divisées en deux grandes catégories, les applications passives et les applications passives.

Les systèmes passifs

Dans les systèmes passifs le MCP est incorporé dans des composants spécifiques du bâtiment, visant à
plusieurs objectifs: augmenter l'inertie thermique, minimiser l'utilisation des dispositifs mécaniques de
chauffage / refroidissement, abaisser la fluctuation de la température intérieure et préserver le confort
de l'occupant. Des exemples de telles applications sont les suivants : MCP en plaques de plâtre et
panneaux muraux, MCP en béton, MCP en briques, MCP en plancher et plafond panneaux et MCP en
vitre et voltes d’ombrages.

Les systèmes actifs

Les applications actives de MCP constituent une technologie améliorée pour l’utilisation du stockage
thermique dans les bâtiments. Contrairement aux applications passives, le stockage thermique peut être
activé par un système mécanique supplémentaire (ventilateurs, pompes, etc), en améliorant de cette
façon les performances de transfert de chaleur et permettant le stockage ou le déstockage de chaleur /
froid sur demande. Ces applications peuvent être utilisées pour des intervalles de temps spécifiques et
sont capable de corréler les besoins avec les fournitures d'énergie.
Des exemples de telle technologie peuvent être trouvés dans les systèmes de ventilation ou des
plafonds/planchers radiants.

MCP intégré dans des systèmes de ventilation

Dans ces types de systèmes, la chaleur est stockée dans un dispositif spécifique (échangeur de chaleur),
thermiquement isolé du reste du bâtiment et utilisé uniquement sur demande et non pas
automatiquement.
Les systèmes de stockage de chaleur latente sont généralement composées de quatre éléments de base:
• L'échangeur de chaleur qui contient le MCP macroencapsulé ou microencapsulé; c’est la composante
principale du système et le lieu où la chaleur est stockée sous forme latente (et sensible).
• un circuit qui permet l'écoulement du fluide de transfert de chaleur (principalement l'air, parfois l'eau).
• une source de chaleur électrique ou solaire qui est utilisé pour chauffer ce fluide.
• un ventilateur ou une pompe qui régule le débit du fluide circulant dans le système.
En plus de ces éléments, différents capteurs sont généralement installés dans différentes parties du
système afin de surveiller la température de l'air à l’entrée / sortie, la température du MCP, le flux d'air,
etc.
Les avantages de cette technologie peuvent être résumés comme suit:
 stockage de grandes quantités d'énergie par unité de masse qui permet une variation de
température plus faible (amélioration du confort thermique)

201
  L'utilisation de l'énergie produite à partir du solaire et de la chaleur des déchets qui vient de
procédés exothermiques
 L'utilisation de l'énergie à prix faible pendant les périodes hors pointe pour le stockage et la
restitution de cette énergie à des moments où les pleins tarifs sont appliqués.
 L'utilisation des systèmes de ventilation à taille réduite et le fonctionnement plus efficace

En conclusion, dans cette partie on a voulu présenter les solutions proposées par le stockage d’énergie
en focalisant sur le stockage thermique sous forme latente. La prochaine partie est consacrée à la
description du développement d’un tel outil : un échangeur de chaleur MCP-Air.

Chapitre 3 Elaboration d’un échangeur thermique

En association avec le projet Stock Air un besoin spécifique a été identifié: le développement d’un
dispositif expérimental qui permet de stocker une quantité de chaleur suffisante pour remplacer la
consommation de chauffage électrique en heure de pointe.
Plus précisément, l'application a été conçue pour un logement de faible énergie, de manière qu'il puisse
être facilement intégré dans un système de ventilation et être en mesure de fournir une puissance de 2
kW sur une période de deux heures, ce qui correspond à la fin d'après-midi (18: 00-20: 00) ou le pic de
demande de puissance est observé.
Ces objectifs ont conduit à certains facteurs de conception, parfois contradictoires entre eux :

 Petit volume (pour l'intégration dans le système de ventilation)

 stockage d'énergie suffisant pour satisfaire la nécessité 4kWh
 plage de température de fonctionnement spécifique
 transfert de chaleur élevé pour un processus de stockage / déstockage rapide
 structure rigide pour un confinement de MCP durable
Les considérations ci-dessus ont conduit à la nécessité du développement d'un échangeur de chaleur
MCP-Air. Une analyse a été effectuée par le laboratoire EDF Matériaux et Mécanique des Composants
(partenaire du projet Stock Air), comparant les propriétés de différents matériaux qui pourraient
convenir à l'unité proposée. La paraffine Microtek 37D a été choisie en raison de ses propriétés
avantageuses (tableau 1). En particulier, il présente une plage de température de fusion / solidification
d'environ 37 ° C, une dilatation de 10% entre les phases et une chaleur latente de fusion de 230 kJ / kg.

Conductivité Capacité Plage de

Densité Chaleur latente
Etat thermique spécifique température
(kg.m-3) ([Link]-1)
(W.m-1K-1) ([Link]-1K-1) de fusion (°C)

Solide 870 0.24 2.36 230 33 - 39,6

202
 Liquide 779 0.18 2.45 226.8 35,5 - 30,3

Tableau 1 : Propriétés de la paraffine Microtek 37D

L'échangeur a été conçu de manière à pouvoir être intégré dans un système ventilation (plenum, faux
plafond. A cet égard, sa taille devait être modeste et a été établi comme suit: 1,05m de longueur, 0,80m
de largeur et 0,25m de hauteur.
La paraffine est macro-encapsulé dans 34 règles de maçon en aluminium, (1x0,10x0,018) m de
dimension et ayant des ouvertures aux deux extrémités (figure 1). Les règles ont été sélectionnées à
cause de la valeur élevée de la conductivité thermique de l'aluminium d'une part et leur prix
relativement faible et leurs facilités d’accès, d'autre part. La paraffine a été insérée sous forme liquide et
les plaques ont été scellées à l’aide des morceaux en bois et une colle élastique. Les plaques ont été
assemblées par deux, puis placés parallèlement l'une à l'autre. Des mesures de poids ont été effectuées
avant et après l'encapsulation de la paraffine, de manière à obtenir la masse exacte du MCP. Le résultat
a été un total de 31,44 kg de MCP.

Règles instrumentées

Panneaux Ailettes Règle Planches

d’isolation de maçon en bois

Figure 1 : L’échangeur en phase de construction

Des ailettes en aluminium, 0,019m d'épaisseur, ont été introduites entre les plaques pour augmenter
l'échange thermique. Trois des plaques ont été équipées avec des composants supplémentaires qui ont
été utilisés pour introduire des capteurs de température à des profondeurs variables à l'intérieur du
MCP, en trois parties différentes des plaques: entrée, milieu et sortie. Des capteurs de température ont
également été installés sur la surface extérieure de chaque plaque aux mêmes points. Des panneaux de

203
polystyrène encerclent les règles afin de les isoler thermiquement du milieu environnant. Enfin, des
planches de bois ont été utilisées pour assembler le tout en assurant la stabilité de la structure.
L'utilisation de l'échangeur de chaleur a nécessité la construction d'une plate-forme composé de deux
ventilateurs centrifuges, deux résistances chauffantes (1KW chacun), un diffuseur d'air et un collecteur
d'air, plusieurs tuyaux de ventilation et raccords pour interconnecter les éléments précédents. La plate-
forme expérimental développé assure la circulation et le chauffage de l'air à travers l'échangeur dans des
conditions souhaités (température et débit d'air) et est utilisé pour coupler l'échangeur avec une pièce
ou un bâtiment.
Un système de contrôle commande a été aussi nécessaire. Ce système est composé d'une chaîne de
composants qui sont utilisés pour l'observation des propriétés thermiques, l’acquisition de données et le
contrôle de la température et le débit d’air. Plusieurs capteurs ont été installés dans différentes parties
de la plate-forme et l'échangeur de chaleur, ce qui permet de surveiller l'état de l'air et du MCP. Plus
précisément, des capteurs de température et d'humidité relative ont été installés avant et après
l'échangeur afin de mesurer les propriétés de l'air à l'entrée et la sortie. Le logiciel LabVIEW a été utilisé
pour l'acquisition des données et le contrôle des dispositifs de la plate-forme expérimentale. L'interface
développée permet une surveillance directe des propriétés de l'air et du MCP, l’automatisation des tests
de caractérisation et l’élaboration de stratégies de contrôle.

Couplage avec une cellule expérimentale

La plate-forme expérimentale a été couplé avec une cellule d'essai expérimental appelé Hybcell, situé
dans le campus d'ENTPE à Lyon, France (El Mankibi, 2003). Hybcell fait 5,1 m de long, 3,5 m de large, 2,9
m de haut et représente un bureau ou une petite salle de réunion. L'échangeur de chaleur est placé à
l'extérieur de la cellule Hybcell et a été couplé à son système de ventilation mécanique. Le logiciel
développé pour contrôle de l'échangeur a été fusionné avec le logiciel existant de l’Hybcell pour un
fonctionnement simultané du système. Le couplage a été réalisé afin d'évaluer la performance de
l'échangeur de chaleur dans des conditions réelles et d'élaborer des stratégies de contrôle.

Étude de caractérisation
Plusieurs tests de vérification et de caractérisation ont été réalisées pour différents débits d'air, visant à:
• Confirmer le bon fonctionnement des différents composants de la plate-forme expérimentale
• Confirmer le bon fonctionnement de l'interface d'acquisition-contrôle-commande
• Obtenir un profil du fonctionnement de l'échangeur pour différents débits d'air
• Vérifier l'uniformité du flux d'air entre les plaques
• Vérifier la répétabilité des phases de charge / décharge
• Acquérir des données expérimentales fiables pour la validation du modèle numérique

Fuite de MCP
Malgré les efforts de confinement étanche du MCP, des fuites ont été observées sur les bouchons en
bois. La fuite a probablement été provoquée par la pression qui a été créé au cours du changement de
phase et la dilatation du MCP. La quantité de MCP fuit n’était pas important, mais le problème devait

204
être résolu car ça pourrait conduire à une nouvelle détérioration de l'échangeur et le manque de
connaissance de la quantité de MCP réel dans les plaques.
Compte tenu de ces faits, l'élaboration de nouvelles plaques a été décidée. Les nouvelles plaques ont été
fabriquées par une usine métallurgique à l'aide du procédé d'extrusion. Le matériau utilisé était une fois
de plus l'aluminium. Les plaques ont deux ouvertures pour le remplissage du MCP et l'évacuation de la
pression provoquée par la dilatation entre les phases solide et liquide. Leurs dimensions étaient les
mêmes que les règles de maçon: (1x0,10x0,018) m.

Résultats des tests de caractérisation

Comme indiqué précédemment, l'étude de caractérisation a été réalisée afin d'obtenir un profil complet
du fonctionnement de l'échangeur et d'obtenir les données expérimentales nécessaires qui permettront
de valider le modèle numérique.
Un protocole a été établi et peut être décrit comme suit:
• Test de deux configurations différentes: avec et sans ailettes
• Réalisation d'un cycle thermique complet avec une température d'entrée d'air de 44°C pendant le
stockage et 26°C pendant la phase de déstockage
• Variation de valeurs du débit d'air de 100 à 500 m3 / h
• augmentation du débit d'air sur un pas de 100 m3 / h
Les résultats obtenus pour les tests avec ailettes sont présentés aux figures suivantes.

45

40
Temperature [°C]

35
T out 100 m3/h [°C]
30 T out 200 m3/h [°C]
T out 300 m3/h [°C]
25 T out 400 m3/h [°C]
T out 500 m3/h [°C]
20
0 1 2 3 4 5 6 7 8 9 10 11 12
Time [h]

Figure 2: Résultats de la caractérisation de l’échangeur: température de l’air à la sortie, charge

205
 45
T out 100 m3/h [°C]
T out 200 m3/h [°C]
40
T out 300 m3/h [°C]
T out 400 m3/h [°C]
35 T out 500 m3/h [°C]
Temperature [°C]

30

25

20
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Time [h]

Figure 3: Résultats de la caractérisation de l’échangeur: température de l’air à la sortie, décharge

Discussion des résultats de caractérisation

Les figures 2 et 3 présentent l'évolution de la température d'air à la sortie de l’échangeur pendant la
charge et la décharge du dispositif pour cinq débits différents. Bien que le transfert de chaleur soit
accéléré par la présence d'ailettes, une période de temps importante est toujours nécessaire pour
charger ou décharger complètement l'échangeur. Pour le débit le plus faible l'intervalle nécessaire a été
11 heures pour le stockage et 18 heures pour le déstockage. Cela peut être inquiétant surtout pour le
processus de déstockage, car la barrière de deux heures (au cours de laquelle la demande de puissance
de pointe se produit) est largement dépassée. D'autre part un débit d'air élevé peut détériorer le confort
des occupants. Les taux élevés de flux d'air peuvent aussi provoquer une surchauffe de la résidence au
cours de la phase de stockage.
Les résultats de caractérisation nous ont donné un aperçu du comportement du dispositif pour deux
configurations différentes et des débits d’air variant. En outre, des données expérimentales fiables ont
été obtenues pour le processus de calibrage et la validation du modèle numérique.

Chapitre 4 Elaboration d’un modèle numérique

Les modèles numériques sont des outils puissants qui peuvent aider à surmonter les limites des études
expérimentales (temps et argent). Dans cette optique, le développement d'un modèle qui simule le
comportement de l’échangeur était nécessaire. Le modèle devrait être à la fois rapide et précis dans la
reproduction des résultats expérimentaux et faciliter le couplage avec un modèle existant de bâtiment
pour le développement des stratégies d'optimisation et de contrôle.

206
Ainsi, un modèle numérique bidimensionnel a été créé dans un environnement Matlab / Simulink, en
utilisant la méthode du bilan de chaleur et la méthode de la capacité thermique apparente. Pour la
représentation numérique de l’échangeur, trois principaux médiums ont été choisi: l'air, le conteneur
(aluminium) et le MCP. La moitié de l'épaisseur de la couche d'air expérimental et de l'épaisseur de la
couche de MCP sont prises en compte dans l'étude numérique, étant donné que, par symétrie, les
résultats peuvent être étendus à l'autre moitié (où s’appliquent des conditions similaires).
Chaque couche a ensuite été discrétisé en n régions égales (nœuds) de longueur dx = L / n, où L est la
longueur totale de l'échangeur (figure 4).
De plus, la couche de MCP a été discrétisé transversalement en m nœuds, afin d’offrir la possibilité d'un
traitement plus approfondi du milieu où le changement de phase se produit. Les résistances thermiques
sont formulées entre les nœuds, avec différents coefficients utilisés pour les nœuds d’entrée et de
sortie. La température de l'air à l’entrée de l’échangeur et le débit d'air sont introduits pour chaque pas
de temps.

̇ ∙ ̇ ∙ ̇ ∙
, , ,

1
1 = + 1 1
∙ℎ 2 ∙
1
= + 2 =
∙ℎ 2 ∙ , ∙ , 2 ,

= +
2 ∙ ∙

1
1 = + =
1 ∙ℎ 2 1 ∙ 1 ∙
, , ,

1 =
= + ∙
2
∙ℎ 2 ∙ =
2 2
, 2 ∙ , ,

1
= + =
∙ℎ 2 ∙ ∙
, , ,

Figure 4 : Discrétisation nodale des couches considérées

𝑇 𝑡 +𝑇 𝑡−𝑑𝑡
La température moyenne 𝑇̅ = 2
est introduite et le bilan thermique est formulé pour chaque
couche de la façon suivante:
Couche d'air:
𝑚̇ 𝑑𝑡 ℎ 𝑆 𝑑𝑡 𝑚̇ 𝑑𝑡 ℎ 𝑆 𝑑𝑡
𝑡−𝑑𝑡
𝑇𝑎,𝑖 = (1 + 2 𝜌 + 2𝜌 ) ̅̅̅̅
𝑇𝑎,𝑖 − 2 𝜌 𝑉 ̅̅̅̅̅̅̅
𝑇𝑎,𝑖−1 − 2 𝜌 𝐶𝑝 𝑉 ̅̅̅̅
𝑇𝑝,𝑖
𝑎 𝑉𝑎 𝑎 𝐶𝑝𝑎 𝑉𝑎 𝑎 𝑎 𝑎 𝑎 𝑎

Couche d’aluminium:

207
 𝑑𝑡 1 2 1 𝑑𝑡
𝑡−𝑑𝑡
𝑇𝑝,𝑖 = [1 + ( ) ( + + )] ̅̅̅̅
𝑇𝑝,𝑖 − ̅̅̅̅̅̅̅
(𝑇 ̅̅̅̅̅̅̅
𝑝,𝑖−1 + 𝑇𝑝,𝑖+1 ) −
2 𝜌𝑝 𝐶𝑝𝑝 𝑉𝑝 𝑅𝑝1 𝑅𝑝2 𝑅𝑝3 2 𝜌𝑝 𝐶𝑝𝑝 𝑉𝑝 𝑅𝑝2
𝑑𝑡 𝑑𝑡
̅̅̅̅
𝑇𝑎,𝑖 − 2 𝜌 ̅̅̅̅̅̅
𝑇 𝑚1,𝑖
2 𝜌𝑝 𝐶𝑝𝑝 𝑉𝑝 𝑅𝑝1 𝑝 𝐶𝑝𝑝 𝑉𝑝 𝑅𝑝3

Couche de MCP:
𝑑𝑡 1 2 1 𝑑𝑡
𝑡−𝑑𝑡
𝑇𝑚𝑗,𝑖 = [1 + (2 𝜌 𝐶𝑝𝑚 𝑉𝑚1
) (𝑅 +𝑅 +𝑅 )] ̅̅̅̅̅̅
𝑇𝑚𝑗,𝑖 − 2 𝜌 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑚 4
̅̅̅̅̅̅̅̅̅
(𝑇 ̅̅̅̅̅̅̅̅̅
𝑚1,𝑖−1 + 𝑇𝑚𝑗,𝑖+1 ) −
𝑚 𝑝3 𝑚4 𝑚7 𝑚
𝑑𝑡 𝑑𝑡
̅̅̅̅
𝑇𝑝,𝑖 − 2 𝜌 ̅̅̅̅̅̅̅̅̅̅̅
𝑇𝑚(𝑗+1),𝑖
2 𝜌𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑝 3 𝑚 𝐶𝑝𝑚 𝑉𝑚1 𝑅𝑚 7

La température de l'air, de l'aluminium et du MCP sont calculées à chaque pas de temps, en utilisant la
température à l’entrée et les températures calculées au pas de temps précédents. Le système
d'équations résultantes est ensuite écrit sous une forme de matrice, ce qui conduit à l'équation finale:

𝑇̅ = 𝐴−1 (𝑇 𝑡−𝑑𝑡 − 𝐵)
Ou 𝑇̅ est une matrice contenant les valeurs moyennes de la température pour les couches d’air,
d’aluminium et de MCP, A est la matrice contenant les coefficients de résistance thermique, 𝑇 𝑡−𝑑𝑡 ) est
une matrice les valeurs de température pour les trois couches au pas de temps précédent et B est un
matrice contenant des valeurs de température initiale pour les trois couches. Enfin, les valeurs de
température pour les trois couches sont calculées par l'équation suivante:
𝑇 𝑡 = 2 𝑇̅ − 𝑇 𝑡−𝑑𝑡
Calibrage du modèle et importance de la capacité spécifique du MCP
Les résultats numériques initiaux ont été promettant, mais pas assez satisfaisant, quand comparé aux
résultats expérimentaux. Après le calibrage de divers facteurs, l'importance des valeurs de la capacité
thermique spécifique (Cp) du MCP a été observée. Dans un premier temps, le modèle utilisait les valeurs
obtenues par le laboratoire d’EDF en utilisant la méthode DSC. Certaines études (Albright et al., Annexe
17 de l’AIE) concluent que le processus de DSC peut fournir des données et que les données fournis de
cette méthode dépendent de la vitesse de chauffage/refroidissement du matériaux investigué.
Cela nous a conduits à la création de courbes de chaleur spécifique adéquates. L'idée initiale était de
créer un triangle pointant vers la valeur de température de changement de phase. Compte tenu des
deux pics observés sur les données de la DSC, un second pic a également été introduit dans la courbe
pour la phase de refroidissement. En outre, il a été décidé de formulé trois courbes différentes selon la
section de MCP considéré (entrée, intermédiaire et de sortie).
La figure 5 présente les résultats expérimentaux et numériques pour l’évolution de la température du
MCP au milieu de l’échangeur pour trois débits différents. La non-linéarité de l'évolution de la
température pendant les changements de phase est reproduite, y compris les deux pics qui se
manifestent pendant la phase de solidification (reliée aux propriétés intrinsèques de la paraffine). Dans
l'ensemble, le modèle développé parvient à simuler avec précision le comportement de tel qu'observé à
travers la plate-forme expérimentale et est considéré comme validé.

