Professional Documents
Culture Documents
90
ECOS 2012
The 25th International Conference on Efficiency, Cost,
Optimization and Simulation of Energy Conversion
Systems and Processes
(Perugia, June 26th-June 29th, 2012)
edited by
Umberto Desideri, Giampaolo Manfrida,
Enrico Sciubba
2012
ECOS 2012
The 25th International Conference on
Efficiency, Cost, Optimization and Simulation
of Energy Conversion Systems and Processes
ii
Scientific Committee
Riccardo Basosi, University of Siena, Italy
Gino Bella, University of Roma Tor Vergata, Italy
Asfaw Beyene, San Diego State University, United States
Ryszard Bialecki, Silesian Institute of Tecnology, Poland
Gianni Bidini, University of Perugia, Italy
Ana M. Blanco-Marigorta, University of Las Palmas de Gran Canaria, Spain
Olav Bolland, University of Science and Technology (NTNU), Norway
Ren Cornelissen, Cornelissen Consulting, The Netherlands
Franco Cotana, University of Perugia, Italy
Alexandru Dobrovicescu, Polytechnical University of Bucharest, Romania
Gheorghe Dumitrascu, Technical University of Iasi, Romania
Brian Elmegaard, Technical University of Denmark , Denmark
Daniel Favrat, EPFL, Switzerland
Michel Feidt, ENSEM - LEMTA University Henri Poincar, France
Daniele Fiaschi, University of Florence, Italy
Marco Frey, Scuola Superiore S. Anna, Italy
Richard A Gaggioli, Marquette University, USA
Carlo N. Grimaldi, University of Perugia, Italy
Simon Harvey, Chalmers University of Technology, Sweden
Hasan Heperkan, Yildiz Technical University, Turkey
Abel Abel Hernandez-Guerrero, University of Guanajuato, Mexico
Jiri Jaromir Kleme, University of Pannonia, Hungary
Zornitza V. Kirova-Yordanova, University "Prof. Assen Zlatarov", Bulgaria
Noam Lior, University of Pennsylvania, United States
Francesco Martelli, University of Florence, Italy
Aristide Massardo, University of Genova, Italy
Jim McGovern, Dublin Institute of Technology, Ireland
Alberto Mirandola, University of Padova, Italy
Michael J. Moran, The Ohio State University, United States
Tatiana Morosuk, Technical University of Berlin, Germany
Pericles Pilidis, University of Cranfield, United Kingdom
Constantine D. Rakopoulos, National Technical University of Athens, Greece
Predrag Raskovic, University of Nis, Serbia and Montenegro
Mauro Reini, University of Trieste, Italy
Gianfranco Rizzo, University of Salerno, Italy
Marc A. Rosen, University of Ontario, Canada
Luis M. Serra, University of Zaragoza, Spain
Gordana Stefanovic, University of Nis, Serbia and Montenegro
Andrea Toffolo, Lule University of Technology, Sweden
Wojciech Stanek, Silesian University of Technology, Poland
George Tsatsaronis, Technical University Berlin, Germany
Antonio Valero, University of Zaragoza, Spain
Michael R. von Spakovsky, Virginia Tech, USA
Stefano Ubertini, Parthenope University of Naples, Italy
Sergio Ulgiati, Parthenope University of Naples, Italy
Sergio Usn, Universidad de Zaragoza, Spain
Roman Weber, Clausthal University of Technology, Germany
Ryohei Yokoyama, Osaka Prefecture University, Japan
Na Zhang, Institute of Engineering Thermophysics, Chinese Academy of Sciences, China
iv
CONTENT MANAGEMENT
The index lists all the papers contained all the eight volumes of the Proceedings of the
ECOS 2012 International Conference.
Page numbers are listed only for papers within the Volume you are looking at.
The ID code allows to trace back the identification number assigned to the paper within
the Conference submission, review and track organization processes.
vii
CONTENT
VOLUME V
V. 1 RENEWABLE ENERGY CONVERSION SYSTEMS
A co-powered concentrated solar power Rankine cycle concept for ....
small size combined heat and power (ID 276)
Alessandro Corsini, Domenico Borello, Franco Rispoli, Eileen Tortora
Pag. 1
Pag. 17
Pag. 26
Alternative feedstock for the biodiesel and energy production: the ....
OVEST project (ID 98)
Matteo Prussi, David Chiaramonti, Lucia Recchia, Francesco Martelli, Fabio
Guidotti
Pag. 38
Pag. 47
Pag. 56
....
Pag. 68
of expansion ....
Pag. 78
....
Pag. 94
....
Pag. 106
Pag. 114
Energy and exergy analysis of the first hybrid solar-gas power plant in ....
Algeria (ID 176)
Fouad Khaldi
Pag. 130
Pag. 144
Pag. 159
....
Pag. 173
Pag. 185
Experimental study of tar and particles content of the produced gas in ....
a double stage downdraft gasifier (ID 487)
Ana Lisbeth Galindo Noguera, Sandra Yamile Giraldo, Rene Lesme-Jan,
Vladimir Melian Cobas, Rubenildo Viera Andrade, Electo Silva Lora
Pag. 197
Pag. 209
Investigations on the use of biogas for small scale decentralized CHP ....
applications with a focus on stability and emissions (ID 140)
Steven MacLean, Eren Tali, Anne Giese, Jrg Leicher
Pag. 218
....
Pag. 229
Mirrors in the sky: status and some supporting materials experiments ....
(ID 184)
Noam Lior
Pag. 253
Numerical parametric study for different cold storage designs and ....
strategies of a solar driven thermoacoustic cooler system (ID 284)
Maxime Perier-Muzet, Pascal Stouffs, Jean-Pierre Bedecarrats, Jean
Castaing-Lasvignottes
Pag. 274
Pag. 290
....
Pag. 309
....
Pag. 331
Pag. 350
Pag. 363
Pag. 381
ix
....
Pag. 373
Vegetable oils of soybean, sunflower and tung as alternative fuels for ....
compression ignition engines (ID 500)
Ricardo Morel Hartmann, Nury Nieto Garzn, Eduardo Morel Hartmann, Amir
Antonio Martins Oliveira Jr, Edson Bazzo, Bruno Okuda, Joselia Piluski
Pag. 409
Pag. 427
V. 2 FUEL CELLS
Comparison study on different SOFC hybrid systems with zero-CO2 ....
emission (ID 196)
Liqiang Duan, Kexin Huang, Xiaoyuan Zhang and Yongping Yang
Pag. 440
Pag. 456
Modelling of a CHP SOFC power system fed with biogas from ....