208
 50 Tpcm,5 exp 100 m3/h (°C)
Tpcm,5 sim 100 m3/h (°C)
Tpcm,5 exp 300 m3/h (°C)
Temperature [°C]

Tpcm,5 sim 300 m3/h (°C)

40
Tpcm,5 exp 500 m3/h (°C)
Tpcm,5 sim 500 m3/h (°C)

30

20
0 5 10 15 20 25 30 35
Temps [h]

Figure 5: Résultats numériques et expérimentaux pour la température du MCP au nœud 5, débit de

100, 300 and 500 m3/h

Chapitre 5: Optimisation de l’échangeur et élaboration des stratégies de

contrôle
Afin d'évaluer la performance de l'échangeur et le tester dans différents scénarios d’opération, un
bâtiment a été conçu numériquement, en utilisant le modèle Hybcell 1.2. Il s’agit d’un modèle développé
au laboratoire LGCB / ENTPE, sous l'environnement Matlab / Simulink en couplant un modèle thermique
et un modèle d'écoulement de l'air sous pression.
Le bâtiment conçu est un local monozone de 80m2 ayant une forme rectangulaire (10 m de long, 8 m de
large et 3m de haut). Les murs sud, est et ouest ont deux fenêtres de 2.5mx1.5. Les propriétés des murs
et des fenêtres ont été définies de manière à correspondre à une maison à faible énergie. Les données
météorologiques ont été acquises en utilisant la station IDMP située au campus de l'ENTPE (Dumortier)
et correspondent aux conditions climatiques Vaulx-en-Velin pour l'année 2013.
Des tests numériques ont été effectués pour le mois de janvier 2013, en utilisant l’échangeur à sa forme
initiale. Plusieurs configurations ont été testées (température à l’entrée de l’échangeur et période de
décharge) sans pour autant pouvoir satisfaire les critères d’effacement énergétique et de confort
thermique.

Première approche d’optimisation

Le couplage numérique entre échangeur et bâtiment ont confirmé les observations expérimentales
concernant un stockage d'énergie limitée et un temps de charge se produise. La première approche
d'optimisation de l’échangeur consistait en l'utilisation du modèle développé pour modifier certains
paramètres du dispositif. Plus précisément, les paramètres étudiés concernent les dimensions de l'unité,
telles que la hauteur des trois couches considérées, PCM, l'air et de l'aluminium (dz), l'épaisseur de la
couche de PCM (dym) et l'épaisseur de la couche d'air (dya).

209
Trois paramètres ont été introduits dans le modèle numérique en tant que facteurs de multiplication des
paramètres en question. La hauteur a été multipliée par 1, 1,5 et 2, l'épaisseur de la couche de PCM a
été multipliée par 1, 2, 3 et 4 et l'épaisseur de la couche d'air a été multipliée par 1, 2, 3 et 4.

Un total de 48 configurations a été produit et testé pour une période de 4 jours, choisie pour des raisons
de rapidité. Trois critères ont été établis: déplacement de la charge de pointe, consommation mensuelle
minimum et maintien du confort thermique. Une diminution de 0,5 ° C de la température intérieure du
bâtiment a été tolérée.

Conception et essaie de trois scénarios

Suite à l’investigation des résultats obtenus, trois configurations ont été sélectionnées car elles
remplissaient les critères établis. La première utilise plus de MCP, ayant ainsi un plus grand volume, mais
consomme moins d'énergie. La seconde et troisième ont moins de MCP, mais consomment plus
d'énergie. Toutes les trois, offrent différents points d'intérêt et ont été choisies pour tester le système de
‘échangeur de chaleur + bâtiment’ à travers différents scénarios.

Effacement énergétique simple

Le premier scénario avait comme but de vérifier la bonne opération de l'échangeur (déplacement de
charge et maintien du confort thermique) pendant une période de temps plus longue ainsi que de
calculer la consommation d'énergie et la facture d'électricité à la fin du mois.

Lors de ce scenario, la charge de l'échangeur s’effectuait entre 22h00-16h00 et la décharge au cours

d'une période de six heures (16: 00-22: 00). La température de l'air à l’entrée de l'échangeur a été réglée
à 38°C et le débit d'air est régulé par un contrôleur PID. Pour les trois configurations choisies, le système
a réussi à maintenir la température de l'air ambiant à 20°C pendant toute la période d’essai.

Charge nocturne de l’échangeur

Une observation plus détaillée des besoins énergétiques de la maison montre que dans certains cas le
chauffage n’est pas nécessaire pendant la journée. En effet, selon les conditions climatiques extérieures
(principalement gains solaires) la puissance nécessaire diminue ou devient même zéro à partir de 10h00
à 16h00. Cette observation nous a conduit à la création d'un scénario de charge nocturne. Dans ce cas,
l'échangeur est chargé pendant la nuit, à partir de 22h00 pour se terminer à 6h00 du matin. Jusqu'au
début de la période de décharge du dispositif (16:00) un chauffage conventionnel est utilisé, si
nécessaire. Cela permet de profiter des tarifs de nuit moins chères que EDF propose.

Les trois configurations ont été testées et ont réussi à fournir de bons résultats, mais avec une
température d'air à l’entrée de l'échangeur augmenté. Même si la consommation d'énergie a été
augmentée en utilisant l'échangeur, une diminution du coût final est observée car l'échangeur est chargé
pendant la nuit.

Occupation et qualité de l’air intérieur

Enfin, un troisième et dernier scénario a été testé en introduisant la présence d'une famille de quatre
personnes au modèle du bâtiment. La simulation d'occupation provoque deux changements importants
par rapport aux deux scénarios précédents: le dégagement de chaleur et de CO2 par chaque occupant. La

210
concentration de CO2 augmentée étant liée à des problèmes de santé (somnolence, maux de tête,
léthargie) et des niveaux de productivité plus faibles (Siskos et al., 2001) une limite de 1000ppm a été
établie pour cette valeur, selon les standards de ASHRAE.

Un système de ventilation plus avancé a été conçu pour ce scénario. Plus précisément une ventilation à
double flux est considéré chauffer l'air frais entrant. De cette façon, l'air utilisé pour ventiler la maison
est plus chaude que l'air extérieur et non polluée en CO2. En outre, un by-pass a également été
considéré, ce qui signifie que l'air entrant peut circuler à travers l'échangeur de chaleur air-PCM ou être
injecté directement dans la maison. Cela permet le renouvellement de l'air intérieur sans affecter l'état
de charge de l’échangeur. La charge nocturne est maintenue pour ce scénario également.

Le système optimisé proposé a réussi à offrir des conditions acceptables intérieures (confort thermique
et QAI) pour les trois configurations, avec une augmentation de la température d’air à l’entrée de
l’échangeur. Le débit d'air de l'échangeur est principalement défini par le contrôleur de la QAI mais cela
n'a pas interféré avec la charge de l'échangeur de chaleur ou de la température de l'air intérieur.
Les besoins de chauffage ont été réduits lorsque les occupants étaient présents dans la maison.

Conclusions de la première approche d’optimisation

La première démarche d'optimisation a été basée sur une étude de modification des dimensions de
l'échangeur de chaleur et de la performance de l'unité lorsqu'il est couplé à une maison, concernant le
déplacement de la charge, le confort thermique et la QAI.

Nous avons pu démontrer les avantages de l'unité optimisée selon différents scénarios. Cependant, la
première approche d'optimisation a présenté certaines limites concernant la quantité de MCP utilisée, le
volume du dispositif et la période nécessaire pour la charge longue.

Deuxième approche d’optimisation

Les limites de la première approche d'optimisation ont démontré la nécessité d'une seconde étude
d'optimisation de l'échangeur, se concentrant sur le temps nécessaire pour la charge et l’énergie stockée
durant cette période. Cette fois-ci, outre la modification des dimensions du dispositif, les propriétés du
MCP ont également été modifié.

Les paramètres de dimensions ont été modifiés de manière à ne pas changer la quantité totale de PCM.
Par exemple, une diminution de l'épaisseur de la couche de MCP résulte à une augmentation
proportionnelle de la hauteur ou la longueur de la même couche (et respectivement de l'échangeur).

En ce qui concerne les propriétés de MCP, les paramètres modifiés étaient la conductivité, la densité et
la chaleur spécifique. Enfin, la conductivité du conteneur a également été étudiée. Diffèrent MCP ont été
examinés et une plage de valeurs a été proposé pour chaque propriété. Le même processus a été
effectué pour le récipient, en investiguant la conductivité des matériaux divers (cuivre, l'acier, etc.). En
conséquence, trois nouvelles valeurs pour chaque propriété ont été testées selon la plage établie.

211
 Les tests d'optimisation ont été réalisés sous l’environnement Matlab pour une charge et décharge
complète ; un critère d’arrêt a été établi pour la fin de ce processus. Pour chacune des trois
configurations de dimensions différentes, la combinaison de toutes les valeurs possibles des propriétés
du MCP a conduit à un total de 243 cas testés.

Des tests pour trois débits différents ont été réalisés (100, 300 et 500 m3/h) et les résultats ont été
classés en fonction de la durée minimale pour le stockage ou déstockage de chaleur. Enfin, des
simulations ont été réalisées pour la dimension initiale de l'échangeur, tout en modifiant les valeurs des
propriétés de PCM. Toutes les configurations combinées un total de 4860 simulations a été réalisée.

Une partie des résultats est présentée au tableau 2 ; il s’agit des configurations qui ont été testées pour
les scénarios mentionnés précédemment.

Conducti-
Densité Capacité
Classifi- Conducti- vité du Epaisseur Longueur Auteur Temps Energie
MCP spécifique
cation vité MCP containe (m) (m) (m) (h) (KWh)
MCP
ur

1 4 0,9 0,7 0,5 0,006 0,1231 0,2216 3,35 2721,2

28 4 1,4 0,7 0,5 0,006 0,1231 0,2216 4,86 4050,71

38 3 0,9 1,25 0,5 0,006 0,1231 0,2216 5,5 4606,99

64 4 0,9 1,5 0,5 0,006 0,1231 0,2216 6,4 5447

Tableau 2: Classification des résultats obtenus pour le chargement de l'unité selon la durée de
chargement plus rapide, taux de 300m3 / h de débit d'air

Une tendance a été détectée pour toutes les conditions de test différents (charge / décharge, débits
différents), même si de petites variations existent : les configurations positionnées aux premières places
sont celles avec des valeurs de chaleur spécifique et de densité plus petites, de couche de MCP plus
mince et des valeurs de conductivité plus grandes.

La comparaison entre la forme initiale et les versions optimisées de l'échangeur montre une charge plus
rapide du dispositif pour la même quantité de chaleur stockée. Les versions optimisées présentent une
diminution d'une heure du temps nécessaire pour la charge.

Les résultats de la deuxième étude d'optimisation nous ont fourni une grande liste de configurations,
chacune présentant des avantages et des inconvénients pour notre cas d’étude. Les scénarios qui ont été
testés avec les résultats de la première approche d'optimisation ont été reproduits en utilisant les
configurations récemment obtenus.

Effacement simple, deuxième approche d’optimisation

L'un des objectifs de la deuxième approche était de réduire la période d’effacement afin de mieux se
conformer au problème du pic de consommation observé en France. Le même scénario s’est donc

212
reproduit, en modifiant toutefois la période d’effacement à 18: 00-20: 00 et en testant les configurations
obtenues avec la deuxième approche.

La première configuration testée est celle qui a présenté la charge la plus rapide (cas 1, tableau 2). Cette
configuration a été testée pour diverses valeurs de la température de l'air à l’entrée de l'échangeur. La
configuration n'a pas fourni de résultats satisfaisants, car la température de l'air intérieur a diminué
pendant la période de déplacement de la charge. Un second test a été réalisé, en utilisant la
configuration 28 (tableau 2) vu que c’est la configuration présentant une augmentation significative de
la quantité d'énergie stockée. Cette configuration a aussi été testée pour des températures de l'air
augmentées, mais aucun d'entre elles n’a réussi à maintenir une température de l'air intérieur stable
pendant la période d’effacement.

Après l'insuffisance des deux premiers cas, la configuration 38 a été testée. Elle correspond à la
configuration nécessitant le même temps pour une charge complète que la version initiale de
l'échangeur. L'énergie stockée augmentée de cette configuration a permis un fonctionnement efficace
du système.

Nous avons ensuite voulu tester les performances de cette configuration avec un scénario plus difficile.
Comme décrit dans le scénario d'occupation de la première étude d'optimisation, en cas de présence
humaine l’air intérieur ne peut plus être réinjecté dans la maison, car il est considéré être pollué. Afin de
surmonter cet obstacle un système de récupération de chaleur a été utilisé, fournissant de l'air à une
température inférieure à celle à l'intérieur. Cette fois, nous avons testé cette option, mais à une
condition plus difficile: la température de l'air à l’'entrée de l'échangeur pendant la période
d’effacement a été prise comme la valeur de la température de l'air intérieur moins 4 °C. Le test de la
38ème configuration dans de telles conditions a montré qu'une augmentation de la température de l'air
au cours de la période de pic était nécessaire afin d'obtenir des conditions de confort thermique stables.

Enfin, une configuration présentant un stockage d'énergie encore plus élevé a été testé (cas 64, tableau
2). Cette configuration a réussi à fournir un fonctionnement efficace du système avec une température,
même avec une température d’air à l’entrée de 38°C de l'air pendant le processus de charge et une
température intérieure moins 4 °C au cours de la décharge.

Charge nocturne, deuxième approche d’optimisation

Les tests présentés dans le paragraphe précédent nous ont fourni avec deux configurations capables de
satisfaire les critères de déplacement de la charge et des conditions intérieures acceptables. Ces deux
configurations ont été utilisées pour reproduire le scénario de charge nocturne.

Deux modifications ont été apportées dans le fonctionnement du système: la période d’effacement a été
réduite à deux heures (18: 00-20: 00) et un chauffage classique a été utilisé de 20 :00 à 22:00.

Le cas 38 a d'abord été testé avec une température d'air à l’entrée de l'échangeur de 44 °C pendant la
charge de l'échangeur. La diminution de la quantité de MCP et la durée de conservation plus longue ont
conduit à une décharge involontaire importante de l'appareil. En conséquence, la température de l'air
intérieur a diminué pendant la période de décharge. Compte tenu de ce fait, nous avons décidé de

213
réduire les pertes de chaleur de l'échangeur par un facteur de 30%. Le dernier essai a ensuite été répété
pour une version mieux isolé de l'échangeur de chaleur. La température de l'air de l'échangeur de sortie
diminue à un rythme plus faible pendant la période de conservation et la température intérieure
présente des variations plus petites.

En outre, le cas 64 a été testé pour le même scénario et conditions. Comme pour le cas 34, la
performance de l'appareil avec l'isolation initiale n’était pas suffisante pour une température de l'air
intérieur stable. La modification de ce paramètre a démontré une meilleure performance.

Occupation et QAI, deuxième approche d’optimisation

Le scénario d'occupation a été reproduit dans le but de tester les performances des configurations 34 et
64. Le système est considéré fonctionner de manière similaire à la première approche d'optimisation, à
la différence d'une période de transfert de charge plus courte. Comme pour la première approche
d'optimisation, la présence d'occupants nécessite un système de récupération de chaleur et un système
by-pass de l'échangeur; elle implique également des gains de chaque personne de chaleur.

La configuration 38 a été testée pour des conditions d'isolation thermique initiales et améliorées. Les
deux ont présentées une performance adéquate, même si le premier cas conduit à une légère baisse de
la température intérieure à la fin de la période d’effacement. La figure 6 présente la performance du
système pour le dernier cas, en zoomant sur les jours 3 à 8, où les besoins énergétiques sont élevés.

Figure 6 : Performance de l'échangeur, jours3-8, 38 cas, dernier cas

214
Enfin, la configuration 64 a également été testée pour le scénario d'occupation, pour une température
de l’air à l’entrée de 38°C et de 40°C et pour une isolation initiale et améliorée. Mis à part le premier cas
(38°C et isolation initiale), les configurations testées ont réussi à maintenir une température stable à
l'intérieur pendant la décharge de l'échangeur.

Conclusions et perspectives
Synopsis
La recherche effectuée lors de cette thèse a été principalement motivée par le problème de la pointe
électrique observé en France au cours de l'hiver en fin d'après-midi. Comme ces pics sont fortement liés
à chauffage électrique, l'objectif de notre travail était d'une part de concevoir et de développer des
outils numériques et expérimentaux fiables, basés sur la technologie de stockage thermique et d'autre
part d'utiliser ces outils afin de développer des stratégies de contrôle efficaces, basées sur des solutions
de déplacement de la charge.
La partie expérimentale de l'étude a consisté à la conception et la construction d'un échangeur de
chaleur MCP-Air comme le principal élément de la plate-forme expérimentale et l'endroit où le stockage
thermique a lieu sous forme latente. Le système de stockage de chaleur est composé d'un ensemble de
plaques MCP (paraffine), calées dans l'échangeur et intégrées dans un système de ventilation.
L'inconvénient initialement observé de la fuite MCP a été surmonté par la construction d'un deuxième
échangeur possédant une structure plus rigoureuse. Les résultats de la caractérisation ont démontré
l'applicabilité du système proposé mais ont aussi révélés des limites, surtout en ce qui concerne le temps
nécessaire pour la charge et la décharge de l'échangeur.
Cette observation a conduit au développement d'un modèle numérique bidimensionnel dans le but de
reproduire le comportement de l'échangeur. Le modèle utilise plusieurs méthodes et techniques, y
compris l'approche du bilan thermique, la capacité thermique apparente et la méthode des différences
finies afin de représenter avec succès le processus de changement de phase et le fonctionnement
général du système. L'importance des valeurs de capacité de chaleur a été révélée et une enquête plus
approfondie de ce paramètre a conduit à la création de courbes adéquates en fonction de la position
considérée de la plaque de MCP. Résultats numériques et expérimentaux ont été confrontés afin de
calibrer et valider le modèle. Suite à cela, le modèle a été couplé à un modèle de bâtiment existant afin
de tester les performances de l'échangeur à propos de l’effacement énergétique, le confort thermique et
la qualité de l'air intérieur lors des différents scénarios.
Deux approches d'optimisation ont été considérées et numériquement exploitées pour surmonter les
limites de la performance initiale de l'échangeur. La première a examiné l'impact de la modification de la
géométrie du dispositif lorsque celui est couplé à l'immeuble. Trois configurations ont été sélectionnées
et testées sur un bâtiment à basse énergie pour une période d'un mois. Les avantages du système
proposé ont été démontrées pour des scénarios présentant une difficulté croissante: déplacement de la
charge simple, charge du dispositif pendant la nuit et enfin fonctionnement pendant la présence des
occupants. La consommation électrique a été déplacée des heures pleines aux heures creuses sans
affecter le confort thermique et la qualité de l'air. Bien que la consommation d'énergie finale mensuelle

215
a été augmentée, la charge nocturne de l'unité a permis le fonctionnement du système pendant les
heures avec des tarifs avantageux, conduisant à un coût d'électricité final réduit. Néanmoins, cette
approche a montré certains inconvénients concernant la masse de MCP utilisée, le volume de
l'échangeur et le temps élevé pour la charge du dispositif.
Cela a conduit à la seconde approche d'optimisation qui, en plus de la géométrie de l'échangeur a
également étudié les propriétés du MCP. Densité, chaleur spécifique et conductivité ont été modifiées,
en cherchant les configurations avec un temps de charge minimum et stockage d'énergie maximum. Les
versions optimisées du dispositif ont été classées selon la durée de charge et un gain d'une heure a été
observé pour le même stockage d'énergie. À la fin, les scénarios testés avec les résultats de la première
approche d'optimisation ont été reproduits et nous avons pu obtenir un système efficace avec une
masse de MCP initiale, capable d’effectuer l’effacement souhaité tout en maintenant des conditions
thermiques et qualité d’air confortables.

Perspectives
Dans cette thèse, l'applicabilité et les avantages de la technologie de stockage de chaleur latente ont été
démontrés par l'étude numérique et expérimentale d'un échangeur de chaleur MCP-Air, intégré dans un
système de ventilation. Plusieurs domaines peuvent être examinés comme une continuité de ce travail.
Tout d'abord, des contrôleurs avancés peuvent être utilisés pour le fonctionnement du système, tels que
ceux utilisant la logique floue. Ces contrôleurs peuvent offrir une approche plus adaptée aux besoins de
l'étude, présenter des variations inferieures et des résultats meilleurs. En outre, des outils prédictifs
peuvent être intégrés dans le système contrôle / commande du système. Ces outils peuvent se focaliser
sur les conditions météorologiques ou la présence des occupants afin d'anticiper les besoins de
chauffage du bâtiment et ajuster le fonctionnement du système en conséquence. Selon les besoins
prévus, les niveaux de stockage d'énergie différents peuvent être décidées, conduisant à une charge
faible, modérée ou complète de l'échangeur. Des méthodes Metaheuristic tels que les algorithmes
génétiques peuvent également être utilisées dans l'étude d'optimisation fournissant une approche
différente, mais plus efficace.
En outre, l'efficacité du système proposé peut être testée pour des propriétés différentes du bâtiment.
La modification de l'inertie ou de l'isolation thermique du bâtiment peuvent constituer des exemples
d'une telle étude. Une conception plus détaillée de l'immeuble avec la séparation des zones selon
l'utilisation et les besoins des occupants peut également être étudié.
La robustesse des solutions numériques proposées peut être testée dans une expérimentation in situ,
que ce soit dans une maison expérimentale ou une habitation réelle avec des occupants qui
comprennent le fonctionnement du système. Le dernier cas nécessiterait une enquête plus large de
l'impact de l'intégration du système dans un environnement habité, mettant l'accent sur les risques
d'incendie et la diffusion de l'odeur du MCP.
L'étude du système peut être étendue à des conditions estivales afin de profiter concept du
refroidissement nocturne, soit par voie expérimentale soit numériquement. Dans le premier cas, le choix
de macroencapsulation et de la structure rigide du second échangeur de chaleur peut permettre un

216
remplacement relativement facile de la paraffine utilisée avec un MCP plus adapté aux conditions d'été
(essentiellement un MCP avec une plage de température de fusion différente, peut-être une capacité de
stockage d'énergie différent). Dans le second cas, l'adaptation à des conditions estivales sera effectuée
par la modification des propriétés du MCP dans le modèle numérique. Osterman et al. sont en train
d'étudier une configuration d'échangeur similaire avec un MCP qui peut satisfaire les deux climats :
hivernal et estival. Il serait intéressant de voir si leurs résultats peuvent contribuer à ce domaine et de
vérifier l'efficacité d’un tel dispositif avec les outils que nous avons développés.
Enfin, même si cette étude a porté sur la question des pics de consommation, la technologie de stockage
thermique peut être couplé à d'autres technologies innovantes existants tels que les énergies
renouvelables, les smart grids, les murs vivants, etc. De cette façon, les avantages de chaque domaine
peuvent être combinés afin d’obtenir des solutions plus larges et plus efficaces en ce qui concerne le
fonctionnement du bâtiment.

217
218
References

219
References

Abedin A., Rosen M. (2011). A Critical Review of Thermochemical Energy Storage Systems. The Open
Renewable Energy Journal, 4, 42-46

ADEME. (2013, a) Bâtiment, Chiffres Clés

ADEME. (2013, b) Climat, air et énergie

ADEME & RTE. (2007), Le contenu en CO2 du kWh électrique: Avantages comparés du contenu marginal
et du contenu par usages sur la base de l’historique

Albright LD (1990) Environment control for animals and plants. ASAE Textbook, 4. American Society of
Agricultural Engineers, Michigan, St. Joseph

Arkar C., Medved S., (2005). Influence of accuracy of thermal property data of a phase change material
on the result of a numerical model of a packed bed latent heat storage with spheres,
Thermochimica Acta, 438, 192-201

Arkar A., Vidrih B., Medved S. (2006). Efficiency of free cooling using latent heat storage integrated into
the ventilation system of a low energy building, International Journal of Refrigeration, 30, 134-143

ASHRAE. (2013). Ventilation for Acceptable Indoor Air Quality (ANSI Approved), Guidelines in ASHRAE
Standard 62.1-2013, Appendix B, Summary of Selected Air Quality

Baetens R., Jelle B. P., Gustavsen A. (2010). Phase change materials for building applications: A state-of-
the-art review. Energy and Buildings, 42, 1361-1368

Borderon J. (2012). Integration des materiaux à changement de phase comme systeme de régulation
dynamique en rénovation thermique, PhD thesis, ENTPE France

Cabeza L.F., Castell A., Barreneche C., de Gracia A., Fernández A.I. (2011). Materials used as PCM in
thermal energy storage in buildings: A review. Renewable and Sustainable Energy Reviews, 15,
1675-1695

Cabeza L., Mehling H. (2008). Heat and cold storage with PCM: An up to date introduction into basics and
applications, Springer , Handbook

Chen H., Cong T.N., Yang W., Tan C., Li Y., Ding Y. (2009). Progress in electrical energy storage system: A
critical review. Progress in Natural Science, 19, 291-312

220
Dolado P., Lazaro A., Marin J.M., Zalba B. (2011). Characterization of melting and solidification in a real-
scale PCM-air heat exchanger: Experimental results and empirical model, Renewable Energy, 36,
2906-2917

Dumortier D. (2003). Lumière naturelle et rayonnement solaire - Mesures au sol et estimations á partir
d'images satellites - Développement de services Web pour l'éclairage naturel des bâtiments, Thèse
d'habilitation, INSA- Univ. Claude Bernard Lyon 1, Septembre 2003.
[Link]

Dutil Y., Rousse D. R., Ben Salah N., Lassue S., Zalewski L., (2011). A review on phase-change materials:
Mathematical modeling and simulations, Renewable and Sustainable Energy Reviews, 15, 112-130

El Mankibi M. (2003). Development and evaluation of experimental and numerical control strategies for
hybrid ventilation, PhD Thesis, INSA Lyon

European Comission. (2002). Energy Performance of Buildings Directive (Directive 2002/91/EC)

European Commission. (2008). A strategy for smart, sustainable and inclusive growth

European Commission. (2011). Energy Roadmap 2050

Farid M., Khudhair A., Razack S., Al-Hallaj S. (2004). A review on phase change energy storage: materials
and applications. Energy Conversion and Management, 45, 1597-1615

Fernandes D., Pitié F., Cáceres G., Baeyens J. (2012). Thermal energy storage: “How previous findings
determine current research priorities” Energy, 39, 246-257

Hadjieva, M., Kanev S., Argirov J. (1992). Thermophysical properties of some paraffins applicable to
thermal energy storage. Solar Energy Materials and Solar Cells, 27, 181-187

Hajat S., Kovats R.S., Atkinson R.W. & Haines A. (2002). Impact of hot temperatures on death in London:
a time series approach. Journal of Epidemiology and Community Health, 56, 367-372

Hale D.V., Hoover M.J., O’Neill M.J. (1971). Phase change materials hand book. Alabaa: Marshal Space
Flight Center.

Heim D., Clarke J.A. (2004). Numerical modelling and thermal simulation of PCM–gypsum composites
with ESP-r, Energy and Buildings, 36, 795-805

Hu H., Argyropoulos S., (1996). Mathematical modelling of solidification and melting: a review, Modelling
and Simulation in Materials Science and Engineering, 4, 371-396.

Idelsohn S., Storti M., Crivelli L. (1994). Numerical methods in phase-change problems, Archives of
Computational Methods in Engineering, 1, 49-74.

221
IEA. Annex 17. (2005). Advanced Thermal Energy Storage through Phase Change Materials and Chemical
Reactions – Feasibility Studies and Demonstration projects, Final report.

IEA. Annex 23. (2013). Applying Energy Storage in Ultra-low Energy Buildings, Final report.

IEA Task 32. (2005). Thermal energy storage for solar and low energy buildings. Handbook

IEA. (2008). Energy efficiency requirements in building codes, energy efficiency policies or new buildings,
IEA Information paper

INSEE. (2011). La précarité énergétique : avoir froid ou dépenser trop pour se chauffer

INSEE. (2012). Bilan énergétique 2012

Khudhair A. M., Farid M. (2003). A review on energy conservation in building applications with thermal
storage by latent heat using phase change materials. Energy Conversion and Management, 45,
263-275

Lamberg P., Sirén K. (2003). Approximate analytical model for solidification in a finite PCM storage with
internal fins, Applied Mathematical Modelling, 7, 491-513

Lazaro A., Dolado P., Marín J.M., Zalba B. (2009). PCM–air heat exchangers for free-cooling applications
in buildings: Experimental results of two real-scale prototypes, Energy Conversion and
Management, 50, 439-443

Lin K., Zhang Y., Di H., Yang R. (2007). Study of an electrical heating system with ductless air supply and
shape-stabilized PCM for thermal storage, Energy Conversion and Management, 48, 2016-2024

Mackenzie J.A., Robertson M.L. (2002). A moving mesh method for the solution of the one-dimensional
phase-field equations. Comput Phys, 181, 526–44.

MEDDE. (2006). Actualisation 2006 du Plan Climat 2004-2012

MEDDE. (2009). La première loi du Grenelle

MEDDE. (2010). Loi Grenelle 2

MEDDTL. (2010). Bilan énergétique de la France rapport du SOeS

Mitchell S. L., Myers T. G. (2012). Application of heat balance integral methods to one-dimensional phase
change problems, International Journal of Differential Equations

222
Nagano K., Takeda S., Mochida T., Shikamura K. (2004). Thermal characteristics of a direct heat exchange
system between granules with phase change material and air, Applied Thermal Engineering, 24,
2131-2144

Nedjar B. (2002). An enthalpy-based finite element method for nonlinear heat problems involving phase
change, Comput Struct, 80, 9–21.

NégaWatt. (2009). La pointe d’électricité en France … zéro pointé !, Dossier de Presse

Osterman E., Butala V., Stritih U. (2013). Room Heat Exchanger with PCM as a Thermal Accumulation for
Heating and Cooling of Buildings. Conference: CLIMA 2013 – 11th REHVA world congress and the
8th international conference on IAQVEC

Parameshwaran R., Kalaiselvam S., Harikrishnan S., Elayaperumal A. (2012). Sustainable thermal energy
storage technologies for buildings: A review, Renewable and Sustainable Energy Reviews, 16,
2394-2433

Pasupathy A., Velraj R., Seeniraj R.V. (2008). Phase change material-based building architecture for
thermal management in residential and commercial establishments, Renewable and Sustainable
Energy Reviews, 12, 39-64

Pomianowski M., Heiselberg P., Zhang Y. (2013). Review of thermal energy storage technologies based
on PCM application in buildings, Energy and Buildings, 67, 56-69

Robine J.-M., Cheung S.L., Le Roy S., Van Oyen H., Griffith C., Michel J.-P. (2008). Death toll exceeded
70,000 in Europe during the summer of 2003. Comptes Rendus Biologies,331, 171-178

Rolph W.D. and Bathe K. J. (1982). An efficient algorithm for analysis of nonlinear heat transfer with
phase changes, Int. J. Numer. Meth. Eng., 18, 119–34

RTE. éCO2mix. [Link]/lienrapide/eco2mix

RTE. (2012, a). Bilan électrique

RTE. (2012, b). La vague de froid de février

RTE. (2013).Bilan électrique

RTE. (2014). Memo

Salyer IO, Sircar AK. (1990). Phase change materials for heating and cooling of residential buildings and
other applications. Proceedings of 25th Intersociety Energy Conversion Engineering Conference,
236-43.

223
Sharma A., Tyagi V.V., Chen C.R., Buddhi D. (2009). Review on thermal energy storage with phase change
materials and applications, Renewable and Sustainable Energy Reviews, 13, 318-345

Siskos P. A., Bouba K. E., Stroubou A.P. (2001). Determination of Selected Pollutants and Measurement
of Physical Parameters for the Evaluation of Indoor Air Quality in School Buildings in Athens,
Greece. Indoor and Built Environment, 10, (3-4), pp185-192.

Soares N., Costa J.J., Gaspar A.R., Santos P. (2013). Review of passive PCM latent heat thermal energy
storage systems towards buildings’ energy efficiency, Energy and Buildings, 59, 82-103

Turnpenny J.R., Etheridge D.W., Reay D.A. (2001). Novel ventilation system for reducing air conditioning
in buildings. Part II: testing of the prototype, Applied Thermal Energy, 21, 1203-1217

UN. (2013). World population projected to reach 9.6 billion by 2050 with most growth in developing
regions, Press release

UNEP SBCI. (2009). Buildings and climate change, Summary for decision makers

United Nations. (1998). Kyoto protocol to the United Nations framework convention on climate change

Verma P., Varun, Singal S.K. (2008). Review of mathematical modeling on latent heat thermal energy
storage systems using phase-change material, Renewable and Sustainable Energy Reviews, 12,
999-1031

Voller V.R., Swenson J.B., Kim W., Paola C. (2006). An enthalpy method for moving boundary problems
on the earth's surface. International Journal of Numerical Methods for Heat & Fluid Flow, 16, 641 -
54.

Voller V.R. and Swaminathan C.R. (1991). General source-based method for solidification phase change
Numer. Heat Transfer, 19, 175–89

Zalba B., Marın J.M., Cabeza L., Mehling H. (2003). Review on thermal energy storage with phase change:
materials, heat transfer analysis and applications, Applied Thermal Engineering, 23, 251-283

Zalba B., Marin J.M., Cabeza L.F., Melhing H. (2004). Free-cooling of buildings with phase change
materials, International Journal of Refrigeration, 27, 839-849

Zhou D., Zhao C.Y., Tian Y. (2012). Review on thermal energy storage with phase change materials
(PCMs) in building applications, Applied Energy, 92, 593-605

224
225
Appendices

226
Appendices

Appendix A1

Second optimization approach, charging of the unit, airflow rate of 300 m3/h

lmpc rmcp cpmcp lalu dym dx dz time energy

4 0,9 0,7 0,5 0,001188 0,12309149 0,22156468 3,35 2721,19836
3 0,9 0,7 0,5 0,001188 0,12309149 0,22156468 3,36666667 2725,61476
2 0,9 0,7 0,5 0,001188 0,12309149 0,22156468 3,4 2729,82398
4 0,9 0,7 1,3 0,001188 0,12309149 0,22156468 3,49722222 2751,40226
3 0,9 0,7 1,3 0,001188 0,12309149 0,22156468 3,51111111 2753,56008
4 0,9 0,7 1,6 0,001188 0,12309149 0,22156468 3,53888889 2758,53639
2 0,9 0,7 1,3 0,001188 0,12309149 0,22156468 3,54166667 2756,04927
3 0,9 0,7 1,6 0,001188 0,12309149 0,22156468 3,55277778 2760,61904
2 0,9 0,7 1,6 0,001188 0,12309149 0,22156468 3,58333333 2762,71056
4 0,9 0,7 0,5 0,002394 0,086711 0,15607979 3,93888889 2387,76575
3 0,9 0,7 0,5 0,002394 0,086711 0,15607979 4,00277778 2394,54682
4 0,9 0,7 1,3 0,002394 0,086711 0,15607979 4,02777778 2396,99669
4 0,9 0,7 1,6 0,002394 0,086711 0,15607979 4,05 2398,51691
3 0,9 0,7 1,3 0,002394 0,086711 0,15607979 4,08888889 2402,4226
3 0,9 0,7 1,6 0,002394 0,086711 0,15607979 4,11388889 2404,59515
2 0,9 0,7 0,5 0,002394 0,086711 0,15607979 4,13333333 2400,4828
2 0,9 0,7 1,3 0,002394 0,086711 0,15607979 4,21111111 2406,52146
2 0,9 0,7 1,6 0,002394 0,086711 0,15607979 4,23611111 2408,47018
4 0,9 0,7 0,5 0,002988 0,07761505 0,13970709 4,26388889 2311,23413
4 0,9 0,7 1,3 0,002988 0,07761505 0,13970709 4,33055556 2316,96383
4 0,9 0,7 1,6 0,002988 0,07761505 0,13970709 4,34722222 2317,87191
3 0,9 0,7 0,5 0,002988 0,07761505 0,13970709 4,36388889 2321,57141
3 0,9 0,7 1,3 0,002988 0,07761505 0,13970709 4,425 2325,73186
3 0,9 0,7 1,6 0,002988 0,07761505 0,13970709 4,44166667 2326,65494
2 0,9 0,7 0,5 0,002988 0,07761505 0,13970709 4,56111111 2329,97805
2 0,9 0,7 1,3 0,002988 0,07761505 0,13970709 4,61666667 2332,9001
2 0,9 0,7 1,6 0,002988 0,07761505 0,13970709 4,63333333 2333,54911
4 1,4 0,7 0,5 0,001188 0,12309149 0,22156468 4,85277778 4050,70151
3 1,4 0,7 0,5 0,001188 0,12309149 0,22156468 4,87777778 4056,91827
2 1,4 0,7 0,5 0,001188 0,12309149 0,22156468 4,93333333 4064,83634
4 1,4 0,7 1,3 0,001188 0,12309149 0,22156468 5,03333333 4087,50912
3 1,4 0,7 1,3 0,001188 0,12309149 0,22156468 5,05833333 4091,67517
4 1,4 0,7 1,6 0,001188 0,12309149 0,22156468 5,08611111 4096,05619