anaerobic digestion of municipal wastes integrated with a solar collector
and storage units (ID 491)
Domenico Borello, Sara Evangelisti, Eileen Tortora
Pag. 472
VOLUME I
I . 1 - SIMULATION OF ENERGY CONVERSION SYSTEMS
A novel hybrid-fuel compressed air energy storage system for Chinas situation (ID 531)
Wenyi Liu, Yongping Yang, Weide Zhang, Gang Xu,and Ying Wu
A review of Stirling engine technologies applied to micro-cogeneration systems (ID 338)
Ana C Ferreira, Manuel L Nunes, Lus B Martins, Senhorinha F Teixeira
An organic Rankine cycle off-design model for the search of the optimal control strategy
(ID 295)
Andrea Toffolo, Andrea Lazzaretto, Giovanni Manente, Marco Paci
Automated superstructure generation and optimization of distributed energy supply
systems (ID 518)
Philip Voll, Carsten Klaffke, Maike Hennen, Andr Bardow
Characterisation and classification of solid recovered fuels (SRF) and model development
of a novel thermal utilization concept through air- gasification (ID 506)
Panagiotis Vounatsos, Konstantinos Atsonios, Mihalis Agraniotis, Kyriakos D. Panopoulos, George
Koufodimos,Panagiotis Grammelis, Emmanuel Kakaras
Design and modelling of a novel compact power cycle for low temperature heat sources
(ID 177)
Jorrit Wronski, Morten Juel Skovrup, Brian Elmegaard, Harald Nes Risl, Fredrik Haglind
Dynamic simulation of combined cycles operating in transient conditions: an innovative
approach to determine the steam drums life consumption (ID 439)
Stefano Bracco
Effect of auxiliary electrical power consumptions on organic Rankine cycle system with
low-temperature waste heat source (ID 235)
Samer Maalouf, Elias Boulawz Ksayer, Denis Clodic
Energetic and exergetic analysis of waste heat recovery systems in the cement industry
(ID 228)
Sotirios Karellas, Aris Dimitrios Leontaritis, Georgios Panousis, Evangelos Bellos, Emmanuel
Kakaras
Energy and exergy analysis of repowering options for Greek lignite-fired power plants (ID
230)
Sotirios Karellas, Aggelos Doukelis, Grammatiki Zanni, Emmanuel Kakaras
Energy saving by a simple solar collector with reflective panels and boiler (ID 366)
Anna Stoppato, Renzo Tosato
Exergetic analysis of biomass fired double-stage Organic Rankine Cycle (ORC) (ID 37)
Markus Preiinger, Florian Heberle, Dieter Brggemann
Experimental tests and modelization of a domestic-scale organic Rankine cycle (ID 156)
Roberto Bracco, Stefano Clemente, Diego Micheli, Mauro Reini
Model of a small steam engine for renewable domestic CHP system (ID 31 )
Giampaolo Manfrida, Giovanni Ferrara, Alessandro Pescioni
Model of vacuum glass heat pipe solar collectors (ID 312)
Daniele Fiaschi, Giampaolo Manfrida
Modelling and exergy analysis of a plasma furnace for aluminum melting process (ID 254)
Luis Enrique Acevedo, Sergio Usn, Javier Uche, Patxi Rodrguez
Modelling and experimental validation of a solar cooling installation (ID 296)
Guillaume Anies, Pascal Stouffs, Jean Castaing-Lasvignottes
The influence of operating parameters and occupancy rate of thermoelectric modules on
the electricity generation (ID 314)
Camille Favarel, Jean-Pierre Bdcarrats, Tarik Kousksou, Daniel Champier
Thermodynamic and heat transfer analysis of rice straw co-firing in a Brazilian pulverised
coal boiler (ID 236)
Raphael Miyake, Alvaro Restrepo, Fbio Kleveston Edson Bazzo, Marcelo Bzuneck
Thermophotovoltaic generation: A state of the art review (ID 88)
Matteo Bosi, Claudio Ferrari, Francesco Melino, Michele Pinelli, Pier Ruggero Spina, Mauro
Venturini
I . 2 HEAT AND MASS TRANSFER
A DNS method for particle motion to establish boundary conditions in coal gasifiers (ID
49)
Efstathios E Michaelides, Zhigang Feng
Effective thermal conductivity with convection and radiation in packed bed (ID 60)
Yusuke Asakuma
Experimental and CFD study of a single phase cone-shaped helical coiled heat exchanger:
an empirical correlation (ID 375)
Daniel Flrez-Orrego, Walter Arias, Diego Lpez, Hctor Velsquez
Thermofluiddynamic model for control analysis of latent heat thermal storage system (ID
207)
Adriano Sciacovelli, Vittorio Verda, Flavio Gagliardi
Towards the development of an efficient immersed particle heat exchanger: particle
transfer from low to high pressure (ID 202)
Luciano A. Catalano, Riccardo Amirante, Stefano Copertino, Paolo Tamburrano, Fabio De Bellis
xi
I . 3 INDUSTRIAL ECOLOGY
Anthropogenic heat and exergy balance of the atmosphere (ID 122)
Asfaw Beyene, David MacPhee, Ron Zevenhoven
Determination of environmental remediation cost of municipal waste in terms of extended
exergy (ID 63)
Candeniz Seckin, Ahmet R. Bayulken
Development of product category rules for the application of life cycle assessment to
carbon capture and storage (537)
Carlo Strazza, Adriana Del Borghi, Michela Gallo
Electricity production from renewable and non-renewable energy sources: a comparison
of environmental, economic and social sustainability indicators with exergy losses
throughout the supply chain (ID 247)
Lydia Stougie, Hedzer van der Kooi, Rob Stikkelman
Exergy analysis of the industrial symbiosis model in Kalundborg
(ID 218)
Alicia Valero Delgado, Sergio Usn, Jorge Costa
Global gold mining: is technological learning overcoming the declining in ore grades? (ID
277)
Adriana Domnguez, Alicia Valero
Personal transportation energy consumption (ID305)
Matteo Muratori, Emmanuele Serra, Vincenzo Marano, Michael Moran
Resource use evaluation of Turkish transportation sector via the extended exergy
accounting method (ID 43)
Candeniz Seckin, Enrico Sciubba, Ahmet R. Bayulken
The impact of higher energy prices on socio-economic inequalities of German social
groups (ID 80)
Holger Schlr, Wolfgang Fischer, Jrgen-Friedrich Hake
VOLUME II
II . 1 EXERGY ANALYSIS AND 2ND LAW ANALYSIS
A comparative analysis of cryogenic recuperative heat exchangers based on exergy
destruction (ID 129)
Adina Teodora Gheorghian, Alexandru Dobrovicescu, Lavinia Grosu, Bogdan Popescu, Claudia
Ionita
A critical exploration of the usefulness of rational efficiency as a performance parameter
for heat exchangers (ID 307)
Jim McGovern, Georgiana Tirca-Dragomirescu, Michel Feidt, Alexandru Dobrovicescu
A new procedure for the design of LNG processes by combining exergy and pinch
analyses (ID 238)
Danahe Marmolejo-Correa, Truls Gundersen
Advances in the distribution of environmental cost of water bodies through the exergy
concept in the Ebro river (ID 258)
Javier Uche Marcuello, Amaya Martnez Gracia, Beatriz Carrasquer lvarez, Antonio Valero Capilla
Application of the entropy generation minimization method to a solar heat exchanger: a
pseudo-optimization design process based on the analysis of the local entropy generation
maps (ID 357)
Giorgio Giangaspero, Enrico Sciubba
Comparative analysis of ammonia and carbon dioxide two-stage cycles for simultaneous
cooling and heating (ID 84)
Alexandru Dobrovicescu, Ciprian Filipoiu, Emilia Cerna Mladin, Valentin Apostol, Liviu Drughean
xii
Optimization and design of pin-fin heat sinks based on minimum entropy generation (ID 6)
Jose-Luis Zuniga-Cerroblanco, Abel Hernandez-Guerrero, Carlos A. Rubio-Jimenez, Cuauhtemoc
Rubio-Arana, Sosimo E. Diaz-Mendez
Performance analysis of a district heating system (ID 271)
Andrej Ljubenko, Alojz Poredo, Tatiana Morosuk, George Tsatsaronis
System analysis of exergy losses in an integrated oxy-fuel combustion power plant (ID 64)
Andrzej Zi bik, Pawe G adysz
What is the cost of losing irreversibly the mineral capital on Earth? (ID 220)
Alicia Valero Delgado, Antonio Valero
II . 2 THERMODYNAMICS
A new polygeneration system for methanol and power based on coke oven gas and coal
gas (ID 252)
Hu Lin, Hongguang Jin, Lin Gao, Rumou Li
Argon-Water closed gas cycle (ID 67)
Federico Fionelli, Giovanni Molinari
Binary alkane mixtures as fluids in Rankine cycles (ID 246)
M. Aslam Siddiqi, Burak Atakan
Excess enthalpies of second generation biofuels (ID 308)
Alejandro Moreau, Jos Juan Segovia, M. Carmen Martn, Miguel ngel Villaman, Csar R.