227
2 1,4 0,7 1,3 0,001188 0,12309149 0,22156468 5,10555556 4094,99408
3 1,4 0,7 1,6 0,001188 0,12309149 0,22156468 5,11111111 4100,16537
2 1,4 0,7 1,6 0,001188 0,12309149 0,22156468 5,15555556 4102,45887
4 0,9 1,25 0,5 0,001188 0,12309149 0,22156468 5,46388889 4598,92341
3 0,9 1,25 0,5 0,001188 0,12309149 0,22156468 5,49444444 4606,98771
2 0,9 1,25 0,5 0,001188 0,12309149 0,22156468 5,55555556 4615,18585
4 0,9 1,25 1,3 0,001188 0,12309149 0,22156468 5,65833333 4638,48075
3 0,9 1,25 1,3 0,001188 0,12309149 0,22156468 5,68333333 4642,09215
4 0,9 1,25 1,6 0,001188 0,12309149 0,22156468 5,71388889 4647,34073
2 0,9 1,25 1,3 0,001188 0,12309149 0,22156468 5,73611111 4645,70143
3 0,9 1,25 1,6 0,001188 0,12309149 0,22156468 5,73888889 4650,96486
4 1,4 0,7 0,5 0,002394 0,086711 0,15607979 5,73888889 3607,85736
2 0,9 1,25 1,6 0,001188 0,12309149 0,22156468 5,79166667 4654,40806
3 1,4 0,7 0,5 0,002394 0,086711 0,15607979 5,83611111 3617,58412
4 1,4 0,7 1,3 0,002394 0,086711 0,15607979 5,84444444 3618,15316
4 1,4 0,7 1,6 0,002394 0,086711 0,15607979 5,875 3620,30233
3 1,4 0,7 1,3 0,002394 0,086711 0,15607979 5,93611111 3626,06128
3 1,4 0,7 1,6 0,002394 0,086711 0,15607979 5,96388889 3627,55253
2 1,4 0,7 0,5 0,002394 0,086711 0,15607979 6,03333333 3626,18948
2 1,4 0,7 1,3 0,002394 0,086711 0,15607979 6,11944444 3631,32663
4 1,84 0,7 0,5 0,001188 0,12309149 0,22156468 6,14444444 5213,86618
2 1,4 0,7 1,6 0,002394 0,086711 0,15607979 6,14722222 3632,89575
3 1,84 0,7 0,5 0,001188 0,12309149 0,22156468 6,18055556 5222,87072
4 1,4 0,7 0,5 0,002988 0,07761505 0,13970709 6,21111111 3502,54953
2 1,84 0,7 0,5 0,001188 0,12309149 0,22156468 6,25 5232,20026
4 1,4 0,7 1,3 0,002988 0,07761505 0,13970709 6,28611111 3508,12827
4 1,4 0,7 1,6 0,002988 0,07761505 0,13970709 6,30833333 3509,2345
4 1,84 0,7 1,3 0,001188 0,12309149 0,22156468 6,34722222 5254,48461
3 1,4 0,7 0,5 0,002988 0,07761505 0,13970709 6,35833333 3516,68712
3 1,84 0,7 1,3 0,001188 0,12309149 0,22156468 6,37777778 5259,18909
4 0,9 1,5 0,5 0,001188 0,12309149 0,22156468 6,4 5446,99531
4 1,84 0,7 1,6 0,001188 0,12309149 0,22156468 6,40833333 5264,20843
3 1,4 0,7 1,3 0,002988 0,07761505 0,13970709 6,425 3520,34817
3 1,84 0,7 1,6 0,001188 0,12309149 0,22156468 6,43611111 5268,13726
2 1,84 0,7 1,3 0,001188 0,12309149 0,22156468 6,43888889 5263,55835
3 0,9 1,5 0,5 0,001188 0,12309149 0,22156468 6,43888889 5456,68106
3 1,4 0,7 1,6 0,002988 0,07761505 0,13970709 6,44722222 3521,20711
4 0,9 1,25 0,5 0,002394 0,086711 0,15607979 6,46388889 4110,6699
2 1,84 0,7 1,6 0,001188 0,12309149 0,22156468 6,49722222 5272,01883
2 0,9 1,5 0,5 0,001188 0,12309149 0,22156468 6,51388889 5467,08696
4 0,9 1,25 1,3 0,002394 0,086711 0,15607979 6,57222222 4120,72458
3 0,9 1,25 0,5 0,002394 0,086711 0,15607979 6,575 4121,79014
4 0,9 1,25 1,6 0,002394 0,086711 0,15607979 6,60555556 4123,07835

228
4 0,9 1,5 1,3 0,001188 0,12309149 0,22156468 6,60833333 5488,70004
3 0,9 1,5 1,3 0,001188 0,12309149 0,22156468 6,63888889 5493,25458
2 1,4 0,7 0,5 0,002988 0,07761505 0,13970709 6,64722222 3527,70537
4 0,9 1,5 1,6 0,001188 0,12309149 0,22156468 6,66944444 5498,28612
3 0,9 1,25 1,3 0,002394 0,086711 0,15607979 6,675 4129,47134
3 0,9 1,5 1,6 0,001188 0,12309149 0,22156468 6,7 5502,80612
2 1,4 0,7 1,3 0,002988 0,07761505 0,13970709 6,70277778 3528,83147
2 0,9 1,5 1,3 0,001188 0,12309149 0,22156468 6,70555556 5498,49131
3 0,9 1,25 1,6 0,002394 0,086711 0,15607979 6,70555556 4131,2026
2 1,4 0,7 1,6 0,002988 0,07761505 0,13970709 6,72222222 3529,28282
2 0,9 1,5 1,6 0,001188 0,12309149 0,22156468 6,76388889 5506,89779
2 0,9 1,25 0,5 0,002394 0,086711 0,15607979 6,79444444 4130,44066
2 0,9 1,25 1,3 0,002394 0,086711 0,15607979 6,88333333 4135,35944
2 0,9 1,25 1,6 0,002394 0,086711 0,15607979 6,91111111 4136,48876
4 0,9 1,25 0,5 0,002988 0,07761505 0,13970709 6,99166667 3993,29705
4 0,9 1,25 1,3 0,002988 0,07761505 0,13970709 7,06666667 3998,39045
4 0,9 1,25 1,6 0,002988 0,07761505 0,13970709 7,09166667 3999,58447
3 0,9 1,25 0,5 0,002988 0,07761505 0,13970709 7,15555556 4008,70851
3 0,9 1,25 1,3 0,002988 0,07761505 0,13970709 7,225 4011,93412
3 0,9 1,25 1,6 0,002988 0,07761505 0,13970709 7,24722222 4012,64634
4 1,84 0,7 0,5 0,002394 0,086711 0,15607979 7,26666667 4674,22742
4 1,84 0,7 1,3 0,002394 0,086711 0,15607979 7,37777778 4684,22791
3 1,84 0,7 0,5 0,002394 0,086711 0,15607979 7,39166667 4686,43293
4 1,84 0,7 1,6 0,002394 0,086711 0,15607979 7,41111111 4686,22201
2 0,9 1,25 0,5 0,002988 0,07761505 0,13970709 7,48055556 4020,43838
3 1,84 0,7 1,3 0,002394 0,086711 0,15607979 7,49444444 4693,88355
3 1,84 0,7 1,6 0,002394 0,086711 0,15607979 7,52777778 4695,69876
2 0,9 1,25 1,3 0,002988 0,07761505 0,13970709 7,53611111 4020,80681
2 0,9 1,25 1,6 0,002988 0,07761505 0,13970709 7,55555556 4021,09391
4 0,9 1,5 0,5 0,002394 0,086711 0,15607979 7,56944444 4888,28128
2 1,84 0,7 0,5 0,002394 0,086711 0,15607979 7,63888889 4696,05309
4 0,9 1,5 1,3 0,002394 0,086711 0,15607979 7,68333333 4898,25756
3 0,9 1,5 0,5 0,002394 0,086711 0,15607979 7,7 4901,04107
4 0,9 1,5 1,6 0,002394 0,086711 0,15607979 7,71666667 4900,25579
2 1,84 0,7 1,3 0,002394 0,086711 0,15607979 7,72777778 4699,85581
2 1,84 0,7 1,6 0,002394 0,086711 0,15607979 7,75555556 4700,78033
3 0,9 1,5 1,3 0,002394 0,086711 0,15607979 7,80277778 4908,05003
3 0,9 1,5 1,6 0,002394 0,086711 0,15607979 7,83611111 4909,67171
4 1,84 0,7 0,5 0,002988 0,07761505 0,13970709 7,85555556 4543,26531
4 1,84 0,7 1,3 0,002988 0,07761505 0,13970709 7,93055556 4548,01756
4 1,84 0,7 1,6 0,002988 0,07761505 0,13970709 7,95277778 4548,68275
2 0,9 1,5 0,5 0,002394 0,086711 0,15607979 7,95555556 4910,57914
3 1,84 0,7 0,5 0,002988 0,07761505 0,13970709 8,03888889 4560,12143

229
4 1,4 1,25 0,5 0,001188 0,12309149 0,22156468 8,04166667 6949,08416
2 0,9 1,5 1,3 0,002394 0,086711 0,15607979 8,04444444 4914,05378
2 0,9 1,5 1,6 0,002394 0,086711 0,15607979 8,07222222 4914,78516
3 1,4 1,25 0,5 0,001188 0,12309149 0,22156468 8,09166667 6961,59089
3 1,84 0,7 1,3 0,002988 0,07761505 0,13970709 8,10555556 4562,43607
3 1,84 0,7 1,6 0,002988 0,07761505 0,13970709 8,12777778 4562,86473
4 0,9 1,5 0,5 0,002988 0,07761505 0,13970709 8,17777778 4751,75024
2 1,4 1,25 0,5 0,001188 0,12309149 0,22156468 8,18611111 6974,27212
4 0,9 1,5 1,3 0,002988 0,07761505 0,13970709 8,25555556 4756,54444
4 1,4 1,25 1,3 0,001188 0,12309149 0,22156468 8,26944444 6993,70274
4 0,9 1,5 1,6 0,002988 0,07761505 0,13970709 8,27777778 4757,23395
3 1,4 1,25 1,3 0,001188 0,12309149 0,22156468 8,30833333 6999,52193
4 1,4 1,25 1,6 0,001188 0,12309149 0,22156468 8,33611111 7003,25793
3 0,9 1,5 0,5 0,002988 0,07761505 0,13970709 8,36944444 4769,29835
3 1,4 1,25 1,6 0,001188 0,12309149 0,22156468 8,375 7008,94283
2 1,4 1,25 1,3 0,001188 0,12309149 0,22156468 8,39166667 7005,80709
2 1,84 0,7 0,5 0,002988 0,07761505 0,13970709 8,40277778 4572,57321
3 0,9 1,5 1,3 0,002988 0,07761505 0,13970709 8,43888889 4771,65522
2 1,4 1,25 1,6 0,001188 0,12309149 0,22156468 8,45555556 7013,99007
2 1,84 0,7 1,3 0,002988 0,07761505 0,13970709 8,45555556 4571,77863
3 0,9 1,5 1,6 0,002988 0,07761505 0,13970709 8,45833333 4771,65745
2 1,84 0,7 1,6 0,002988 0,07761505 0,13970709 8,475 4571,80027
2 0,9 1,5 0,5 0,002988 0,07761505 0,13970709 8,74722222 4781,84169
2 0,9 1,5 1,3 0,002988 0,07761505 0,13970709 8,8 4780,80501
2 0,9 1,5 1,6 0,002988 0,07761505 0,13970709 8,81944444 4780,58098
4 1,4 1,5 0,5 0,001188 0,12309149 0,22156468 9,45555556 8258,17535
4 1,4 1,25 0,5 0,002394 0,086711 0,15607979 9,48611111 6264,13865
3 1,4 1,5 0,5 0,001188 0,12309149 0,22156468 9,51388889 8272,79346
4 1,4 1,25 1,3 0,002394 0,086711 0,15607979 9,6 6272,75807
2 1,4 1,5 0,5 0,001188 0,12309149 0,22156468 9,62777778 8288,12379
4 1,4 1,25 1,6 0,002394 0,086711 0,15607979 9,63333333 6273,8938
3 1,4 1,25 0,5 0,002394 0,086711 0,15607979 9,64722222 6279,39634
4 1,4 1,5 1,3 0,001188 0,12309149 0,22156468 9,69444444 8303,86193
3 1,4 1,5 1,3 0,001188 0,12309149 0,22156468 9,74166667 8311,04883
3 1,4 1,25 1,3 0,002394 0,086711 0,15607979 9,75 6284,12702
4 1,4 1,5 1,6 0,001188 0,12309149 0,22156468 9,76388889 8313,21971
3 1,4 1,25 1,6 0,002394 0,086711 0,15607979 9,78333333 6285,05994
3 1,4 1,5 1,6 0,001188 0,12309149 0,22156468 9,81111111 8319,69698
2 1,4 1,5 1,3 0,001188 0,12309149 0,22156468 9,83888889 8317,75922
2 1,4 1,5 1,6 0,001188 0,12309149 0,22156468 9,90555556 8325,19696
2 1,4 1,25 0,5 0,002394 0,086711 0,15607979 9,96944444 6290,28675
2 1,4 1,25 1,3 0,002394 0,086711 0,15607979 10,0527778 6290,47023
2 1,4 1,25 1,6 0,002394 0,086711 0,15607979 10,0777778 6289,95409

230
4 1,4 1,25 0,5 0,002988 0,07761505 0,13970709 10,2277778 6093,41045
4 1,84 1,25 0,5 0,001188 0,12309149 0,22156468 10,2555556 9004,16912
4 1,4 1,25 1,3 0,002988 0,07761505 0,13970709 10,3 6096,17448
3 1,84 1,25 0,5 0,001188 0,12309149 0,22156468 10,3194444 9020,21704
4 1,4 1,25 1,6 0,002988 0,07761505 0,13970709 10,3222222 6096,05825
2 1,84 1,25 0,5 0,001188 0,12309149 0,22156468 10,4444444 9037,23226
3 1,4 1,25 0,5 0,002988 0,07761505 0,13970709 10,4666667 6114,26025
4 1,84 1,25 1,3 0,001188 0,12309149 0,22156468 10,4972222 9049,7159
3 1,4 1,25 1,3 0,002988 0,07761505 0,13970709 10,5277778 6113,71451
3 1,4 1,25 1,6 0,002988 0,07761505 0,13970709 10,5472222 6113,03966
3 1,84 1,25 1,3 0,001188 0,12309149 0,22156468 10,55 9057,88272
4 1,84 1,25 1,6 0,001188 0,12309149 0,22156468 10,5694444 9058,89005
3 1,84 1,25 1,6 0,001188 0,12309149 0,22156468 10,6194444 9065,84701
2 1,84 1,25 1,3 0,001188 0,12309149 0,22156468 10,6555556 9064,91632
2 1,84 1,25 1,6 0,001188 0,12309149 0,22156468 10,7222222 9071,602
2 1,4 1,25 0,5 0,002988 0,07761505 0,13970709 10,9333333 6127,55713
2 1,4 1,25 1,3 0,002988 0,07761505 0,13970709 10,975 6122,72854
2 1,4 1,25 1,6 0,002988 0,07761505 0,13970709 10,9916667 6121,50923
4 1,4 1,5 0,5 0,002394 0,086711 0,15607979 11,1222222 7462,43613
4 1,4 1,5 1,3 0,002394 0,086711 0,15607979 11,2305556 7468,59522
4 1,4 1,5 1,6 0,002394 0,086711 0,15607979 11,2638889 7469,02378
3 1,4 1,5 0,5 0,002394 0,086711 0,15607979 11,3111111 7479,50027
3 1,4 1,5 1,3 0,002394 0,086711 0,15607979 11,4083333 7481,67481
3 1,4 1,5 1,6 0,002394 0,086711 0,15607979 11,4388889 7481,53538
2 1,4 1,5 0,5 0,002394 0,086711 0,15607979 11,6861111 7491,14228
2 1,4 1,5 1,3 0,002394 0,086711 0,15607979 11,7583333 7487,32971
2 1,4 1,5 1,6 0,002394 0,086711 0,15607979 11,7833333 7485,98325
4 1,4 1,5 0,5 0,002988 0,07761505 0,13970709 11,9722222 7261,24408
4 1,4 1,5 1,3 0,002988 0,07761505 0,13970709 12,0333333 7261,21851
4 1,84 1,25 0,5 0,002394 0,086711 0,15607979 12,0416667 8144,87025
4 1,4 1,5 1,6 0,002988 0,07761505 0,13970709 12,0555556 7260,43604
4 1,84 1,5 0,5 0,001188 0,12309149 0,22156468 12,0777778 10716,0106
4 1,84 1,25 1,3 0,002394 0,086711 0,15607979 12,1472222 8149,52749
3 1,84 1,5 0,5 0,001188 0,12309149 0,22156468 12,1555556 10735,4403
4 1,84 1,25 1,6 0,002394 0,086711 0,15607979 12,1805556 8149,62355
3 1,84 1,25 0,5 0,002394 0,086711 0,15607979 12,2472222 8163,34754
3 1,4 1,5 0,5 0,002988 0,07761505 0,13970709 12,25 7284,50148
3 1,4 1,5 1,3 0,002988 0,07761505 0,13970709 12,2972222 7280,38398
2 1,84 1,5 0,5 0,001188 0,12309149 0,22156468 12,3 10754,6024
3 1,4 1,5 1,6 0,002988 0,07761505 0,13970709 12,3166667 7279,1252
4 1,84 1,5 1,3 0,001188 0,12309149 0,22156468 12,3222222 10759,8732
3 1,84 1,25 1,3 0,002394 0,086711 0,15607979 12,3388889 8163,43481
3 1,84 1,25 1,6 0,002394 0,086711 0,15607979 12,3694444 8162,65375

231
3 1,84 1,5 1,3 0,001188 0,12309149 0,22156468 12,3861111 10769,7757
4 1,84 1,5 1,6 0,001188 0,12309149 0,22156468 12,3972222 10768,404
3 1,84 1,5 1,6 0,001188 0,12309149 0,22156468 12,4583333 10776,5384
2 1,84 1,5 1,3 0,001188 0,12309149 0,22156468 12,5111111 10778,1158
2 1,84 1,5 1,6 0,001188 0,12309149 0,22156468 12,5805556 10783,3378
2 1,84 1,25 0,5 0,002394 0,086711 0,15607979 12,65 8174,76938
2 1,84 1,25 1,3 0,002394 0,086711 0,15607979 12,7166667 8168,67644
2 1,84 1,25 1,6 0,002394 0,086711 0,15607979 12,7388889 8166,55374
2 1,4 1,5 0,5 0,002988 0,07761505 0,13970709 12,7888889 7297,52668
2 1,4 1,5 1,3 0,002988 0,07761505 0,13970709 12,8166667 7288,82411
2 1,4 1,5 1,6 0,002988 0,07761505 0,13970709 12,8277778 7286,20058
4 1,84 1,25 0,5 0,002988 0,07761505 0,13970709 12,95 7926,03738
4 1,84 1,25 1,3 0,002988 0,07761505 0,13970709 13,0055556 7924,42165
4 1,84 1,25 1,6 0,002988 0,07761505 0,13970709 13,025 7923,08943
3 1,84 1,25 0,5 0,002988 0,07761505 0,13970709 13,25 7950,77035
3 1,84 1,25 1,3 0,002988 0,07761505 0,13970709 13,2888889 7944,44086
3 1,84 1,25 1,6 0,002988 0,07761505 0,13970709 13,3055556 7942,404
2 1,84 1,25 0,5 0,002988 0,07761505 0,13970709 13,8305556 7963,49843
2 1,84 1,25 1,3 0,002988 0,07761505 0,13970709 13,8472222 7952,08359
2 1,84 1,25 1,6 0,002988 0,07761505 0,13970709 13,8583333 7948,87702
4 1,84 1,5 0,5 0,002394 0,086711 0,15607979 14,1222222 9709,04488
4 1,84 1,5 1,3 0,002394 0,086711 0,15607979 14,2138889 9709,58873
4 1,84 1,5 1,6 0,002394 0,086711 0,15607979 14,2472222 9708,5047
3 1,84 1,5 0,5 0,002394 0,086711 0,15607979 14,3666667 9730,64245
3 1,84 1,5 1,3 0,002394 0,086711 0,15607979 14,4388889 9725,07053
3 1,84 1,5 1,6 0,002394 0,086711 0,15607979 14,4666667 9722,76414
2 1,84 1,5 0,5 0,002394 0,086711 0,15607979 14,8333333 9741,86014
2 1,84 1,5 1,3 0,002394 0,086711 0,15607979 14,8777778 9729,10751
2 1,84 1,5 1,6 0,002394 0,086711 0,15607979 14,8972222 9725,27614
4 1,84 1,5 0,5 0,002988 0,07761505 0,13970709 15,1583333 9449,59214
4 1,84 1,5 1,3 0,002988 0,07761505 0,13970709 15,1944444 9443,34098
4 1,84 1,5 1,6 0,002988 0,07761505 0,13970709 15,2111111 9440,69964
3 1,84 1,5 0,5 0,002988 0,07761505 0,13970709 15,5055556 9476,69797
3 1,84 1,5 1,3 0,002988 0,07761505 0,13970709 15,525 9465,23566
3 1,84 1,5 1,6 0,002988 0,07761505 0,13970709 15,5361111 9461,39965
2 1,84 1,5 1,3 0,002988 0,07761505 0,13970709 16,1666667 9470,034
2 1,84 1,5 1,6 0,002988 0,07761505 0,13970709 16,1722222 9465,10446
2 1,84 1,5 0,5 0,002988 0,07761505 0,13970709 16,175 9487,94705

232
Appendix A2

Second optimization approach, charging of the unit, airflow rate of 500 m3/h

lmpc rmcp cpmcp lalu dym dx dz time energy

2 0,9 0,7 0,5 0,001188 0,1 0,27272727 2,77777778 2652,65696
2 0,9 0,7 1,3 0,001188 0,1 0,27272727 2,77777778 2650,77826
2 0,9 0,7 1,6 0,001188 0,1 0,27272727 2,77777778 2649,91746
3 0,9 0,7 0,5 0,001188 0,1 0,27272727 2,77777778 2657,71478
3 0,9 0,7 1,3 0,001188 0,1 0,27272727 2,77777778 2657,57411
3 0,9 0,7 1,6 0,001188 0,1 0,27272727 2,77777778 2656,69384
4 0,9 0,7 0,5 0,001188 0,1 0,27272727 2,77777778 2657,55059
4 0,9 0,7 1,3 0,001188 0,1 0,27272727 2,77777778 2659,44246
4 0,9 0,7 1,6 0,001188 0,1 0,27272727 2,77777778 2659,19077
4 0,9 0,7 0,5 0,002394 0,1 0,13533835 2,84444444 2297,15087
4 0,9 0,7 1,3 0,002394 0,1 0,13533835 2,87777778 2302,05838
4 0,9 0,7 1,6 0,002394 0,1 0,13533835 2,88611111 2302,69499
3 0,9 0,7 0,5 0,002394 0,1 0,13533835 2,91388889 2298,39692
3 0,9 0,7 1,3 0,002394 0,1 0,13533835 2,94166667 2301,85971
3 0,9 0,7 1,6 0,002394 0,1 0,13533835 2,95277778 2303,30612
2 0,9 0,7 0,5 0,002394 0,1 0,13533835 3,05 2300,30655
2 0,9 0,7 1,3 0,002394 0,1 0,13533835 3,07777778 2302,91553
2 0,9 0,7 1,6 0,002394 0,1 0,13533835 3,08611111 2303,4191
4 0,9 0,7 0,5 0,002988 0,1 0,10843373 3,15555556 2241,58036
4 0,9 0,7 1,3 0,002988 0,1 0,10843373 3,175 2244,4303
4 0,9 0,7 1,6 0,002988 0,1 0,10843373 3,18055556 2244,94781
3 0,9 0,7 0,5 0,002988 0,1 0,10843373 3,25833333 2242,83665
3 0,9 0,7 1,3 0,002988 0,1 0,10843373 3,275 2244,48023
3 0,9 0,7 1,6 0,002988 0,1 0,10843373 3,28055556 2244,86963
4 1,4 0,7 0,5 0,001188 0,1 0,27272727 3,29166667 3798,73377
3 1,4 0,7 0,5 0,001188 0,1 0,27272727 3,31944444 3803,74429
2 1,4 0,7 0,5 0,001188 0,1 0,27272727 3,37222222 3808,03264
4 1,4 0,7 1,3 0,001188 0,1 0,27272727 3,40833333 3826,35389
3 1,4 0,7 1,3 0,001188 0,1 0,27272727 3,43333333 3828,70642
4 1,4 0,7 1,6 0,001188 0,1 0,27272727 3,44166667 3832,02752
2 0,9 0,7 0,5 0,002988 0,1 0,10843373 3,46111111 2244,93795
3 1,4 0,7 1,6 0,001188 0,1 0,27272727 3,46388889 3833,6793
2 0,9 0,7 1,3 0,002988 0,1 0,10843373 3,47777778 2246,14602
2 0,9 0,7 1,6 0,002988 0,1 0,10843373 3,48055556 2246,06458
2 1,4 0,7 1,3 0,001188 0,1 0,27272727 3,48055556 3829,04145
2 1,4 0,7 1,6 0,001188 0,1 0,27272727 3,51388889 3834,77381

233
4 0,9 1,25 0,5 0,001188 0,1 0,27272727 3,69722222 4312,92006
3 0,9 1,25 0,5 0,001188 0,1 0,27272727 3,73055556 4319,58132
2 0,9 1,25 0,5 0,001188 0,1 0,27272727 3,79166667 4324,47423
4 0,9 1,25 1,3 0,001188 0,1 0,27272727 3,82222222 4342,86512
3 0,9 1,25 1,3 0,001188 0,1 0,27272727 3,85 4345,41077
4 0,9 1,25 1,6 0,001188 0,1 0,27272727 3,85833333 4349,61005
3 0,9 1,25 1,6 0,001188 0,1 0,27272727 3,88611111 4352,13607
2 0,9 1,25 1,3 0,001188 0,1 0,27272727 3,90555556 4345,99048
2 0,9 1,25 1,6 0,001188 0,1 0,27272727 3,94166667 4352,15727
4 1,4 0,7 0,5 0,002394 0,1 0,13533835 4,13888889 3474,97549
4 1,84 0,7 0,5 0,001188 0,1 0,27272727 4,15 4890,49067
4 1,4 0,7 1,3 0,002394 0,1 0,13533835 4,17777778 3481,12649
3 1,84 0,7 0,5 0,001188 0,1 0,27272727 4,18611111 4897,60958
4 1,4 0,7 1,6 0,002394 0,1 0,13533835 4,18888889 3482,31804
3 1,4 0,7 0,5 0,002394 0,1 0,13533835 4,23888889 3476,24521
2 1,84 0,7 0,5 0,001188 0,1 0,27272727 4,25555556 4903,18792
3 1,4 0,7 1,3 0,002394 0,1 0,13533835 4,27222222 3480,39153
3 1,4 0,7 1,6 0,002394 0,1 0,13533835 4,28333333 3481,26915
4 1,84 0,7 1,3 0,001188 0,1 0,27272727 4,28333333 4922,40357
3 1,84 0,7 1,3 0,001188 0,1 0,27272727 4,31666667 4926,15827
4 0,9 1,5 0,5 0,001188 0,1 0,27272727 4,32222222 5110,03849
4 1,84 0,7 1,6 0,001188 0,1 0,27272727 4,32222222 4930,08034
3 1,84 0,7 1,6 0,001188 0,1 0,27272727 4,35277778 4931,68595
3 0,9 1,5 0,5 0,001188 0,1 0,27272727 4,35833333 5117,14647
2 1,84 0,7 1,3 0,001188 0,1 0,27272727 4,37777778 4926,19369
2 1,84 0,7 1,6 0,001188 0,1 0,27272727 4,41666667 4932,78832
2 0,9 1,5 0,5 0,001188 0,1 0,27272727 4,43333333 5123,80573
2 1,4 0,7 0,5 0,002394 0,1 0,13533835 4,44166667 3477,88567
4 0,9 1,5 1,3 0,001188 0,1 0,27272727 4,45833333 5142,90452
2 1,4 0,7 1,3 0,002394 0,1 0,13533835 4,46944444 3480,17953
2 1,4 0,7 1,6 0,002394 0,1 0,13533835 4,47777778 3480,45039
3 0,9 1,5 1,3 0,001188 0,1 0,27272727 4,49166667 5146,16609
4 0,9 1,5 1,6 0,001188 0,1 0,27272727 4,49722222 5150,07163
3 0,9 1,5 1,6 0,001188 0,1 0,27272727 4,52777778 5151,71636
2 0,9 1,5 1,3 0,001188 0,1 0,27272727 4,55555556 5146,38655
4 1,4 0,7 0,5 0,002988 0,1 0,10843373 4,59166667 3400,24208
2 0,9 1,5 1,6 0,001188 0,1 0,27272727 4,59444444 5152,56389
4 1,4 0,7 1,3 0,002988 0,1 0,10843373 4,61388889 3403,67686
4 1,4 0,7 1,6 0,002988 0,1 0,10843373 4,61944444 3404,11737
4 0,9 1,25 0,5 0,002394 0,1 0,13533835 4,65833333 3960,72893
4 0,9 1,25 1,3 0,002394 0,1 0,13533835 4,69722222 3967,05781
4 0,9 1,25 1,6 0,002394 0,1 0,13533835 4,70833333 3967,88735
3 1,4 0,7 0,5 0,002988 0,1 0,10843373 4,73888889 3400,71996

234
3 1,4 0,7 1,3 0,002988 0,1 0,10843373 4,75833333 3402,6233
3 1,4 0,7 1,6 0,002988 0,1 0,10843373 4,76388889 3403,0023
3 0,9 1,25 0,5 0,002394 0,1 0,13533835 4,77222222 3962,05683
3 0,9 1,25 1,3 0,002394 0,1 0,13533835 4,80555556 3965,9977
3 0,9 1,25 1,6 0,002394 0,1 0,13533835 4,81666667 3967,04463
2 0,9 1,25 0,5 0,002394 0,1 0,13533835 5 3963,38216
2 0,9 1,25 1,3 0,002394 0,1 0,13533835 5,02777778 3965,25705
2 0,9 1,25 1,6 0,002394 0,1 0,13533835 5,03611111 3965,32725
2 1,4 0,7 0,5 0,002988 0,1 0,10843373 5,03611111 3402,28399
2 1,4 0,7 1,3 0,002988 0,1 0,10843373 5,05 3402,62404
2 1,4 0,7 1,6 0,002988 0,1 0,10843373 5,05277778 3402,28419
4 0,9 1,25 0,5 0,002988 0,1 0,10843373 5,16666667 3878,1797
4 0,9 1,25 1,3 0,002988 0,1 0,10843373 5,18888889 3881,58296
4 0,9 1,25 1,6 0,002988 0,1 0,10843373 5,19444444 3882,03391
4 1,84 0,7 0,5 0,002394 0,1 0,13533835 5,23333333 4505,68973
4 1,84 0,7 1,3 0,002394 0,1 0,13533835 5,27222222 4511,97751
4 1,84 0,7 1,6 0,002394 0,1 0,13533835 5,28611111 4513,07564
3 0,9 1,25 0,5 0,002988 0,1 0,10843373 5,33333333 3878,13632
3 0,9 1,25 1,3 0,002988 0,1 0,10843373 5,35 3879,84163
3 0,9 1,25 1,6 0,002988 0,1 0,10843373 5,35833333 3880,33085
3 1,84 0,7 0,5 0,002394 0,1 0,13533835 5,36111111 4506,59043
3 1,84 0,7 1,3 0,002394 0,1 0,13533835 5,39722222 4511,06983
3 1,84 0,7 1,6 0,002394 0,1 0,13533835 5,40833333 4511,83396
4 1,4 1,25 0,5 0,001188 0,1 0,27272727 5,40833333 6521,70627
4 0,9 1,5 0,5 0,002394 0,1 0,13533835 5,45 4712,47137
3 1,4 1,25 0,5 0,001188 0,1 0,27272727 5,45555556 6530,50169
4 0,9 1,5 1,3 0,002394 0,1 0,13533835 5,48888889 4719,05417
4 0,9 1,5 1,6 0,002394 0,1 0,13533835 5,50277778 4720,12432
2 1,4 1,25 0,5 0,001188 0,1 0,27272727 5,55 6538,35181
4 1,4 1,25 1,3 0,001188 0,1 0,27272727 5,56111111 6558,92123
3 0,9 1,5 0,5 0,002394 0,1 0,13533835 5,58333333 4713,64829
3 1,4 1,25 1,3 0,001188 0,1 0,27272727 5,60277778 6562,29054
4 1,4 1,25 1,6 0,001188 0,1 0,27272727 5,60555556 6566,55232
2 1,84 0,7 0,5 0,002394 0,1 0,13533835 5,61388889 4507,22906
3 0,9 1,5 1,3 0,002394 0,1 0,13533835 5,61666667 4717,45088
3 0,9 1,5 1,6 0,002394 0,1 0,13533835 5,62777778 4718,43615
2 1,84 0,7 1,3 0,002394 0,1 0,13533835 5,64166667 4508,66433
3 1,4 1,25 1,6 0,001188 0,1 0,27272727 5,64722222 6569,80242
2 1,84 0,7 1,6 0,002394 0,1 0,13533835 5,65 4508,84129
2 0,9 1,25 0,5 0,002988 0,1 0,10843373 5,66388889 3878,82182
2 0,9 1,25 1,3 0,002988 0,1 0,10843373 5,675 3878,65161
2 0,9 1,25 1,6 0,002988 0,1 0,10843373 5,68055556 3878,58602
2 1,4 1,25 1,3 0,001188 0,1 0,27272727 5,68611111 6563,2407

235
2 1,4 1,25 1,6 0,001188 0,1 0,27272727 5,72777778 6569,45829
4 1,84 0,7 0,5 0,002988 0,1 0,10843373 5,8 4413,60578
4 1,84 0,7 1,3 0,002988 0,1 0,10843373 5,82222222 4417,07301
4 1,84 0,7 1,6 0,002988 0,1 0,10843373 5,82777778 4417,44077
2 0,9 1,5 0,5 0,002394 0,1 0,13533835 5,84722222 4713,92802
2 0,9 1,5 1,3 0,002394 0,1 0,13533835 5,875 4715,47289
2 0,9 1,5 1,6 0,002394 0,1 0,13533835 5,88333333 4715,44348
3 1,84 0,7 0,5 0,002988 0,1 0,10843373 5,98888889 4413,44952
3 1,84 0,7 1,3 0,002988 0,1 0,10843373 6,00555556 4415,09335
3 1,84 0,7 1,6 0,002988 0,1 0,10843373 6,01111111 4415,16633
4 0,9 1,5 0,5 0,002988 0,1 0,10843373 6,03888889 4617,09293
4 0,9 1,5 1,3 0,002988 0,1 0,10843373 6,06111111 4620,65591
4 0,9 1,5 1,6 0,002988 0,1 0,10843373 6,06666667 4620,86291
3 0,9 1,5 0,5 0,002988 0,1 0,10843373 6,23333333 4616,56781
3 0,9 1,5 1,3 0,002988 0,1 0,10843373 6,25 4618,01217
3 0,9 1,5 1,6 0,002988 0,1 0,10843373 6,25555556 4617,99126
4 1,4 1,5 0,5 0,001188 0,1 0,27272727 6,34166667 7752,31255
2 1,84 0,7 0,5 0,002988 0,1 0,10843373 6,35555556 4412,71416
2 1,84 0,7 1,3 0,002988 0,1 0,10843373 6,36666667 4412,25405
2 1,84 0,7 1,6 0,002988 0,1 0,10843373 6,37222222 4412,13227
3 1,4 1,5 0,5 0,001188 0,1 0,27272727 6,39722222 7763,03201
4 1,4 1,5 1,3 0,001188 0,1 0,27272727 6,50555556 7792,28819
2 1,4 1,5 0,5 0,001188 0,1 0,27272727 6,50833333 7771,92779
3 1,4 1,5 1,3 0,001188 0,1 0,27272727 6,55555556 7796,81354
4 1,4 1,5 1,6 0,001188 0,1 0,27272727 6,55555556 7801,16673
3 1,4 1,5 1,6 0,001188 0,1 0,27272727 6,60277778 7804,08651
2 0,9 1,5 0,5 0,002988 0,1 0,10843373 6,61666667 4615,48864
2 0,9 1,5 1,3 0,002988 0,1 0,10843373 6,625 4614,69971
2 0,9 1,5 1,6 0,002988 0,1 0,10843373 6,63055556 4614,47557
2 1,4 1,5 1,3 0,001188 0,1 0,27272727 6,65555556 7798,07605
2 1,4 1,5 1,6 0,001188 0,1 0,27272727 6,69722222 7802,81378
4 1,4 1,25 0,5 0,002394 0,1 0,13533835 6,81944444 6043,41988
4 1,4 1,25 1,3 0,002394 0,1 0,13533835 6,86111111 6050,42461
4 1,84 1,25 0,5 0,001188 0,1 0,27272727 6,86944444 8454,55952
4 1,4 1,25 1,6 0,002394 0,1 0,13533835 6,87222222 6051,42002
3 1,84 1,25 0,5 0,001188 0,1 0,27272727 6,93055556 8466,1779
3 1,4 1,25 0,5 0,002394 0,1 0,13533835 6,98611111 6044,05921
3 1,4 1,25 1,3 0,002394 0,1 0,13533835 7,01944444 6047,81139
3 1,4 1,25 1,6 0,002394 0,1 0,13533835 7,03055556 6048,15874
4 1,84 1,25 1,3 0,001188 0,1 0,27272727 7,03888889 8495,65541
2 1,84 1,25 0,5 0,001188 0,1 0,27272727 7,05 8475,37019
4 1,84 1,25 1,6 0,001188 0,1 0,27272727 7,08888889 8503,95478
3 1,84 1,25 1,3 0,001188 0,1 0,27272727 7,09166667 8500,10212