Chamorro, Rosa M. Villaman
Local stability analysis of a Curzon-Ahlborn engine considering the Van der Waals
equation state in the maximum ecological regime (ID 281)
Ricardo Richard Pez-Hernndez, Pedro Portillo-Daz, Delfino Ladino-Luna,
Marco Antonio Barranco-Jimnez
Some remarks on the Carnot's theorem (ID 325)
Julian Gonzalez Ayala, Fernando Angulo-Brown
The Dead State (ID 340)
Richard A. Gaggioli
The magnetocaloric energy conversion (ID 97)
Andrej Kitanovski, Jaka Tusek, Alojz Poredos
VOLUME III
THERMO-ECONOMIC ANALYSIS AND OPTIMIZATION
A comparison of optimal operation of residential energy systems using clustered demand
patterns based on Kullback-Leibler divergence (ID 142)
Akira Yoshida, Yoshiharu Amano, Noboru Murata, Koichi Ito, Takumi Hashizume
A Model for Simulation and Optimal Design of a Solar Heating System with Seasonal
Storage (ID 51)
Gianfranco Rizzo
A thermodynamic and economic comparative analysis of combined gas-steam and gas
turbine air bottoming cycle (ID 232)
Tadeusz Chmielniak, Daniel Czaja, Sebastian Lepszy
Application of an alternative thermoeconomic approach to a two-stage vapor compression
refrigeration cycle with intercooling (ID 135)
Atilio Barbosa Loureno, Jos Joaquim Conceio Soares Santos, Joo Luiz Marcon Donatelli
Comparative performance of advanced power cycles for low temperature heat sources
(ID 109)
Guillaume Becquin, Sebastian Freund
xiv
Comparison of nuclear steam power plant and conventional steam power plant through
energy level and thermoeconomic analysis (ID 251)
S. Khamis Abadi, Mohammad Hasan Khoshgoftar Manesh, M. Baghestani, H. Ghalami, Majid
Amidpour
Economic and exergoeconomic analysis of micro GT and ORC cogeneration systems
(ID 87)
Audrius Bagdanavicius, Robert Sansom, Nick Jenkins, Goran Strbac
Exergoeconomic comparison of wet and dry cooling technologies for the Rankine cycle of
a solar thermal power plant (ID 300)
Philipp Habl, Ana M. Blanco-Marigorta, Berit Erlach
Influence of renewable generators on the thermo-economic multi-level optimization of a
poly-generation smart grid (101)
Massimo Rivarolo, Andrea Greco, Francesca Travi, Aristide F. Massardo
Local stability analysis of a thermoeconomic model of an irreversible heat engine working
at different criteria of performance (ID 289)
Marco A. Barranco-Jimnez, Norma Snchez-Salas, Israel Reyes-Ramrez, Lev Guzmn-Vargas
Multicriteria optimization of a distributed trigeneration system in an industrial area (ID 154)
Dario Buoro, Melchiorre Casisi, Alberto de Nardi, Piero Pinamonti, Mauro Reini
On the effect of eco-indicator selection on the conclusions obtained from an
exergoenvironmental analysis (ID 275)
Tatiana Morosuk, George Tsatsaronis, Christopher Koroneos
Optimisation of supply temperature and mass flow rate for a district heating network
(ID 104)
Marouf Pirouti, Audrius Bagdanavicius, Jianzhong Wu, Janaka Ekanayake
Optimization of energy supply systems in consideration of hierarchical relationship
between design and operation (ID 389)
Ryohei Yokoyama, Shuhei Ose
The fuel impact formula revisited (ID 279)
Cesar Torres, Antonio Valero
The introduction of exergy analysis to the thermo-economic modelling and optimisation of
a marine combined cycle system (ID 61)
George G. Dimopoulos, Chariklia A. Georgopoulou, Nikolaos M.P. Kakalis
The relationship between costs and environmental impacts in power plants: an exergybased study (ID 272)
Fontina Petrakopoulou, Yolanda Lara, Tatiana Morosuk, Alicia Boyano, George Tsatsaronis
Thermo-ecological evaluation of biomass integrated gasification gas turbine based
cogeneration technology (ID 441)
Wojciech Stanek, Lucyna Czarnowska, Jacek Kalina
Thermo-ecological optimization of a heat exchanger through empirical modeling (ID 501)
Ireneusz Szczygie , Wojciech Stanek, Lucyna Czarnowska, Marek Rojczyk
Thermoeconomic analysis and optimization in a combined cycle power plant including a
heat transformer for energy saving (ID 399)
Elizabeth Corts Rodrguez, Jos Luis Castilla Carrillo, Claudia A. Ruiz Mercado, Wilfrido Rivera
Gmez-Franco
Thermoeconomic analysis and optimization of a hybrid solar-electric heating in a fluidized
bed dryer (ID 400)
Elizabeth Corts Rodrguez, Felipe de Jess Ojeda Cmara, Isaac Pilatowsky Figueroa
Thermoeconomic approach for the analysis of low temperature district heating systems
(ID 208)
Vittorio Verda, Albana Kona
xv
VOLUME IV
IV . 1 - FLUID DYNAMICS AND POWER PLANT COMPONENTS
A control oriented simulation model of a multistage axial compressor (ID 444)
Lorenzo Damiani, Giampaolo Crosa, Angela Trucco
A flexible and simple device for in-cylinder flow measurements: experimental and
numerical validation (ID 181)
Andrea Dai Zotti, Massimo Masi, Marco Antonello
CFD Simulation of Entropy Generation in Pipeline for Steam Transport in Real Industrial
Plant (ID 543)
Goran Vu kovi , Gradimir Ili , Mi a Vuki , Milan Bani , Gordana Stefanovi
Feasibility Study of Turbo expander Installation in City Gate Station (ID 168)
Navid Zehtabiyan Rezaie, Majid Saffar-Avval
GTL and RME combustion analysis in a transparent CI engine by means of IR digital
imaging (ID 460)
Ezio Mancaruso, Luigi Sequino, Bianca Maria Vaglieco
Some aspects concerning fluid flow and turbulence modeling in 4-valve engines (ID 116)
Zoran Stevan Jovanovic, Zoran Masonicic, Miroljub Tomic
IV . 2 - SYSTEM OPERATION CONTROL DIAGNOSIS AND PROGNOSIS
Adapting the operation regimes of trigeneration systems to renewable energy systems
integration (ID 188)
Liviu Ruieneanu, Mihai Paul Mircea
Advanced electromagnetic sensors for sustainable monitoring of industrial processes
(ID 145)
Uro Puc, Andreja Abina, Anton Jegli , Pavel Cevc, Aleksander Zidanek
Assessment of stresses and residual life of plant components in view of life-time
extension of power plants (ID 453)
Anna Stoppato, Alberto Benato and Alberto Mirandola
Control strategy for minimizing the electric power consumption of hybrid ground source
heat pump system (ID 244)
Zoi Sagia, Constantinos Rakopoulos
Exergetic evaluation of heat pump booster configurations in a low temperature district
heating network (ID 148)
Torben Ommen, Brian Elmegaard
xv i
VOLUME VI
VI . 1 - CARBON CAPTURE AND SEQUESTRATION
A novel coal-based polygeneration system cogenerating power, natural gas and liquid fuel
with CO2 capture (ID 96)
Sheng Li, Hongguang Jin, Lin Gao
Analysis and optimization of CO2 capture in a Chinas existing coal-fired power plant
(ID 532)
Gang Xu, Yongping Yang, Shoucheng Li, Wenyi Liu and Ying Wu
Analysys of four-end high temperature membrane air separator in a supercritical power
plant with oxy-type pulverized fuel boiler (ID 442)
Janusz Kotowicz, Sebastian Stanis aw Michalski