236
3 1,84 1,25 1,6 0,001188 0,1 0,27272727 7,14166667 8507,72452
2 1,84 1,25 1,3 0,001188 0,1 0,27272727 7,2 8501,14332
2 1,84 1,25 1,6 0,001188 0,1 0,27272727 7,24444444 8506,72323
2 1,4 1,25 0,5 0,002394 0,1 0,13533835 7,31111111 6042,03562
2 1,4 1,25 1,3 0,002394 0,1 0,13533835 7,33611111 6042,4716
2 1,4 1,25 1,6 0,002394 0,1 0,13533835 7,34444444 6042,28252
4 1,4 1,25 0,5 0,002988 0,1 0,10843373 7,54722222 5925,35247
4 1,4 1,25 1,3 0,002988 0,1 0,10843373 7,56388889 5928,17518
4 1,4 1,25 1,6 0,002988 0,1 0,10843373 7,56944444 5928,21262
3 1,4 1,25 0,5 0,002988 0,1 0,10843373 7,78611111 5922,78945
3 1,4 1,25 1,3 0,002988 0,1 0,10843373 7,79722222 5923,29805
3 1,4 1,25 1,6 0,002988 0,1 0,10843373 7,80277778 5923,1298
4 1,4 1,5 0,5 0,002394 0,1 0,13533835 7,98888889 7203,8837
4 1,4 1,5 1,3 0,002394 0,1 0,13533835 8,02777778 7210,59919
4 1,4 1,5 1,6 0,002394 0,1 0,13533835 8,03888889 7211,25557
4 1,84 1,5 0,5 0,001188 0,1 0,27272727 8,06666667 10064,9833
3 1,84 1,5 0,5 0,001188 0,1 0,27272727 8,13888889 10078,2364
3 1,4 1,5 0,5 0,002394 0,1 0,13533835 8,18055556 7203,24892
3 1,4 1,5 1,3 0,002394 0,1 0,13533835 8,21111111 7206,25418
3 1,4 1,5 1,6 0,002394 0,1 0,13533835 8,21944444 7205,84496
4 1,84 1,5 1,3 0,001188 0,1 0,27272727 8,24444444 10107,774
2 1,4 1,25 0,5 0,002988 0,1 0,10843373 8,25555556 5918,01625
2 1,4 1,25 1,3 0,002988 0,1 0,10843373 8,25833333 5915,73063
2 1,4 1,25 1,6 0,002988 0,1 0,10843373 8,26111111 5915,00691
2 1,84 1,5 0,5 0,001188 0,1 0,27272727 8,28055556 10089,3286
4 1,84 1,5 1,6 0,001188 0,1 0,27272727 8,29722222 10115,8437
3 1,84 1,5 1,3 0,001188 0,1 0,27272727 8,30833333 10113,1569
3 1,84 1,5 1,6 0,001188 0,1 0,27272727 8,35833333 10119,6491
2 1,84 1,5 1,3 0,001188 0,1 0,27272727 8,43333333 10113,3529
2 1,84 1,5 1,6 0,001188 0,1 0,27272727 8,48055556 10118,6898
2 1,4 1,5 0,5 0,002394 0,1 0,13533835 8,55555556 7198,58705
2 1,4 1,5 1,3 0,002394 0,1 0,13533835 8,575 7197,54877
2 1,4 1,5 1,6 0,002394 0,1 0,13533835 8,58333333 7196,94797
4 1,84 1,25 0,5 0,002394 0,1 0,13533835 8,64444444 7864,92581
4 1,84 1,25 1,3 0,002394 0,1 0,13533835 8,68055556 7871,16873
4 1,84 1,25 1,6 0,002394 0,1 0,13533835 8,69166667 7871,52753
4 1,4 1,5 0,5 0,002988 0,1 0,10843373 8,82777778 7064,48212
4 1,4 1,5 1,3 0,002988 0,1 0,10843373 8,83888889 7066,86546
4 1,4 1,5 1,6 0,002988 0,1 0,10843373 8,84444444 7066,84694
3 1,84 1,25 0,5 0,002394 0,1 0,13533835 8,85 7863,6224
3 1,84 1,25 1,3 0,002394 0,1 0,13533835 8,87777778 7865,94696
3 1,84 1,25 1,6 0,002394 0,1 0,13533835 8,88888889 7865,98675
3 1,4 1,5 0,5 0,002988 0,1 0,10843373 9,10277778 7059,89598

237
3 1,4 1,5 1,3 0,002988 0,1 0,10843373 9,10833333 7059,47154
3 1,4 1,5 1,6 0,002988 0,1 0,10843373 9,11111111 7058,64561
2 1,84 1,25 0,5 0,002394 0,1 0,13533835 9,25555556 7857,63485
2 1,84 1,25 1,3 0,002394 0,1 0,13533835 9,27222222 7855,74676
2 1,84 1,25 1,6 0,002394 0,1 0,13533835 9,27777778 7854,5519
4 1,84 1,25 0,5 0,002988 0,1 0,10843373 9,54444444 7713,58268
4 1,84 1,25 1,3 0,002988 0,1 0,10843373 9,55277778 7715,4173
4 1,84 1,25 1,6 0,002988 0,1 0,10843373 9,55833333 7715,18256
2 1,4 1,5 0,5 0,002988 0,1 0,10843373 9,64166667 7050,93943
2 1,4 1,5 1,3 0,002988 0,1 0,10843373 9,64166667 7047,6122
2 1,4 1,5 1,6 0,002988 0,1 0,10843373 9,64166667 7046,14729
3 1,84 1,25 0,5 0,002988 0,1 0,10843373 9,83888889 7707,42289
3 1,84 1,25 1,3 0,002988 0,1 0,10843373 9,84444444 7706,71445
3 1,84 1,25 1,6 0,002988 0,1 0,10843373 9,84444444 7705,33047
4 1,84 1,5 0,5 0,002394 0,1 0,13533835 10,1277778 9381,25959
4 1,84 1,5 1,3 0,002394 0,1 0,13533835 10,1583333 9386,92146
4 1,84 1,5 1,6 0,002394 0,1 0,13533835 10,1694444 9386,90955
3 1,84 1,5 0,5 0,002394 0,1 0,13533835 10,3666667 9378,49963
3 1,84 1,5 1,3 0,002394 0,1 0,13533835 10,3861111 9379,08604
3 1,84 1,5 1,6 0,002394 0,1 0,13533835 10,3944444 9378,18158
2 1,84 1,25 1,6 0,002988 0,1 0,10843373 10,4111111 7689,65971
2 1,84 1,25 1,3 0,002988 0,1 0,10843373 10,4138889 7691,59712
2 1,84 1,25 0,5 0,002988 0,1 0,10843373 10,4194444 7696,07868
2 1,84 1,5 0,5 0,002394 0,1 0,13533835 10,8333333 9368,2032
2 1,84 1,5 1,3 0,002394 0,1 0,13533835 10,8416667 9363,94453
2 1,84 1,5 1,6 0,002394 0,1 0,13533835 10,8444444 9361,94171
4 1,84 1,5 0,5 0,002988 0,1 0,10843373 11,1611111 9201,27505
4 1,84 1,5 1,3 0,002988 0,1 0,10843373 11,1638889 9202,27661
4 1,84 1,5 1,6 0,002988 0,1 0,10843373 11,1638889 9201,11213
3 1,84 1,5 1,6 0,002988 0,1 0,10843373 11,4944444 9187,2467
3 1,84 1,5 1,3 0,002988 0,1 0,10843373 11,4972222 9189,28268
3 1,84 1,5 0,5 0,002988 0,1 0,10843373 11,5027778 9191,96386
2 1,84 1,5 1,6 0,002988 0,1 0,10843373 12,1472222 9163,99594
2 1,84 1,5 1,3 0,002988 0,1 0,10843373 12,15 9166,30838
2 1,84 1,5 0,5 0,002988 0,1 0,10843373 12,1638889 9172,94647

238
Appendix A3

Second optimization approach, discharging of the unit, airflow rate of 300m3/h

lmpc rmcp cpmcp lalu dym dx dz time energy

4 0,9 0,7 0,5 0,001188 0,12309149 0,22156468 3,91944444 -2974,7809
3 0,9 0,7 0,5 0,001188 0,12309149 0,22156468 3,94166667 -2981,09273
2 0,9 0,7 0,5 0,001188 0,12309149 0,22156468 3,99444444 -2991,65713
4 0,9 0,7 1,3 0,001188 0,12309149 0,22156468 4,09166667 -2964,45559
3 0,9 0,7 1,3 0,001188 0,12309149 0,22156468 4,11388889 -2968,85148
4 0,9 0,7 1,6 0,001188 0,12309149 0,22156468 4,14166667 -2960,38531
2 0,9 0,7 1,3 0,001188 0,12309149 0,22156468 4,16111111 -2977,55192
3 0,9 0,7 1,6 0,001188 0,12309149 0,22156468 4,16388889 -2964,68235
2 0,9 0,7 1,6 0,001188 0,12309149 0,22156468 4,21111111 -2972,31006
4 0,9 0,7 0,5 0,002394 0,086711 0,15607979 4,88888889 -3063,37966
3 0,9 0,7 0,5 0,002394 0,086711 0,15607979 4,975 -3071,34117
4 0,9 0,7 1,3 0,002394 0,086711 0,15607979 4,98611111 -3054,64353
4 0,9 0,7 1,6 0,002394 0,086711 0,15607979 5,00833333 -3051,57547
3 0,9 0,7 1,3 0,002394 0,086711 0,15607979 5,06666667 -3062,88894
3 0,9 0,7 1,6 0,002394 0,086711 0,15607979 5,09444444 -3059,55128
2 0,9 0,7 0,5 0,002394 0,086711 0,15607979 5,15555556 -3089,7016
2 0,9 0,7 1,3 0,002394 0,086711 0,15607979 5,23888889 -3081,29751
2 0,9 0,7 1,6 0,002394 0,086711 0,15607979 5,26666667 -3078,45007
4 0,9 0,7 0,5 0,002988 0,07761505 0,13970709 5,40555556 -3125,2794
4 0,9 0,7 1,3 0,002988 0,07761505 0,13970709 5,45 -3091,73517
4 0,9 0,7 1,6 0,002988 0,07761505 0,13970709 5,46944444 -3086,25521
3 0,9 0,7 0,5 0,002988 0,07761505 0,13970709 5,53333333 -3129,92443
3 0,9 0,7 1,3 0,002988 0,07761505 0,13970709 5,6 -3122,83928
3 0,9 0,7 1,6 0,002988 0,07761505 0,13970709 5,61666667 -3118,31445
4 1,4 0,7 0,5 0,001188 0,12309149 0,22156468 5,68055556 -4548,40605
3 1,4 0,7 0,5 0,001188 0,12309149 0,22156468 5,71944444 -4559,31239
2 0,9 0,7 0,5 0,002988 0,07761505 0,13970709 5,78611111 -3138,12985
2 1,4 0,7 0,5 0,001188 0,12309149 0,22156468 5,79722222 -4573,52334
2 0,9 0,7 1,3 0,002988 0,07761505 0,13970709 5,84166667 -3131,16335
2 0,9 0,7 1,6 0,002988 0,07761505 0,13970709 5,85833333 -3128,66041
4 1,4 0,7 1,3 0,001188 0,12309149 0,22156468 5,89166667 -4522,19295
3 1,4 0,7 1,3 0,001188 0,12309149 0,22156468 5,92777778 -4530,54487
4 1,4 0,7 1,6 0,001188 0,12309149 0,22156468 5,95555556 -4515,54206
3 1,4 0,7 1,6 0,001188 0,12309149 0,22156468 6,00277778 -4535,74419
2 1,4 0,7 1,3 0,001188 0,12309149 0,22156468 6,01111111 -4555,77202
2 1,4 0,7 1,6 0,001188 0,12309149 0,22156468 6,07222222 -4548,59365

239
4 0,9 1,25 0,5 0,001188 0,12309149 0,22156468 6,39444444 -5201,36131
3 0,9 1,25 0,5 0,001188 0,12309149 0,22156468 6,43888889 -5214,16343
2 0,9 1,25 0,5 0,001188 0,12309149 0,22156468 6,52777778 -5231,34885
4 0,9 1,25 1,3 0,001188 0,12309149 0,22156468 6,63333333 -5188,97068
3 0,9 1,25 1,3 0,001188 0,12309149 0,22156468 6,675 -5196,14476
4 0,9 1,25 1,6 0,001188 0,12309149 0,22156468 6,69166667 -5164,94176
3 0,9 1,25 1,6 0,001188 0,12309149 0,22156468 6,74444444 -5188,46104
2 0,9 1,25 1,3 0,001188 0,12309149 0,22156468 6,75555556 -5211,4364
2 0,9 1,25 1,6 0,001188 0,12309149 0,22156468 6,825 -5202,97265
4 1,4 0,7 0,5 0,002394 0,086711 0,15607979 7,14444444 -4730,36923
4 1,84 0,7 0,5 0,001188 0,12309149 0,22156468 7,18611111 -5936,63692
3 1,84 0,7 0,5 0,001188 0,12309149 0,22156468 7,23611111 -5949,97794
4 1,4 0,7 1,3 0,002394 0,086711 0,15607979 7,25555556 -4717,87529
3 1,4 0,7 0,5 0,002394 0,086711 0,15607979 7,275 -4743,56754
4 1,4 0,7 1,6 0,002394 0,086711 0,15607979 7,28611111 -4712,8617
2 1,84 0,7 0,5 0,001188 0,12309149 0,22156468 7,33888889 -5971,24182
3 1,4 0,7 1,3 0,002394 0,086711 0,15607979 7,38055556 -4731,32377
3 1,4 0,7 1,6 0,002394 0,086711 0,15607979 7,41388889 -4726,48666
4 1,84 0,7 1,3 0,001188 0,12309149 0,22156468 7,43055556 -5904,35165
4 0,9 1,5 0,5 0,001188 0,12309149 0,22156468 7,48333333 -6216,60601
3 1,84 0,7 1,3 0,001188 0,12309149 0,22156468 7,48888889 -5931,94794
4 1,84 0,7 1,6 0,001188 0,12309149 0,22156468 7,51944444 -5914,71331
3 0,9 1,5 0,5 0,001188 0,12309149 0,22156468 7,53888889 -6230,61353
2 1,4 0,7 0,5 0,002394 0,086711 0,15607979 7,54166667 -4772,83479
3 1,84 0,7 1,6 0,001188 0,12309149 0,22156468 7,56388889 -5923,32
2 1,84 0,7 1,3 0,001188 0,12309149 0,22156468 7,58333333 -5948,42244
2 1,4 0,7 1,3 0,002394 0,086711 0,15607979 7,63888889 -4760,89583
2 0,9 1,5 0,5 0,001188 0,12309149 0,22156468 7,64722222 -6252,99955
2 1,84 0,7 1,6 0,001188 0,12309149 0,22156468 7,65555556 -5940,15956
2 1,4 0,7 1,6 0,002394 0,086711 0,15607979 7,66944444 -4755,96669
4 0,9 1,5 1,3 0,001188 0,12309149 0,22156468 7,75 -6202,317
3 0,9 1,5 1,3 0,001188 0,12309149 0,22156468 7,79722222 -6211,8767
4 0,9 1,5 1,6 0,001188 0,12309149 0,22156468 7,82777778 -6194,0755
3 0,9 1,5 1,6 0,001188 0,12309149 0,22156468 7,87222222 -6203,2592
2 0,9 1,5 1,3 0,001188 0,12309149 0,22156468 7,89444444 -6229,38989
4 1,4 0,7 0,5 0,002988 0,07761505 0,13970709 7,90833333 -4834,69264
4 1,4 0,7 1,3 0,002988 0,07761505 0,13970709 7,93888889 -4782,83947
4 1,4 0,7 1,6 0,002988 0,07761505 0,13970709 7,95555556 -4775,65031
2 0,9 1,5 1,6 0,001188 0,12309149 0,22156468 7,96944444 -6220,29034
4 0,9 1,25 0,5 0,002394 0,086711 0,15607979 8,05 -5421,64828
3 1,4 0,7 0,5 0,002988 0,07761505 0,13970709 8,09444444 -4843,13282
3 1,4 0,7 1,3 0,002988 0,07761505 0,13970709 8,16111111 -4833,24914
4 0,9 1,25 1,3 0,002394 0,086711 0,15607979 8,16388889 -5407,43367

240
3 1,4 0,7 1,6 0,002988 0,07761505 0,13970709 8,17777778 -4826,20947
3 0,9 1,25 0,5 0,002394 0,086711 0,15607979 8,2 -5436,75244
4 0,9 1,25 1,6 0,002394 0,086711 0,15607979 8,2 -5401,73498
3 0,9 1,25 1,3 0,002394 0,086711 0,15607979 8,30833333 -5423,22737
3 0,9 1,25 1,6 0,002394 0,086711 0,15607979 8,34166667 -5417,9278
2 1,4 0,7 0,5 0,002988 0,07761505 0,13970709 8,46388889 -4856,63355
2 0,9 1,25 0,5 0,002394 0,086711 0,15607979 8,49722222 -5470,91324
2 1,4 0,7 1,3 0,002988 0,07761505 0,13970709 8,52222222 -4846,6232
2 1,4 0,7 1,6 0,002988 0,07761505 0,13970709 8,54166667 -4842,90627
2 0,9 1,25 1,3 0,002394 0,086711 0,15607979 8,6 -5457,03298
2 0,9 1,25 1,6 0,002394 0,086711 0,15607979 8,63055556 -5452,23391
4 0,9 1,25 0,5 0,002988 0,07761505 0,13970709 8,90833333 -5543,64335
4 0,9 1,25 1,3 0,002988 0,07761505 0,13970709 8,93055556 -5484,43143
4 0,9 1,25 1,6 0,002988 0,07761505 0,13970709 8,95 -5475,73028
4 1,84 0,7 0,5 0,002394 0,086711 0,15607979 9,05277778 -6198,87405
3 0,9 1,25 0,5 0,002988 0,07761505 0,13970709 9,11944444 -5553,54834
4 1,84 0,7 1,3 0,002394 0,086711 0,15607979 9,16944444 -6183,07806
3 0,9 1,25 1,3 0,002988 0,07761505 0,13970709 9,18611111 -5542,52362
3 0,9 1,25 1,6 0,002988 0,07761505 0,13970709 9,20277778 -5534,39844
4 1,84 0,7 1,6 0,002394 0,086711 0,15607979 9,20555556 -6176,95597
3 1,84 0,7 0,5 0,002394 0,086711 0,15607979 9,21944444 -6217,28023
3 1,84 0,7 1,3 0,002394 0,086711 0,15607979 9,33055556 -6201,76329
3 1,84 0,7 1,6 0,002394 0,086711 0,15607979 9,36666667 -6195,25985
4 1,4 1,25 0,5 0,001188 0,12309149 0,22156468 9,38611111 -8023,68156
4 0,9 1,5 0,5 0,002394 0,086711 0,15607979 9,43055556 -6494,91765
3 1,4 1,25 0,5 0,001188 0,12309149 0,22156468 9,45555556 -8042,01043
2 0,9 1,25 0,5 0,002988 0,07761505 0,13970709 9,53055556 -5569,00142
4 0,9 1,5 1,3 0,002394 0,086711 0,15607979 9,54722222 -6478,04616
2 1,84 0,7 0,5 0,002394 0,086711 0,15607979 9,55555556 -6256,09172
4 0,9 1,5 1,6 0,002394 0,086711 0,15607979 9,58333333 -6471,6788
2 0,9 1,25 1,3 0,002988 0,07761505 0,13970709 9,59166667 -5557,50505
2 1,4 1,25 0,5 0,001188 0,12309149 0,22156468 9,59166667 -8070,47498
3 0,9 1,5 0,5 0,002394 0,086711 0,15607979 9,60277778 -6514,16138
2 0,9 1,25 1,6 0,002988 0,07761505 0,13970709 9,61111111 -5553,50749
2 1,84 0,7 1,3 0,002394 0,086711 0,15607979 9,66111111 -6240,6533
4 1,4 1,25 1,3 0,001188 0,12309149 0,22156468 9,68611111 -8007,87751
2 1,84 0,7 1,6 0,002394 0,086711 0,15607979 9,69166667 -6235,03267
3 0,9 1,5 1,3 0,002394 0,086711 0,15607979 9,71666667 -6497,60622
3 1,4 1,25 1,3 0,001188 0,12309149 0,22156468 9,74722222 -8020,1588
3 0,9 1,5 1,6 0,002394 0,086711 0,15607979 9,75277778 -6491,22205
4 1,4 1,25 1,6 0,001188 0,12309149 0,22156468 9,775 -7997,42091
3 1,4 1,25 1,6 0,001188 0,12309149 0,22156468 9,83333333 -8009,5579
2 1,4 1,25 1,3 0,001188 0,12309149 0,22156468 9,86944444 -8043,12848