Analysis of potential improvements to the lignite-fired oxy-fuel power unit (ID 413)
Marcin Liszka, Jakub Tuka, Grzegorz Nowak, Grzegorz Szapajko
Biogas Upgrading: Global Warming Potential of Conventional and Innovative
Technologies (ID 240)
Katherine Starr, Xavier Gabarrell Durany, Gara Villalba Mendez, Laura Talens Peiro, Lidia
Lombardi
Capture of carbon dioxide using gas hydrate technology (ID 103)
Beatrice Castellani, Mirko Filipponi, Sara Rinaldi, Federico Rossi
Carbon dioxide mineralisation and integration with flue gas desulphurisation applied to a
modern coal-fired power plant (ID 179)
Ron Zevenhoven, Johan Fagerlund, Thomas Bjrklf, Magdalena Mkel, Olav Eklund
Carbon dioxide storage by mineralisation applied to a lime kiln (ID 226)
Ins Sofia Soares Romo, Matias Eriksson, Experience Nduagu, Johan Fagerlund, Licnio Manuel
Gando-Ferreira, Ron Zevenhoven
Comparison of IGCC and CFB cogeneration plants equipped with CO2 removal (ID 380)
Marcin Liszka, Tomasz Malik, Micha Budnik, Andrzej Zi bik
Concept of a capture ready combined heat and power plant (ID 231)
Piotr Henryk Lukowicz, Lukasz Bartela
xv ii
Cryogenic method for H2 and CH4 recovery from a rich CO2 stream in pre-combustion
CCS schemes (ID 508)
Konstantinos Atsonios, Kyriakos D. Panopoulos, Angelos Doukelis, Antonis Koumanakos,
Emmanuel Kakaras
Design and optimization of ITM oxy-combustion power plant (ID 495)
Surekha Gunasekaran, Nicholas David Mancini, Alexander Mitsos
Implementation of a CCS technology: the ZECOMIX experimental platform (ID 222)
Antonio Calabr, Stefano Cassani, Leandro Pagliari, Stefano Stendardo
Influence of regeneration condition on cyclic CO2 capture using pre-treated dispersed
CaO as high temperature sorbent (ID 221)
Stefano Stendardo, Antonio Calabr
Investigation of an innovative process for biogas up-grading pilot plant preliminary
results (ID 56)
Lidia Lombardi, Renato Baciocchi, Ennio Antonio Carnevale, Andrea Corti, Giulia Costa, Tommaso
Olivieri, Alessandro Paradisi, Daniela Zingaretti
Method of increasing the efficiency of a supercritical lignite-fired oxy-type fluidized bed
boiler and high-temperature three - end membrane for air separation (ID 438)
Janusz Kotowicz, Adrian Balicki
Monitoring of carbon dioxide uptake in accelerated carbonation processes applied to air
pollution control residues (ID 539)
Felice Alfieri, Peter J Gunning, Michela Gallo, Adriana Del Borghi, Colin D Hills
Process efficiency and optimization of precipitated calcium carbonate (PCC) production
from steel converter slag (ID 114)
Hannu-Petteri Mattila, Inga Grigali nait , Arshe Said, Sami Filppula, Carl-Johan Fogelholm, Ron
Zevenhoven
Production of Mg(OH)2 for CO2 Emissions Removal Applications: Parametric and Process
Evaluation (ID 245)
Experience Ikechukwu Nduagu, Ins Romo, Ron Zevenhoven
Thermodynamic analysis of a supercritical power plant with oxy type pulverized fuel
boiler, carbon dioxide capture system (CC) and four-end high temperature membrane air
seprator (ID 411)
Janusz Kotowicz, Sebastian Stanis aw Michalski
VI . 2 - PROCESS INTEGRATION AND HEAT EXCHANGER NETWORKS
A multi-objective optimization technique for co- processing in the cement production (ID
42)
Maria Luiza Grillo Ren, Rogrio Jos da Silva, Mirian de Lourdes Noronha Motta Melo, Jos
Joaquim Conceio Soares Santos
Comparison of options for debottlenecking the recovery boiler at kraft pulp mills
Economic performance and CO2 emissions (ID 449)
Johanna Jnsson, Karin Pettersson, Simon Harvey, Thore Berntsson
Demonstrating an integral approach for industrial energy saving (ID 541)
Ren Cornelissen, Geert van Rens, Jos Sentjens, Henk Akse, Ton Backx, Arjan van der Weiden,
Jo Vandenbroucke
Maximising the use of renewables with variable availability (ID 494)
Andreja Nemet, Jiri Jaromr Kleme, Petar Sabev Varbanov, Zdravko Kravanja
Methodology for the improvement of large district heating networks (ID 46)
Anna Volkova, Vladislav Mashatin, Aleksander Hlebnikov, Andres Siirde
Optimal mine site energy supply (ID 306)
Monica Carvalho, Dean Lee Millar
xv iii
VOLUME VII
VII . 1 - BUILDING, URBAN AND COMPLEX ENERGY SYSTEMS
A linear programming model for the optimal assessment of sustainable energy action
plans (ID 398)
Gianfranco Rizzo, Giancarlo Savino
A natural gas fuelled 10 kW electric power unit based on a Diesel automotive internal
combustion engine and suitable for cogeneration (ID 477)
Pietro Capaldi
Adjustment of envelopes characteristics to climatic conditions for saving heating and
cooling energy in buildings (ID 430)
Christos Tzivanidis, Kimon Antonopoulos, Foteini Gioti
An exergy based method for the optimal integration of a building and its heating plant.
Part 1: comparison of domestic heating systems based on renewable sources (ID 81)
Marta Cianfrini, Enrico Sciubba, Claudia Toro
Analysis of different typologies of natural insulation materials with economic and
performances evaluation of the same buildings (ID 28)
Umberto Desideri, Daniela Leonardi, Livia Arcioni
Complex networks approach to the Italian photovoltaic energy distribution system (ID 470)
Luca Valori, Giovanni Luca Giannuzzi, Tiziano Squartini, Diego Garlaschelli, Riccardo Basosi
Design of a multi-purpose building "to zero energy consumption" according to European
Directive 2010/31/CE: Architectural and plant solutions (ID 29)
Umberto Desideri, Livia Arcioni, Daniela Leonardi, Luca Cesaretti ,Perla Perugini, Elena Agabitini,
Nicola Evangelisti
Effect of initial systems on the renewal planning of energy supply systems for a hospital
(ID 107)
Shu Yoshida, Koichi Ito, Yoshiharu Amano, Shintaro Ishikawa, Takahiro Sushi, Takumi Hashizume
Effects of insulation and phase change materials (PCM) combinations on the energy
consumption for buildings indoor thermal comfort (ID 387)
Christos Tzivanidis, Kimon Antonopoulos, Eleutherios Kravvaritis
Energetic evaluation of a smart controlled greenhouse for tomato cultivation (ID 150)
Nickey Van den Bulck, Mathias Coomans, Lieve Wittemans, Kris Goen, Jochen Hanssens, Kathy
Steppe, Herman Marien, Johan Desmedt
Energy networks in sustainable cities: temperature and energy consumption monitoring in
urban area (ID 190)
Luca Giaccone, Alessandra Guerrisi, Paolo Lazzeroni and Michele Tartaglia
Extended exergy analysis of the economy of Nova Scotia, Canada
David C Bligh, V.Ismet Ugursal
(ID 215)
Feasibility study and design of a low-energy residential unit in Sagarmatha Park (Nepal)
for envirnomental impact reduction of high altitude buildings (ID 223)
Umberto Desideri, Stefania Proietti, Paolo Sdringola, Elisa Vuillermoz
Fire and smoke spread in low-income housing in Mexico (ID 379)
Raul R. Flores-Rodriguez, Abel Hernandez-Guerrero, Cuauhtemoc Rubio-Arana, Consuelo A.