241
2 0,9 1,5 0,5 0,002394 0,086711 0,15607979 9,95555556 -6554,73422
2 1,4 1,25 1,6 0,001188 0,12309149 0,22156468 9,95555556 -8032,15336
4 1,84 0,7 0,5 0,002988 0,07761505 0,13970709 10,0138889 -6339,88187
4 1,84 0,7 1,3 0,002988 0,07761505 0,13970709 10,0277778 -6272,51466
4 1,84 0,7 1,6 0,002988 0,07761505 0,13970709 10,0472222 -6262,83474
2 0,9 1,5 1,3 0,002394 0,086711 0,15607979 10,0583333 -6538,39308
2 0,9 1,5 1,6 0,002394 0,086711 0,15607979 10,0916667 -6532,66739
3 1,84 0,7 0,5 0,002988 0,07761505 0,13970709 10,2527778 -6352,57326
3 1,84 0,7 1,3 0,002988 0,07761505 0,13970709 10,3166667 -6339,74141
3 1,84 0,7 1,6 0,002988 0,07761505 0,13970709 10,3333333 -6330,59429
4 0,9 1,5 0,5 0,002988 0,07761505 0,13970709 10,4305556 -6643,01845
4 0,9 1,5 1,3 0,002988 0,07761505 0,13970709 10,4416667 -6572,0766
4 0,9 1,5 1,6 0,002988 0,07761505 0,13970709 10,4583333 -6562,17127
3 0,9 1,5 0,5 0,002988 0,07761505 0,13970709 10,6777778 -6656,25798
2 1,84 0,7 0,5 0,002988 0,07761505 0,13970709 10,7138889 -6370,01191
3 0,9 1,5 1,3 0,002988 0,07761505 0,13970709 10,7416667 -6642,99847
3 0,9 1,5 1,6 0,002988 0,07761505 0,13970709 10,7583333 -6633,31805
2 1,84 0,7 1,3 0,002988 0,07761505 0,13970709 10,7694444 -6356,92114
2 1,84 0,7 1,6 0,002988 0,07761505 0,13970709 10,7888889 -6352,47933
4 1,4 1,5 0,5 0,001188 0,12309149 0,22156468 11,0138889 -9607,36581
3 1,4 1,5 0,5 0,001188 0,12309149 0,22156468 11,0972222 -9629,46461
2 0,9 1,5 0,5 0,002988 0,07761505 0,13970709 11,1555556 -6674,78505
2 0,9 1,5 1,3 0,002988 0,07761505 0,13970709 11,2138889 -6661,113
2 0,9 1,5 1,6 0,002988 0,07761505 0,13970709 11,2333333 -6656,15094
2 1,4 1,5 0,5 0,001188 0,12309149 0,22156468 11,2583333 -9664,02832
4 1,4 1,5 1,3 0,001188 0,12309149 0,22156468 11,3416667 -9589,81897
3 1,4 1,5 1,3 0,001188 0,12309149 0,22156468 11,4138889 -9604,67213
4 1,4 1,5 1,6 0,001188 0,12309149 0,22156468 11,4361111 -9578,42333
3 1,4 1,5 1,6 0,001188 0,12309149 0,22156468 11,5083333 -9592,64
2 1,4 1,5 1,3 0,001188 0,12309149 0,22156468 11,5611111 -9632,70401
2 1,4 1,5 1,6 0,001188 0,12309149 0,22156468 11,6527778 -9619,99087
4 1,4 1,25 0,5 0,002394 0,086711 0,15607979 11,8166667 -8403,98365
4 1,84 1,25 0,5 0,001188 0,12309149 0,22156468 11,9333333 -10512,7649
4 1,4 1,25 1,3 0,002394 0,086711 0,15607979 11,9361111 -8382,76701
4 1,4 1,25 1,6 0,002394 0,086711 0,15607979 11,975 -8374,2116
3 1,84 1,25 0,5 0,001188 0,12309149 0,22156468 12,025 -10536,8005
3 1,4 1,25 0,5 0,002394 0,086711 0,15607979 12,0333333 -8429,28955
3 1,4 1,25 1,3 0,002394 0,086711 0,15607979 12,15 -8408,65273
3 1,4 1,25 1,6 0,002394 0,086711 0,15607979 12,1861111 -8400,0314
2 1,84 1,25 0,5 0,001188 0,12309149 0,22156468 12,2 -10574,847
4 1,84 1,25 1,3 0,001188 0,12309149 0,22156468 12,275 -10494,416
3 1,84 1,25 1,3 0,001188 0,12309149 0,22156468 12,3527778 -10511,2957
4 1,84 1,25 1,6 0,001188 0,12309149 0,22156468 12,3722222 -10482,1696

242
3 1,84 1,25 1,6 0,001188 0,12309149 0,22156468 12,45 -10498,0664
2 1,4 1,25 0,5 0,002394 0,086711 0,15607979 12,475 -8482,25713
2 1,84 1,25 1,3 0,001188 0,12309149 0,22156468 12,5111111 -10541,9188
2 1,4 1,25 1,3 0,002394 0,086711 0,15607979 12,5777778 -8461,49647
2 1,84 1,25 1,6 0,001188 0,12309149 0,22156468 12,6055556 -10528,2802
2 1,4 1,25 1,6 0,002394 0,086711 0,15607979 12,6111111 -8454,25178
4 1,4 1,25 1,3 0,002988 0,07761505 0,13970709 13,0416667 -8506,12802
4 1,4 1,25 0,5 0,002988 0,07761505 0,13970709 13,0555556 -8597,9813
4 1,4 1,25 1,6 0,002988 0,07761505 0,13970709 13,0555556 -8493,0437
3 1,4 1,25 0,5 0,002988 0,07761505 0,13970709 13,3583333 -8616,30551
3 1,4 1,25 1,3 0,002988 0,07761505 0,13970709 13,4194444 -8599,1181
3 1,4 1,25 1,6 0,002988 0,07761505 0,13970709 13,4333333 -8586,46367
4 1,4 1,5 0,5 0,002394 0,086711 0,15607979 13,8444444 -10075,0118
2 1,4 1,25 0,5 0,002988 0,07761505 0,13970709 13,9583333 -8640,01596
4 1,4 1,5 1,3 0,002394 0,086711 0,15607979 13,9666667 -10049,4301
4 1,4 1,5 1,6 0,002394 0,086711 0,15607979 14,0027778 -10039,0953
2 1,4 1,25 1,3 0,002988 0,07761505 0,13970709 14,0055556 -8622,32299
4 1,84 1,5 0,5 0,001188 0,12309149 0,22156468 14,0194444 -12597,6701
2 1,4 1,25 1,6 0,002988 0,07761505 0,13970709 14,0222222 -8616,01189
3 1,4 1,5 0,5 0,002394 0,086711 0,15607979 14,1027778 -10105,8864
3 1,84 1,5 0,5 0,001188 0,12309149 0,22156468 14,1277778 -12626,4135
3 1,4 1,5 1,3 0,002394 0,086711 0,15607979 14,2166667 -10080,6771
3 1,4 1,5 1,6 0,002394 0,086711 0,15607979 14,25 -10070,97
2 1,84 1,5 0,5 0,001188 0,12309149 0,22156468 14,3333333 -12672,3076
4 1,84 1,5 1,3 0,001188 0,12309149 0,22156468 14,3833333 -12577,7672
3 1,84 1,5 1,3 0,001188 0,12309149 0,22156468 14,4777778 -12597,7669
4 1,84 1,5 1,6 0,001188 0,12309149 0,22156468 14,4916667 -12563,4509
3 1,84 1,5 1,6 0,001188 0,12309149 0,22156468 14,5833333 -12582,6163
2 1,4 1,5 0,5 0,002394 0,086711 0,15607979 14,6166667 -10169,2595
2 1,84 1,5 1,3 0,001188 0,12309149 0,22156468 14,6666667 -12634,7477
2 1,4 1,5 1,3 0,002394 0,086711 0,15607979 14,7166667 -10144,7826
2 1,4 1,5 1,6 0,002394 0,086711 0,15607979 14,75 -10135,7156
2 1,84 1,5 1,6 0,001188 0,12309149 0,22156468 14,7666667 -12618,9175
4 1,84 1,25 0,5 0,002394 0,086711 0,15607979 14,9861111 -11029,9263
4 1,84 1,25 1,3 0,002394 0,086711 0,15607979 15,1055556 -11001,7894
4 1,84 1,25 1,6 0,002394 0,086711 0,15607979 15,1416667 -10990,5955
4 1,4 1,5 1,3 0,002988 0,07761505 0,13970709 15,2472222 -10197,4247
4 1,4 1,5 1,6 0,002988 0,07761505 0,13970709 15,2583333 -10181,9163
3 1,84 1,25 0,5 0,002394 0,086711 0,15607979 15,2638889 -11063,8186
4 1,4 1,5 0,5 0,002988 0,07761505 0,13970709 15,2805556 -10308,1528
3 1,84 1,25 1,3 0,002394 0,086711 0,15607979 15,375 -11036,2148
3 1,84 1,25 1,6 0,002394 0,086711 0,15607979 15,4083333 -11025,4326
3 1,4 1,5 0,5 0,002988 0,07761505 0,13970709 15,6388889 -10330,6373

243
3 1,4 1,5 1,3 0,002988 0,07761505 0,13970709 15,6888889 -10309,8244
3 1,4 1,5 1,6 0,002988 0,07761505 0,13970709 15,7 -10294,4252
2 1,84 1,25 0,5 0,002394 0,086711 0,15607979 15,8222222 -11133,2077
2 1,84 1,25 1,3 0,002394 0,086711 0,15607979 15,9166667 -11106,3695
2 1,84 1,25 1,6 0,002394 0,086711 0,15607979 15,9472222 -11096,4327
2 1,4 1,5 0,5 0,002988 0,07761505 0,13970709 16,3361111 -10358,7345
2 1,4 1,5 1,3 0,002988 0,07761505 0,13970709 16,3722222 -10337,1393
2 1,4 1,5 1,6 0,002988 0,07761505 0,13970709 16,3861111 -10329,5302
4 1,84 1,25 1,3 0,002988 0,07761505 0,13970709 16,4833333 -11163,3968
4 1,84 1,25 1,6 0,002988 0,07761505 0,13970709 16,4916667 -11146,1165
4 1,84 1,25 0,5 0,002988 0,07761505 0,13970709 16,5305556 -11284,9315
3 1,84 1,25 0,5 0,002988 0,07761505 0,13970709 16,9194444 -11309,6749
3 1,84 1,25 1,3 0,002988 0,07761505 0,13970709 16,9611111 -11286,5617
3 1,84 1,25 1,6 0,002988 0,07761505 0,13970709 16,9694444 -11269,7812
4 1,84 1,5 0,5 0,002394 0,086711 0,15607979 17,5666667 -13225,5845
2 1,84 1,25 0,5 0,002988 0,07761505 0,13970709 17,6694444 -11340,2478
4 1,84 1,5 1,3 0,002394 0,086711 0,15607979 17,6722222 -13191,2466
2 1,84 1,25 1,3 0,002988 0,07761505 0,13970709 17,6972222 -11316,3782
2 1,84 1,25 1,6 0,002988 0,07761505 0,13970709 17,7083333 -11308,0772
4 1,84 1,5 1,6 0,002394 0,086711 0,15607979 17,7083333 -13177,7069
3 1,84 1,5 0,5 0,002394 0,086711 0,15607979 17,8916667 -13266,6061
3 1,84 1,5 1,3 0,002394 0,086711 0,15607979 17,9888889 -13233,1213
3 1,84 1,5 1,6 0,002394 0,086711 0,15607979 18,0222222 -13220,0325
2 1,84 1,5 0,5 0,002394 0,086711 0,15607979 18,5388889 -13349,851
2 1,84 1,5 1,3 0,002394 0,086711 0,15607979 18,6222222 -13316,9857
2 1,84 1,5 1,6 0,002394 0,086711 0,15607979 18,65 -13305,1019
4 1,84 1,5 1,3 0,002988 0,07761505 0,13970709 19,2666667 -13382,9083
4 1,84 1,5 1,6 0,002988 0,07761505 0,13970709 19,2666667 -13361,9197
4 1,84 1,5 0,5 0,002988 0,07761505 0,13970709 19,3527778 -13529,7808
3 1,84 1,5 0,5 0,002988 0,07761505 0,13970709 19,8027778 -13560,1758
3 1,84 1,5 1,3 0,002988 0,07761505 0,13970709 19,825 -13531,5647
3 1,84 1,5 1,6 0,002988 0,07761505 0,13970709 19,8277778 -13511,0897
2 1,84 1,5 0,5 0,002988 0,07761505 0,13970709 20,6694444 -13595,9078
2 1,84 1,5 1,3 0,002988 0,07761505 0,13970709 20,6777778 -13566,3011
2 1,84 1,5 1,6 0,002988 0,07761505 0,13970709 20,6833333 -13555,9734

244
Appendix A4

Second optimization approach, discharging of the unit, airflow rate of 500m3/h

lmpc rmcp cpmcp lalu dym dx dz time energy

3 0,9 0,7 0,5 0,001188 0,15151515 0,18 2,77777778 -3129,93354
4 0,9 0,7 0,5 0,001188 0,15151515 0,18 2,77777778 -3117,60215
4 0,9 0,7 1,3 0,001188 0,15151515 0,18 2,78333333 -3126,97305
2 0,9 0,7 0,5 0,001188 0,15151515 0,18 2,80277778 -3146,56144
4 0,9 0,7 1,6 0,001188 0,15151515 0,18 2,80277778 -3125,29714
3 0,9 0,7 1,3 0,001188 0,15151515 0,18 2,80833333 -3133,43984
3 0,9 0,7 1,6 0,001188 0,15151515 0,18 2,82777778 -3134,06485
2 0,9 0,7 1,3 0,001188 0,15151515 0,18 2,86111111 -3146,67889
2 0,9 0,7 1,6 0,001188 0,15151515 0,18 2,88055556 -3146,353
4 0,9 0,7 0,5 0,002394 0,07518797 0,18 3,61944444 -3162,91918
4 0,9 0,7 1,3 0,002394 0,07518797 0,18 3,66666667 -3158,52281
4 0,9 0,7 1,6 0,002394 0,07518797 0,18 3,68055556 -3156,53257
3 0,9 0,7 0,5 0,002394 0,07518797 0,18 3,70833333 -3174,46447
3 0,9 0,7 1,3 0,002394 0,07518797 0,18 3,75 -3170,39114
3 0,9 0,7 1,6 0,002394 0,07518797 0,18 3,76388889 -3168,57539
2 0,9 0,7 0,5 0,002394 0,07518797 0,18 3,89166667 -3200,68513
2 0,9 0,7 1,3 0,002394 0,07518797 0,18 3,92777778 -3195,89855
2 0,9 0,7 1,6 0,002394 0,07518797 0,18 3,93888889 -3194,06827
4 1,4 0,7 0,5 0,001188 0,15151515 0,18 3,94444444 -4766,54194
3 1,4 0,7 0,5 0,001188 0,15151515 0,18 3,98888889 -4781,06113
4 1,4 0,7 1,3 0,001188 0,15151515 0,18 4,02777778 -4773,34584
4 1,4 0,7 1,6 0,001188 0,15151515 0,18 4,05277778 -4771,62075
3 1,4 0,7 1,3 0,001188 0,15151515 0,18 4,06666667 -4784,34603
2 1,4 0,7 0,5 0,001188 0,15151515 0,18 4,07222222 -4805,04764
3 1,4 0,7 1,6 0,001188 0,15151515 0,18 4,09166667 -4783,43324
2 1,4 0,7 1,3 0,001188 0,15151515 0,18 4,14444444 -4802,72616
2 1,4 0,7 1,6 0,001188 0,15151515 0,18 4,16944444 -4803,03879
4 0,9 0,7 1,6 0,002988 0,06024096 0,18 4,3 -3416,18088
4 0,9 0,7 1,3 0,002988 0,06024096 0,18 4,30555556 -3428,87344
3 0,9 0,7 0,5 0,002988 0,06024096 0,18 4,33055556 -3359,65644
3 0,9 0,7 1,3 0,002988 0,06024096 0,18 4,33333333 -3329,41802
3 0,9 0,7 1,6 0,002988 0,06024096 0,18 4,33888889 -3324,10109
4 0,9 1,25 0,5 0,001188 0,15151515 0,18 4,44166667 -5448,35467
2 0,9 0,7 0,5 0,002988 0,06024096 0,18 4,475 -3253,53777
3 0,9 1,25 0,5 0,001188 0,15151515 0,18 4,49166667 -5464,58815
2 0,9 0,7 1,3 0,002988 0,06024096 0,18 4,50555556 -3249,50488