Caldera-Briseo
xix
Optimal lighting control strategies in supermarkets for energy efficiency applications via
digital dimmable technology (ID 136)
Salvador Acha, Nilay Shah, Jon Ashford, David Penfold
Optimising the arrangement of finance towards large scale refurbishment of housing stock
using mathematical programming and optimisationg (ID 127)
Mark Gerard Jennings, Nilay Shah, David Fisk
Optimization of thermal insulation to save energy in buildings (ID 174)
Milorad Boji , Marko Mileti , Vesna Marjanovi , Danijela Nikoli , Jasmina Skerli
Residential solar-based seasonal thermal storage system in cold climate: building
envelope and thermal storage (ID 342)
Alexandre Hugo and Radu Zmeureanu
Simultaneous production of domestic hot water and space cooling with a heat pump in a
Swedish Passive House (ID 55)
Johannes Persson, Mats Westermark
SOFC micro-CHP integration in residential buildings (ID 201)
Umberto Desideri, Giovanni Cinti, Gabriele Discepoli, Elena Sisani, Daniele Penchini
The effect of shading of building integrated photovoltaics on roof surface temperature and
heat transfer in buildings (ID 83)
Eftychios Vardoulakis, Dimitrios Karamanis
The influence of glazing systems on energy performance and thermal comfort in nonresidential buildings (ID 206)
Cinzia Buratti, Elisa Moretti, Elisa Belloni
Thermal analysis of a greenhouse heated by solar energy and seasonal thermal energy
storage in soil (ID 405)
Yong Li, Jin Xu, Ru-Zhu Wang
Thermodynamic analysis of a combined cooling, heating and power system under part
load condition (ID 476)
Qiang Chen, Jianjiao Zheng, Wei Han, Jun Sui, Hong-guang Jin
VII . 2 - COMBUSTION, CHEMICAL REACTORS
Baffle as a cost-effective design improvement for volatile combustion rate increase in
biomass boilers of simple construction (ID 233)
Borivoj Stepanov, Ivan Peenjanski, Biljana Miljkovi
Characterization of CH4-H2-air mixtures in the high-pressure DHARMA reactor (ID 287)
Vincenzo Moccia, Jacopo D'Alessio
Development of a concept for efficiency improvement and decreased NOx production for
natural gas-fired glass melting furnaces by switching to a propane exhaust gas fired
process (ID 146)
Jrn Benthin, Anne Giese
Experimental analysis of inhibition phenomenon management for Solid Anaerobic
Digestion Batch process (ID 348)
Francesco Di Maria, Giovanni Gigliotti, Alessio Sordi, Caterina Micale, Claudia Zadra, Luisa
Massaccesi
Experimental investigations of the combustion process of n-butanol/diesel blend in an
optical high swirl CI engine (ID 85)
Simona Silvia Merola, G. Valentino, C. Tornatore, L. Marchitto , F. E. Corcione
Flameless oxidation as a means to reduce NOx emissions in glass melting furnaces
(ID 141)
Jrg Leicher, Anne Giese
xx
VOLUME VIII
VIII . 1 - ENERGY SYSTEMS : ENVIRONMENTAL AND SUSTAINABILITY ISSUES
A multi-criteria decision analysis tool to support electricity planning (ID 467)
Fernando Ribeiro, Paula Ferreira, Madalena Arajo
Comparison of sophisticated life cycle impact assessment methods for assessing
environmental impacts in a LCA study of electricity production (ID 259)
Jens Buchgeister
Defossilisation assessment of biodiesel life cycle production using the ExROI indicator
(ID 304)
Emilio Font de Mora, Csar Torres, Antonio Valero, David Zambrana
Design strategy of geothermal plants for water dominant medium-low temperature
reservoirs based on sustainability issues (ID 99)
Alessandro Franco, Maurizio Vaccaro
Energetic and environmental benefits from waste management: experimental analysis of
the sustainable landfill (ID 33)
Francesco Di Maria, Alessandro Canovai, Federico Valentini, Alessio Sordi, Caterina Micale
Environmental assessment of energy recovery technologies for the treatment and
disposal of municipal solid waste using life cycle assessment (LCA): a case study of Brazil
(ID 512)
Marcio Montagnana Vicente Leme, Mateus Henrique Rocha, Electo Eduardo Silva Lora,Osvaldo
Jos Venturini, Bruno Marciano Lopes, Claudio Homero Ferreira
How will renewable power generation be affected by climate change? The case of a
metropolitan region in Northwest Germany (ID 503)
Jakob Wachsmuth, Andrew Blohm, Stefan Gling-Reisemann, Tobias Eickemeier, Rebecca
Gasper, Matthias Ruth, Snke Sthrmann
Impact of nuclear power plant on Thailand power development plan (ID 474)
Raksanai Nidhiritdhikrai, Bundhit Eua-arporn
Improving sustainability of maritime transport through utilization of liquefied natural gas
(LNG) for propulsion (ID 496)
Fabio Burel, Rodolfo Taccani, Nicola Zuliani
Life cycle assessment of thin film non conventional photovoltaics: the case of dye
sensitized solar cells (ID 471)
Maria Laura Parisi, Adalgisa Sinicropi, Riccardo Basosi
Low CO2 emission hybrid solar CC power system (ID 175)
Yuanyuan Li, Na Zhang, Ruixian Cai
Low exergy solutions as a contribution to climate adapted and resilient power supply
(ID 489)
Stefan Goessling-Reisemann, Thomas Bloethe
On the use of MPT to derive optimal RES electricity generation mixes (ID 459)
Paula Ferreira, Jorge Cunha
xxi
Stability and limit cycles in an exergy-based model of population dynamics (ID 128)
Enrico Sciubba, Federico Zullo
The influence of primary measures for reducing NOx emissions on energy steam boiler
efficiency (ID 125)
Goran Stupar, Dragan Tucakovi , Titoslav ivanovi , Milo Banjac, Sr an Beloevi ,Vladimir
Beljanski, Ivan Tomanovi , Nenad Crnomarkovi , Miroslav Sijer
The Lethe city car of the University of Roma 1: final proposed configuration (ID 45)
Roberto Capata, Enrico Sciubba
VIII . 2 - POSTER SESSION
A variational optimization of a finite-time thermal cycle with a Stefan-Boltzmann heat
transfer law (ID 333)
Juan C.Chimal-Eguia, Norma Sanchez-Salas
Modeling and simulation of a boiler unit for steam power plants (ID 545)
Luca Moliterno, Claudia Toro
Numerical Modelling of straw combustion in a moving bed combustor (ID 412)
Biljana Miljkovi, Ivan Peenjanski, Borivoj Stepanov, Vladimir Milosavljevi, Vladimir Rajs
Physicochemical evaluation of the properties of the coke formed at radiation area of light
hydrocarbons pyrolysis furnace in petrochemical industry (ID 10)
Jaqueline Saavedra Rueda , Anglica Mara Carreo Parra, Mara del Rosario Prez Trejos,
Dionisio Laverde Catao, Diego Bonilla Duarte, Jorge Leonardo Rodrguez Jimnez, Laura Mara
Daz Burgos
Rotor TG cooled (ID 121)
Chiara Durastante, Paolo Petroni, Michela Spagnoli, Vincenzo Rizzica, Jrg Helge Wirfs
Study of the phase change in binary alloy (ID 534)
Aroussia Jaouahdou, Mohamed J. Safi, Herve Muhr
Technip initiatives in renewable energies and sustainable technologies (ID 527)
Pierfrancesco Palazzo, Corrado Pigna
xxii
ECOS 2012
VOLUME V
Abstract:
The present work investigates the matching of an advanced small scale combined heat and power Rankine
cycle plant to end-user thermal and electric load. The power plant consists of a conc entrated solar power
field co-powered by a biomass furnac e to produce steam in a Rankine cycle, with a Combined Heat and
Power configuration. A hotel was selected as the end user. The power plant design and its operation were
modelled and investigated by adopting transient simulations wit h a hourly distribution. The study of the load
matching of the proposed renewable power technology and the final user has been carried out by comparing
two different load tracking scenarios, i.e. the thermal and t he electric demands. As a result, the power output
follows fairly well the given load curves, supplying, on a selected winter day, about 50 GJ/d of thermal
energy and the 6 GJ/d of electric energy, with reduced energy dumps when matching the load. Furthermore,
for the same winter day, the system allows the reduction of about 4 103 kgCO2 of greenhous e gas
emissions.
Keywords:
Co-powered Concentrated Solar Power, Rankine Cycle, Transient Simulation, Load Matching.
1. Introduction
In recent years the use of Combined Heat and Power (CHP) was commonly considered to supply
energy to end users in the service or residential sectors. The basic argument in favour of CHP is the
possibility to obtain electric and thermal energy in situ, improving the power generation efficiency
and reducing the losses usually related to the energy distribution [1, 2]. Notably among the existing
CHP technologies, only some exceptions are based on the exploitation of different fuels from
natural gas, i.e. small-scale power plants based on biomass derived fuel exploitation, like wood or
biogas [3, 4].
In most applications the main factor which determines the economic viability of CHP schemes is
the high utilisation of heat and electric energy, which are produced simultaneously. Most of the
literature indicates that a CHP plant needs to be fully utilised providing heat and power for a
minimum duty of 4,500 h per annum to gain its breakeven point [5].
When designing renewable energy based CHP technologies, in a distributed generation concept, one
of the key factors is the capability of tracking the time-dependent end-user load. Renewable Energy
Sources (RES), intermittent by nature, produce inconsistently and somewhat unpredictably power
outputs uncorrelated with the end user power demands, typically variable according to predictable
daily load profiles. As a consequence of this mismatch the available RES energy may not meet the
energy demand, resulting in deficit and surplus energy situations.
Several solutions have been proposed to attenuate the RES- user matching inconsistency. The
conventional remedial strategy is to plug the supply gap providing alternative capacity, known as
spinning reserve [6]. Among the solutions devoted to RES electric grid integration, it is worth
mentioning the use of high capacity energy storage to save the produced energy surplus and
postponing the energy surplus delivery [7,8], or combining renewable energy sources with
complementary intermittencies [9].
In this respect, the present study investigates a CHP scheme combining a parabolic trough field for
concentrated solar power (CSP), a thermal energy storage and a biomass furnace as complementary
source. It is worth noting that the biomass source is a sui generis RES, in fact its storage simplicity
permits to customize the power production management, exactly like the fossil fuel sources.
Concerning the parabolic trough field, that device was selected for its high worldwide development
among the CSP systems [10]. Nonetheless, an important aspect of these plants is the size, which, is
usually large. In fact solar trough plants are characterised by multi-MW sizes, which range up to
about 50 MWel for parabolic trough systems. Also the biomass power plants are usually rated in the
range 5100 MW. Even so, while CSP plants size is still growing [11, 12], in the biomass field
there are several applications on small-scale biomass power plants [13, 14].
The aim to exploit CSP technology and limit the plant footprint led to the design of a small scale
plant, recently presented in [15, 16], composed by a 2,580 m2 parabolic trough field, a thermal
energy storage system (TES) and a 1,163 kW biomass furnace to face the solar source fluctuations.
A heat transfer fluid (HTF), i.e. diathermic oil, is heated by the parabolic through field and biomass
furnace and subsequently it is sent to a heat recovery steam generator where it produces low
enthalpy saturated steam that is sent to a 130 kW reciprocating steam engine for the electric energy
production. Moreover, the Rankine cycle (RC) economizer is fed by the exhaust gases derived from
the biomass combustion. A heat recovery for thermal energy production is obtained, using hot water
as heat carrier, in a back-pressure scheme at 134 C and 300 kPa.
The investigations on the proposed RES-based small- scale CHP Rankine cycle plant, when matched
to a typical hotelier end-user were carried out by transient model simulations. The selection of a
hotel as end-user was made for its high heat/electricity consumption ratio. The system matching
behaviour is analyzed for both thermal and electric load tracking with the aim to demonstrate its
capability to meet the end-users energy request on a 24 hour period in a winter day as more
challenging for the solar field performance.
The transient model and the simulations were performed in the TRNSYS environment [17]
supported by the in- house made types of the biomass furnace and reciprocating steam engine and
the STEC component model library [18]. The software TRNSYS was selected as it is a well-known
instrument to model complex energy systems, as demonstrated by several studies appeared in the
open literature which mostly deal on RES applications in a few fields like small- islands stand alone
power systems [8, 19], or, more related to the present paper, on CSP field simulations [20], TES
behaviour in solar trough plants [21] and matching to a hotel end-user [16].
power output deficits and/or furnace start-up problems related to the Direct Normal Insulation
(DNI) sudden variations.
kWth
kWth
kWth
kWel
kWth
Size
1,200
360
1,163
130
1,240
C
kJ/kg K
kPa
300/240
2.36/ 2.19
800
kPa
C
kg/s
kW
kW
2,800/300
230/134
0.51
130
1,100
The HTF circuit supplies the thermal energy to the RC for the production of saturated steam to be
expanded in a 130 kW reciprocating steam engine fitted with an electric generator. According to a
bottomer CHP configuration, the expanded steam is condensed producing a thermal power output
3
available at a constant temperature of 80 C, i.e. the temperature demand of typical district heating
networks. Figure 2 illustrates the temperature-entropy diagram of the Rankine cycle and the
thermodynamic parameters in the reference points.The main components and system
thermodynamic parameters, subdivided in diathermic oil and water/steam circuit, are described in
Table 1. Additional details concerning the power system components are given in [15].
The temperature-heat diagram is shown in Figure 3. The exhaust gas, diathermic oil and watersteam temperatures with the Rankine cycle exchanged heat rate are shown. In particular, two lines
are plotted for the exhaust gas respectively showing the temperature evolution at CSP design
operation with the biomass furnace working at 35% duty rate (Gas-35%), and at 100% of the
biomass heat contribution (Gas-100%). In between these two limiting lines the solar contribution to
the Rankine cycle spans from maximum (Gas-35%) to zero (Gas-100%). Notably, the pinch point
temperature difference for the evaporator is set to 10 C.
406,73
407,15
503,29
503,29
406,73
3,00
28,00
28,00
28,00
[kg/m ]
931,78
932,75
827,10
13,99
1,83
u [kJ/kg]
561,27
562,06
987,39
2604,02
2345,94
h [kJ/kg]
561,60
565,06
990,78
2804,11
2509,69
1,67
1,67
2,61
6,21
6,46
0,0
0,0
0,0
1,0
0,9
T [K]
P [bar]
s [kJ/kg K]
quality
diagram in Figure 4. The present solar-biomass CHP plant is broadly based on a configuration
recently investigated and assessed [15].
The base-line model has been implemented by a control logic targeted to the tracking of different
loads, namely heat or power demands. The development of the load tracking strategy has been
based on the definition of algebraic correlations between the HTF flow rate, directly related to the
RES power input, and the system thermal power output (P th ) or the system electric output (Pel),
respectively. The HTF flow rate was selected as the reference parameter because it governs the
actual power outputs according to the instantaneous renewable energy availability. A sensitivity
analysis, was carried out on the power system configuration by varying
and recording Pel and
Pth values. Figure 5shows the values obtained with the sensitivity analysis (grey lines) and the
corresponding trend lines (black lines) and equations. The HTF control equations, accordingly
derived, read as
,
.