245
2 0,9 0,7 1,6 0,002988 0,06024096 0,18 4,50833333 -3248,37371
4 0,9 1,25 1,3 0,001188 0,15151515 0,18 4,52777778 -5454,76553
4 0,9 1,25 1,6 0,001188 0,15151515 0,18 4,55555556 -5455,29759
3 0,9 1,25 1,3 0,001188 0,15151515 0,18 4,575 -5469,68741
2 0,9 1,25 0,5 0,001188 0,15151515 0,18 4,58611111 -5491,39319
3 0,9 1,25 1,6 0,001188 0,15151515 0,18 4,6 -5467,67968
2 0,9 1,25 1,3 0,001188 0,15151515 0,18 4,66388889 -5491,47961
4 0,9 0,7 0,5 0,002988 0,06024096 0,18 4,66388889 -3630,77828
2 0,9 1,25 1,6 0,001188 0,15151515 0,18 4,68888889 -5489,19899
4 1,84 0,7 0,5 0,001188 0,15151515 0,18 4,99166667 -6213,90236
3 1,84 0,7 0,5 0,001188 0,15151515 0,18 5,05 -6234,84716
4 1,84 0,7 1,3 0,001188 0,15151515 0,18 5,08611111 -6223,84718
4 1,84 0,7 1,6 0,001188 0,15151515 0,18 5,11666667 -6221,74528
3 1,84 0,7 1,3 0,001188 0,15151515 0,18 5,13888889 -6239,49298
2 1,84 0,7 0,5 0,001188 0,15151515 0,18 5,15555556 -6263,851
3 1,84 0,7 1,6 0,001188 0,15151515 0,18 5,16666667 -6238,10351
4 0,9 1,5 0,5 0,001188 0,15151515 0,18 5,2 -6506,95097
2 1,84 0,7 1,3 0,001188 0,15151515 0,18 5,23888889 -6264,41417
3 0,9 1,5 0,5 0,001188 0,15151515 0,18 5,25833333 -6527,4394
2 1,84 0,7 1,6 0,001188 0,15151515 0,18 5,26388889 -6261,74967
4 1,4 0,7 0,5 0,002394 0,07518797 0,18 5,28333333 -4879,73679
4 0,9 1,5 1,3 0,001188 0,15151515 0,18 5,29722222 -6516,03897
4 0,9 1,5 1,6 0,001188 0,15151515 0,18 5,32777778 -6514,71781
4 1,4 0,7 1,3 0,002394 0,07518797 0,18 5,33055556 -4872,69686
4 1,4 0,7 1,6 0,002394 0,07518797 0,18 5,34722222 -4868,60783
3 0,9 1,5 1,3 0,001188 0,15151515 0,18 5,35 -6531,30231
2 0,9 1,5 0,5 0,001188 0,15151515 0,18 5,37222222 -6557,52067
3 0,9 1,5 1,6 0,001188 0,15151515 0,18 5,38055556 -6530,06002
3 1,4 0,7 0,5 0,002394 0,07518797 0,18 5,41666667 -4897,73858
2 0,9 1,5 1,3 0,001188 0,15151515 0,18 5,45555556 -6558,33842
3 1,4 0,7 1,3 0,002394 0,07518797 0,18 5,46111111 -4891,01815
3 1,4 0,7 1,6 0,002394 0,07518797 0,18 5,475 -4888,16894
2 0,9 1,5 1,6 0,001188 0,15151515 0,18 5,48333333 -6556,85575
2 1,4 0,7 0,5 0,002394 0,07518797 0,18 5,68055556 -4936,99707
2 1,4 0,7 1,3 0,002394 0,07518797 0,18 5,72222222 -4930,0881
2 1,4 0,7 1,6 0,002394 0,07518797 0,18 5,73611111 -4927,55112
4 0,9 1,25 0,5 0,002394 0,07518797 0,18 5,95 -5590,97266
4 0,9 1,25 1,3 0,002394 0,07518797 0,18 6 -5583,19768
4 0,9 1,25 1,6 0,002394 0,07518797 0,18 6,01388889 -5578,40541
3 0,9 1,25 0,5 0,002394 0,07518797 0,18 6,1 -5611,61731
3 0,9 1,25 1,3 0,002394 0,07518797 0,18 6,14444444 -5604,48033
3 0,9 1,25 1,6 0,002394 0,07518797 0,18 6,16111111 -5600,83649
4 1,4 0,7 1,6 0,002988 0,06024096 0,18 6,27777778 -5277,19241

246
4 1,4 0,7 1,3 0,002988 0,06024096 0,18 6,28333333 -5295,8924
3 1,4 0,7 1,6 0,002988 0,06024096 0,18 6,31388889 -5136,98108
3 1,4 0,7 1,3 0,002988 0,06024096 0,18 6,31944444 -5145,08515
3 1,4 0,7 0,5 0,002988 0,06024096 0,18 6,32777778 -5191,41966
2 0,9 1,25 0,5 0,002394 0,07518797 0,18 6,39722222 -5657,01625
2 0,9 1,25 1,3 0,002394 0,07518797 0,18 6,43888889 -5649,14673
2 0,9 1,25 1,6 0,002394 0,07518797 0,18 6,45277778 -5646,25564
4 1,4 1,25 0,5 0,001188 0,15151515 0,18 6,52222222 -8388,98985
2 1,4 0,7 0,5 0,002988 0,06024096 0,18 6,53333333 -5028,11131
2 1,4 0,7 1,3 0,002988 0,06024096 0,18 6,55 -5023,50983
2 1,4 0,7 1,6 0,002988 0,06024096 0,18 6,55833333 -5021,43612
3 1,4 1,25 0,5 0,001188 0,15151515 0,18 6,59722222 -8417,41976
4 1,4 1,25 1,3 0,001188 0,15151515 0,18 6,63333333 -8401,73555
4 1,4 1,25 1,6 0,001188 0,15151515 0,18 6,66666667 -8401,71645
4 1,84 0,7 0,5 0,002394 0,07518797 0,18 6,68611111 -6390,93608
3 1,4 1,25 1,3 0,001188 0,15151515 0,18 6,70277778 -8423,38137
3 1,4 1,25 1,6 0,001188 0,15151515 0,18 6,73333333 -8420,13347
4 1,84 0,7 1,3 0,002394 0,07518797 0,18 6,73611111 -6381,42098
2 1,4 1,25 0,5 0,001188 0,15151515 0,18 6,73888889 -8458,0561
4 1,84 0,7 1,6 0,002394 0,07518797 0,18 6,75277778 -6376,08946
2 1,4 1,25 1,3 0,001188 0,15151515 0,18 6,83333333 -8457,08357
4 1,4 0,7 0,5 0,002988 0,06024096 0,18 6,83333333 -5606,13224
3 1,84 0,7 0,5 0,002394 0,07518797 0,18 6,85277778 -6414,51635
2 1,4 1,25 1,6 0,001188 0,15151515 0,18 6,86388889 -8455,10851
3 1,84 0,7 1,3 0,002394 0,07518797 0,18 6,9 -6406,35681
3 1,84 0,7 1,6 0,002394 0,07518797 0,18 6,91666667 -6402,22575
4 0,9 1,5 0,5 0,002394 0,07518797 0,18 6,96388889 -6695,22083
4 0,9 1,5 1,3 0,002394 0,07518797 0,18 7,01388889 -6684,80986
4 0,9 1,5 1,6 0,002394 0,07518797 0,18 7,03055556 -6679,54269
4 0,9 1,25 1,6 0,002988 0,06024096 0,18 7,05833333 -6048,73136
4 0,9 1,25 1,3 0,002988 0,06024096 0,18 7,07777778 -6069,88252
3 0,9 1,25 1,3 0,002988 0,06024096 0,18 7,10277778 -5897,27061
3 0,9 1,25 1,6 0,002988 0,06024096 0,18 7,10555556 -5887,84693
3 0,9 1,25 0,5 0,002988 0,06024096 0,18 7,12777778 -5950,47458
3 0,9 1,5 0,5 0,002394 0,07518797 0,18 7,13333333 -6719,94171
3 0,9 1,5 1,3 0,002394 0,07518797 0,18 7,18611111 -6711,41992
2 1,84 0,7 0,5 0,002394 0,07518797 0,18 7,18888889 -6466,33241
3 0,9 1,5 1,6 0,002394 0,07518797 0,18 7,2 -6707,62335
2 1,84 0,7 1,3 0,002394 0,07518797 0,18 7,23055556 -6457,59803
2 1,84 0,7 1,6 0,002394 0,07518797 0,18 7,24722222 -6454,00424
2 0,9 1,25 0,5 0,002988 0,06024096 0,18 7,35277778 -5763,48962
2 0,9 1,25 1,3 0,002988 0,06024096 0,18 7,36944444 -5758,28181
2 0,9 1,25 1,6 0,002988 0,06024096 0,18 7,37777778 -5755,87458

247
2 0,9 1,5 0,5 0,002394 0,07518797 0,18 7,48611111 -6774,25411
2 0,9 1,5 1,3 0,002394 0,07518797 0,18 7,53055556 -6765,2466
2 0,9 1,5 1,6 0,002394 0,07518797 0,18 7,54444444 -6761,38768
4 1,4 1,5 0,5 0,001188 0,15151515 0,18 7,65833333 -10038,9453
4 0,9 1,25 0,5 0,002988 0,06024096 0,18 7,70277778 -6424,53316
3 1,4 1,5 0,5 0,001188 0,15151515 0,18 7,74444444 -10070,5118
4 1,4 1,5 1,3 0,001188 0,15151515 0,18 7,77777778 -10054,6101
4 1,4 1,5 1,6 0,001188 0,15151515 0,18 7,81388889 -10053,5856
3 1,4 1,5 1,3 0,001188 0,15151515 0,18 7,85833333 -10079,5573
3 1,4 1,5 1,6 0,001188 0,15151515 0,18 7,89444444 -10077,4039
2 1,4 1,5 0,5 0,001188 0,15151515 0,18 7,91111111 -10121,1586
4 1,84 0,7 1,6 0,002988 0,06024096 0,18 7,91944444 -6915,37649
4 1,84 0,7 1,3 0,002988 0,06024096 0,18 7,93611111 -6940,05595
3 1,84 0,7 1,3 0,002988 0,06024096 0,18 7,975 -6743,30766
3 1,84 0,7 1,6 0,002988 0,06024096 0,18 7,975 -6732,22627
3 1,84 0,7 0,5 0,002988 0,06024096 0,18 8,00833333 -6803,24566
2 1,4 1,5 1,3 0,001188 0,15151515 0,18 8,01388889 -10120,6285
2 1,4 1,5 1,6 0,001188 0,15151515 0,18 8,04722222 -10117,1854
4 0,9 1,5 1,6 0,002988 0,06024096 0,18 8,24444444 -7244,53921
2 1,84 0,7 0,5 0,002988 0,06024096 0,18 8,25 -6590,45196
4 0,9 1,5 1,3 0,002988 0,06024096 0,18 8,26111111 -7270,67955
2 1,84 0,7 1,3 0,002988 0,06024096 0,18 8,27222222 -6583,83976
2 1,84 0,7 1,6 0,002988 0,06024096 0,18 8,28055556 -6581,55647
4 1,84 1,25 0,5 0,001188 0,15151515 0,18 8,29722222 -10981,7975
3 0,9 1,5 1,3 0,002988 0,06024096 0,18 8,30277778 -7064,92127
3 0,9 1,5 1,6 0,002988 0,06024096 0,18 8,30277778 -7053,36702
3 0,9 1,5 0,5 0,002988 0,06024096 0,18 8,34166667 -7127,6522
3 1,84 1,25 0,5 0,001188 0,15151515 0,18 8,39166667 -11017,8049
4 1,84 1,25 1,3 0,001188 0,15151515 0,18 8,425 -10999,6922
4 1,84 1,25 1,6 0,001188 0,15151515 0,18 8,46111111 -10998,2796
3 1,84 1,25 1,3 0,001188 0,15151515 0,18 8,51111111 -11026,7906
3 1,84 1,25 1,6 0,001188 0,15151515 0,18 8,54722222 -11023,8214
2 1,84 1,25 0,5 0,001188 0,15151515 0,18 8,57222222 -11071,1797
2 0,9 1,5 0,5 0,002988 0,06024096 0,18 8,58888889 -6904,58534
2 0,9 1,5 1,3 0,002988 0,06024096 0,18 8,61111111 -6898,20074
2 0,9 1,5 1,6 0,002988 0,06024096 0,18 8,61944444 -6895,54209
4 1,84 0,7 0,5 0,002988 0,06024096 0,18 8,66388889 -7344,87647
2 1,84 1,25 1,3 0,001188 0,15151515 0,18 8,68055556 -11071,6085
2 1,84 1,25 1,6 0,001188 0,15151515 0,18 8,71388889 -11068,2429
4 1,4 1,25 0,5 0,002394 0,07518797 0,18 8,71388889 -8658,04044
4 1,4 1,25 1,3 0,002394 0,07518797 0,18 8,76388889 -8644,74857
4 1,4 1,25 1,6 0,002394 0,07518797 0,18 8,78055556 -8638,19515
3 1,4 1,25 0,5 0,002394 0,07518797 0,18 8,93055556 -8691,21662

248
3 1,4 1,25 1,3 0,002394 0,07518797 0,18 8,97777778 -8679,93997
3 1,4 1,25 1,6 0,002394 0,07518797 0,18 8,99444444 -8674,58473
4 0,9 1,5 0,5 0,002988 0,06024096 0,18 9,02222222 -7694,57105
2 1,4 1,25 0,5 0,002394 0,07518797 0,18 9,36666667 -8760,41284
2 1,4 1,25 1,3 0,002394 0,07518797 0,18 9,40833333 -8748,49103
2 1,4 1,25 1,6 0,002394 0,07518797 0,18 9,42222222 -8744,07388
4 1,84 1,5 0,5 0,001188 0,15151515 0,18 9,75 -13151,2103
3 1,84 1,5 0,5 0,001188 0,15151515 0,18 9,86111111 -13194,5724
4 1,84 1,5 1,3 0,001188 0,15151515 0,18 9,88611111 -13173,3027
4 1,84 1,5 1,6 0,001188 0,15151515 0,18 9,92777778 -13171,812
3 1,84 1,5 1,3 0,001188 0,15151515 0,18 9,98888889 -13206,0402
3 1,84 1,5 1,6 0,001188 0,15151515 0,18 10,0277778 -13203,1162
2 1,84 1,5 0,5 0,001188 0,15151515 0,18 10,075 -13259,6935
2 1,84 1,5 1,3 0,001188 0,15151515 0,18 10,1861111 -13260,0892
4 1,4 1,5 0,5 0,002394 0,07518797 0,18 10,2 -10375,4283
2 1,84 1,5 1,6 0,001188 0,15151515 0,18 10,2222222 -13256,28
4 1,4 1,5 1,3 0,002394 0,07518797 0,18 10,25 -10359,6977
4 1,4 1,5 1,6 0,002394 0,07518797 0,18 10,2666667 -10351,155
4 1,4 1,25 1,6 0,002988 0,06024096 0,18 10,3 -9370,36699
4 1,4 1,25 1,3 0,002988 0,06024096 0,18 10,325 -9403,21138
3 1,4 1,25 1,3 0,002988 0,06024096 0,18 10,3666667 -9138,62869
3 1,4 1,25 1,6 0,002988 0,06024096 0,18 10,3666667 -9123,94439
3 1,4 1,25 0,5 0,002988 0,06024096 0,18 10,4194444 -9220,31224
3 1,4 1,5 0,5 0,002394 0,07518797 0,18 10,4555556 -10415,0279
3 1,4 1,5 1,3 0,002394 0,07518797 0,18 10,5027778 -10401,5529
3 1,4 1,5 1,6 0,002394 0,07518797 0,18 10,5166667 -10395,6095
2 1,4 1,25 0,5 0,002988 0,06024096 0,18 10,7305556 -8932,14078
2 1,4 1,25 1,3 0,002988 0,06024096 0,18 10,7444444 -8923,39488
2 1,4 1,25 1,6 0,002988 0,06024096 0,18 10,75 -8919,67375
2 1,4 1,5 0,5 0,002394 0,07518797 0,18 10,9638889 -10497,961
2 1,4 1,5 1,3 0,002394 0,07518797 0,18 11,0027778 -10484,1584
2 1,4 1,5 1,6 0,002394 0,07518797 0,18 11,0166667 -10478,2959
4 1,84 1,25 0,5 0,002394 0,07518797 0,18 11,0361111 -11356,4427
4 1,84 1,25 1,3 0,002394 0,07518797 0,18 11,0833333 -11338,8645
4 1,84 1,25 1,6 0,002394 0,07518797 0,18 11,0972222 -11329,9206
4 1,4 1,25 0,5 0,002988 0,06024096 0,18 11,2805556 -9950,17668
3 1,84 1,25 0,5 0,002394 0,07518797 0,18 11,3111111 -11400,3951
3 1,84 1,25 1,3 0,002394 0,07518797 0,18 11,3555556 -11385,408
3 1,84 1,25 1,6 0,002394 0,07518797 0,18 11,3694444 -11378,8597
2 1,84 1,25 0,5 0,002394 0,07518797 0,18 11,8611111 -11490,8977
2 1,84 1,25 1,3 0,002394 0,07518797 0,18 11,8944444 -11475,0229
2 1,84 1,25 1,6 0,002394 0,07518797 0,18 11,9083333 -11469,1808
4 1,4 1,5 1,6 0,002988 0,06024096 0,18 12,0416667 -11228,386

249
4 1,4 1,5 1,3 0,002988 0,06024096 0,18 12,075 -11267,8209
3 1,4 1,5 1,6 0,002988 0,06024096 0,18 12,1138889 -10933,7838
3 1,4 1,5 1,3 0,002988 0,06024096 0,18 12,1166667 -10951,4469
3 1,4 1,5 0,5 0,002988 0,06024096 0,18 12,1861111 -11049,6876
2 1,4 1,5 0,5 0,002988 0,06024096 0,18 12,5416667 -10703,9369
2 1,4 1,5 1,3 0,002988 0,06024096 0,18 12,55 -10693,5219
2 1,4 1,5 1,6 0,002988 0,06024096 0,18 12,5527778 -10689,2042
4 1,84 1,5 0,5 0,002394 0,07518797 0,18 12,9222222 -13612,2248
4 1,84 1,5 1,3 0,002394 0,07518797 0,18 12,9638889 -13590,7128
4 1,84 1,5 1,6 0,002394 0,07518797 0,18 12,975 -13579,6697
4 1,84 1,25 1,6 0,002988 0,06024096 0,18 13,0194444 -12289,389
4 1,84 1,25 1,3 0,002988 0,06024096 0,18 13,0527778 -12332,2952
3 1,84 1,25 1,6 0,002988 0,06024096 0,18 13,0888889 -11967,0763
3 1,84 1,25 1,3 0,002988 0,06024096 0,18 13,0972222 -11986,5012
3 1,84 1,25 0,5 0,002988 0,06024096 0,18 13,1777778 -12094,2554
4 1,4 1,5 0,5 0,002988 0,06024096 0,18 13,2055556 -11922,3368
3 1,84 1,5 0,5 0,002394 0,07518797 0,18 13,2416667 -13665,1638
3 1,84 1,5 1,3 0,002394 0,07518797 0,18 13,2805556 -13647,0099
3 1,84 1,5 1,6 0,002394 0,07518797 0,18 13,2916667 -13638,7362
2 1,84 1,25 0,5 0,002988 0,06024096 0,18 13,5555556 -11715,5714
2 1,84 1,25 1,3 0,002988 0,06024096 0,18 13,5583333 -11704,2406
2 1,84 1,25 1,6 0,002988 0,06024096 0,18 13,5611111 -11699,4572
2 1,84 1,5 0,5 0,002394 0,07518797 0,18 13,8805556 -13773,1141
2 1,84 1,5 1,3 0,002394 0,07518797 0,18 13,9083333 -13753,747
2 1,84 1,5 1,6 0,002394 0,07518797 0,18 13,9194444 -13746,3786
4 1,84 1,25 0,5 0,002988 0,06024096 0,18 14,2861111 -13048,4652
4 1,84 1,5 1,6 0,002988 0,06024096 0,18 15,2138889 -14726,7405
4 1,84 1,5 1,3 0,002988 0,06024096 0,18 15,2583333 -14778,2241
3 1,84 1,5 1,6 0,002988 0,06024096 0,18 15,2888889 -14341,4416
3 1,84 1,5 1,3 0,002988 0,06024096 0,18 15,3 -14364,7637
3 1,84 1,5 0,5 0,002988 0,06024096 0,18 15,4083333 -14494,2783
2 1,84 1,5 1,3 0,002988 0,06024096 0,18 15,825 -14025,1578
2 1,84 1,5 1,6 0,002988 0,06024096 0,18 15,8277778 -14019,0778
2 1,84 1,5 0,5 0,002988 0,06024096 0,18 15,8333333 -14039,118
4 1,84 1,5 0,5 0,002988 0,06024096 0,18 16,7166667 -15635,2295

250