Hotel
43,000
350
8,640
1,656
2,580
12 326
5.23
7.44
Figure 6 shows the monthly distribution of the electric and equivalent thermal load for the selected
end-user; the average daily energy demand (dot sign) is represented in relationship with the daily
average power demand (x sign) and the power demand excursion (bar). It is evident that the electric
energy request has an almost constant behaviour with average daily energy demand always below
200 GJ/day. Whereas the thermal monthly profile has a seasonal connotation which entails a
thermal load range from 250 GJ on the summer period to 1,370 GJ on the winter one. It is worth
noting that generally the average power demand is positioned on the lower part of the power
demand excursion bars, indicating that the energy demand is composed by frequent low power
demand values and rare high power values. This behaviour is highlighted in the summer equivalent
6
thermal load curves (from June to September) of both the end users, when high peaks of cooling
energy are requested during the day.
Legend:
Left axis
DNI [W/m2 ]
938.06
918.06
848.06
560.83
18.33
0
0
0
0
0
0
0
As far as the biomass is concerned, the thermo-chemical characteristics are typical of short rotation
forestry derived woody pellet, with a lower heating value of about 17 MJ/kg and high carbon and
oxygen ratios.
800
Pel,d (kW)
Pth,d (kW)
90
700
80
600
70
500
60
50
400
40
300
30
200
20
100
10
0
12
16
20
0
24
Figure 7. End user electric and thermal load for a typical winter day [22].
Looking at the RC system performance Table 4, the value of 1.2 for the primary energy ratio
demonstrates that the presented solar-biomass Rankine cycle systems can effectively allow the
saving of conventional primary energy sources in each presented scenario. Looking at the electric
output, globally the system produces more electric energy than the need with a peak
production/request ratio of 124% for the electric tracking.
RC system
RES system
Electric Tracking
4,277.53
4,172.09
18,132.39
18.71
990.84
22,304.48
2,064.37
1,664.68
124.01
19.80
0.44
17,291.93
11,656.13
148.35
40.09
7.49
9.26
77.53
11.94
Thermal Tracking
4,277.53
4,092.72
17,221.31
19.20
941.06
21,314.03
2,017.10
1,664.46
121.19
25.80
8.32
16,895.49
11,653.99
144.98
33.27
2.26
9.46
79.27
11.94
1.21
1.24
For the primary energy ratio evaluation, the values for the reference electric and thermal efficiencies are
el
= 0.38 and
th
= 0.8.
configuration, the thermal behaviour is worst, with a high rate of surplus distributed all over the
reference year and a deficit peak during the summer period, as the electric energy request is not
sufficiently high to let the system to produce the requested thermal energy too. The deficit and
surplus events have a quadruple explanation. The first one is that half of the results are loadindependent, e.g. when discussing the electric tracking configuration, the thermal output does not
follow any production law, but is dependent from the electric production trend, without any
correlation to the thermal load. Secondly, in most of the occasions the gaps with the requested load
are entailed to the used correlation among load energy and hot thermal fluid flowrate, which do not
perfectly fit the sensitivity analysis data, conducing to gaps between the desired output and the
obtained one. Nevertheless, those gaps are not particularly remarkable. The third reason, instead,
explains the high surplus peaks that occur, by observing that sometimes there are contemporarily an
elevated available solar supply and full thermal energy storage. In those cases the system, which has
to deliver the collected heat, sends all the hot flowrate directly to the Rankine cycle. The last reason
is that the biomass furnace is always on duty, even if on a minimum rate, supplying energy also in
extremely low energy request.
Figure 8. Hotel electric and thermal power surplus/deficit behaviour during a one year period
under electric and thermal load tracking conditions.
the peak hours and to buffer it (PT ES,d) in the day time when the sun DNI falls below 3,000 kJ/h m2 .
The passage to the electric load tracking logic (Figure 9.b) appears to influence remarkably the RES
power inputs/outputs and the TES charge/discharge cycle. In particular, the TES charge cycle is no
more driven by solar radiation a.m. and p.m. peaks and it is shifted in the afternoon hours when the
overall electric power request reduces. This circumstance causes the shifting of the TES discharge
cycle to the evening time and unbalances the power input from the biomass furnace which is mainly
concentrated in the early morning hours. This finding confirms that the TES and the biomass
furnace have complementary behaviours by implementing an effective reserve to the solar source.
PCSP
PTES,c
800
kW
700
PCSP
PTES,c
800
Pb
Peg
kW
700
PTES,d
600
Pb
Peg
PTES,d
600
500
500
400
400
300
300
200
200
100
100
0
a)
12
16
20
24
b)
hours
12
16
20
24
hours
Figure 9. RES power contribution with the a) thermaland b) electric load tracking matching.
The matching of the power plant with the end-user demand, as driven respectively by the thermal
and electric profile, is described in Figure 10 and Figure 11, by plotting the thermal power output
(Pth ) against the thermal power request (Pth,d) (Figure 10.a and Figure 11.a), and the electric power
output (Pel) against the electric demand (Pel,d) (Figure 10.b and Figure 11.b).
Pth,d
Pth
8 00
kW
200
kW
180
7 00
Pel,d
Pel
160
6 00
140
5 00
120
4 00
100
3 00
80
60
2 00
40
1 00
20
a)
12
16
20
24
4
8
12
16
20
24
b)
hours
Figure 10. Thermal a) and electric b) behaviour with the thermal load tracking matching.
hours
In the thermal load tracking case, Figure 10.a, the thermal load (Pth,d) is completely satisfied by the
solar-biomass plant output (Pth ). The exceeding heat production during the periods of minimum
request is consequent to the control regime of the biomass furnace which is kept at a constant
minimum level. When looking at the electric matching, Figure 10.b, it is remarkable that the power
plant electric output (Pel) mimics the shape of the leading load component. As a result, the correct
11
sizing of the solar-biomass CHP system provides a fair matching in the period of peak electric
request, while the load tracking logic drives the system to an over-production of electricity during
the remaining duty time.
Looking at the electric load tracking case, as a matter of fact, the thermodynamic characteristics of
the solar-biomass CHP system determine the significant overproduction of the thermal power
output when the overall control is given to the electricity production. Figure 11.a shows the electric
peak request in the early morning which, giving rise to the intervention of the biomass, in absence
of any direct or stored solar contribution, results in a large surplus of heat availability. Moving to
the electric matching, Figure 11.b, it is shown that the delivered electric power (Pel) follows fairly
the load (Pel,d) between 4 a.m. and 12 p.m. while keeping it nearly constant in the remaining hours.
800
Pth,d
kW
200
Pel,d
Pel
kW
Pth
180
700
160
600
140
500
120
100
400
80
300
60
200
40
20
100
0
a)
12
16
20
12
16
20
24
24
b)
hours
hours
Figure 11. Thermal a) and electric b) behaviour with the electric load tracking matching.
5. Environomic issues
here the environmental and economic aspects will be analyzed. An effect of the application of this
system are the entailed Greenhouse Gases (GHG) emission savings related to the solar fraction,
estimated by means of emission factors related to the Italian thermoelectric power stations at
reference year 2003 [22]. The emissions savings are evaluated considering the entire electric energy
supply, in the hypothesis of grid transfer of the surplus, and the fraction of thermal energy supplied
to the end users, in the hypothesis of dump of the thermal energy surplus
The result is a higher emission saving in the thermal tracking scenario, which avoids the emissions
of 661 ton/y of carbon dioxide.
Table 5. Global emission savings for a typical winter day from solar fraction.
Electric tracking
Thermal tracking
CO2 [ton/y]
555.29
661.00
SOx [ton/y]
0.58
0.69
NOx [ton/y]
0.35
0.41
TSP [ton/y]
0.02
0.03
Another essential environmental aspect is the land use of the plant. Considering the net land use,
Table 6, the plant needs about 13,000 m2 , nevertheless, taking into account security distances and
the need of space for the power conversion block the needed surface amounts to 31,000 m2 . Even if
12
the solar field accounts for the 51.56% of the global footprintit is meaningful to reflect on the higher
specific power of the parabolic trough field, i.e. 0.46 kWp /m2 , when compared to the most
commercial photovoltaic power plants (0.17 kWp /m2 ) referring only to the devices surface area.
Concerning the plant costs, Table 7 indicates that the parabolic trough field with the thermal energy
storage are the most expensive devices of the proposed system. In particular, the capital cost of a
solar trough field with thermal storage has been evaluated in 4,820 $/kW for the reference year
2006 [10]. It is worth noting that these data refer to large CSP technologies and must be considered
only as a rough estimate of the present CSP device. Referring to the other technologies, the capital
costs have been obtained by private communications with producers. In the utilities heading, it
entails costs for electric panels, electric and hydraulic connections, civil works &c.
It is obvious that such high costs are constraining to the development of the proposed system when
thinking to the standard fossil fuel based power technologies. Nevertheless, in a fossil fuel free
power generation perspective, given from the exhaustion of fossil energy sources and from the need
to pull down the fossil sources related emissions, the current high costs become a side issue in
behalf of the sustainable development of the energy sector.
Table 6. Plant land use.
Solar field
TES
Biomass furnace, filter and stack
Biomass storage
Buildings (Rankine cycle elements, desalting units, offices)
Total
Table 7. Plant estimated capital costs.
Technology
CSP field with TES
Biomass furnace
Economizer
Evaporator
Steam engine
Condenser
Utilities
Total
Cost []
7,870,000
130,000
15,000
45,000
220,000
15,000
300,000
8,595,000
6. Conclusions
A model of a combined solar-biomass CHP plant devoted to feed an hotelier end- user was
presented. The well-established TRNSYS software was adopted for transient simulation.
An analysis of thermal and electrical power production on a yearly basis demonstrated the
feasibility of the present configuration in satisfying the energy requirements of the hotel using a
fully renewable and sustainable approach.
Furthermore the model was matched with a thermal and an electric winter day load in transient
simulations. The results for the two different load tracking scenarios were compared in terms of
delivered power, matched load, RC system efficiencies and global GHG emission savings.
13
When looking at the output performance, the results show a most suitable behaviour for the thermal
load tracking scenario, as it delivers both electric and thermal energy with less gap from the enduser requested energy.
The Primary Energy Ratio values, in both electric and thermal tracking cases, indicate the capability
of the system to save energy in comparison of two separated plant for the single electric and thermal
energy production. Nevertheless, as the high plant capital costs, i.e. 7.2 k$/kW, are mostly related to
the solar section, the improvements must be oriented to the exploitation of low-tech solar field
entailing the passage from parabolic troughs to Compound Parabolic Concentrators (CPC) and the
adoption of Direct Steam Generators (DSG) systems with supercritical steam Rankine cycles.
Nomenclature
CHP
CSP
DNI
Eel
Eg
Eth
HTF
Pb
Pb,min
PCSP
Peg
Pel
Pel,d
P
PT ES,d
Pth
Pth,d
RC
RES
TES
el
th
14
References
[1] Maidment, G.G., Zhao, X., Riffat, S.B. and Prosser G., Application of combined heat-andpower and absorption cooling in supermarkets, Applied Energy 63 (1999) 169-190.
[2] Maidment, G.G. and Tozer, R.M., Combined cooling heat and power in supermarkets,
Applied Thermal Engineering 22 (2002) 653-665.
[3] Corsini, V. Naso, G. Mattei, P. Venturini. Biomass co-firing: analysis of the main technica l
problems in coal power plants. 15th European Biomass Conference, 7-11 May, Berlin,
Germany.
[4] Corsini, V. Naso, G. Mattei, P. Venturini. Biomass co-firing: estimation of fuel requirements
and land needed to feed some Italian coal power plants. 15th European Biomass Conference,
7-11 May, Berlin, Germany.
[5] Eastop T.D. and Croft D.R., Energy efficiency for engineers and technologists (1st ed.),
Longman Scientific and Technical, Halow, Essex, 1990, p. 335.
[6] Frhlke, K. and Haidn, O.J., Spinning reserve system based on H2 /O2 combustion. Energy
Conversion, S0196-8904(96)00128-8.
[7] Klouwani, S., Agbossou, K. and Chahine R., Model for energy conversion in renewable
energy system with hydrogen storage. Journal of Power Sources 140 (2005) 392-399.
[8] Corsini, A., Rispoli, F., Gamberale, M., and Tortora, E., 2009, Assessment of H2- and H2Obased renewable energy-buffering systems in minor islands, Renewable Energy 34 (2009)
279288.
[9] Sinden, G. Environmental Change Institute University of Oxford, The practicalities o f
developing renewable energy stand-by capacity and intermittency. Submission to The Science
and Technology Select Committee of the House of Lords, 2004.
[10] Sargent & Lundy LLC Consulting Group Chicago, Illinois, Assessment of Parabolic
Trough and Power Tower Solar Technology Cost and Performance Forecasts, NREL/SR-55034440, October 2003.
[11] Jun Li, Scaling up concentrating solar thermal technology in China, Renewable and
Sustainable Energy Reviews 13 (2009) 2051-2060.
[12] Al-Soud, M.S. and Hyayshat E.S., A 50 MW concentrating solar power plant for Jordan,
Journal of Cleaner production 17 (2008) 625-635.
[13] Dong, L., Liu, H. and Riffat, S., Development of small-scale and micro-scale biomass
fuelled CHP systems. A literature review, Applied Thermal Engineering 29(2009) 2119-2126.
[14] Badami, M. and Mura, M., Preliminary design and controlling strategy of small-scale wood
waste Rankine Cycle (RC) with a reciprocating steam engine (SE), Energy 34 (2009) 13151324.
[15] Borello, D., Corsini, A., Rispoli, F. and Tortora E., A combined solar-biomass Rankine
cycle concept for small-size cogeneration, Proceedings of ECOS 2009 Conference.
[16] Borello, D., Corsini, A., Rispoli, F. and Tortora E., Load matching for a combined solarbiomass Rankine cycle plant Proceedings of ASME-ATI-UIT 2010 Conference.
[17] Klein, S.A., Beckam, W.A., Mitchell, J.W., Braun, J.E., Evan,s B.L., Kummert J.P., et al.,
2000, TRNSYS a transient system simulation program. Version 15.1, Madison: Solar
Energy Laboratory, University of Wisconsin; 2000.
[18] Schwarzbzl, P., Eiden, U., Pitz-Paal, R., Jones, S., 2002, A TRNSYS model library for
solar thermal electric components (STEC). A reference manual. Release 2.2.
[19] Corsini, A., Gamberale, M., and Rispoli F., 2006, Assessment of renewable energy
solutions in an Italian small island energy system using a transient simulation model, ASME
Journal of Solar Energy Engineering 2006;128:23744.
15
[20] Jones, S.A., Pitz-Paal, R., Schwarzboezl, P., Blair, N. and Cable R., 2001, TRNSYS
modeling of the SEGS VI parabolic trough solar electric generating system, Proceedings o f
Solar Forum 2001: Solar Energy: The Power to Choose, April 21-25, 2001, Washington, DC.
[21] Kolb, G.J. and Hassani, V., 2006, Performance of thermocline energy storage proposed for
the 1 MW Saguaro solar trough plant, ISEC2006-99005.
[22] Macchi, E., Campanari, S., Silva, P., La micro-cogenerazione a gas naturale, Polipress Politecnico di Milano, 2005.
[23] National Action Plan for Energy Efficiency Sector Collaborative on Energy Efficiency,
Hotel Energy Use Profile, EPA Summer Workshop Report, 2007.
[24] SEL, 25 July 2003, Generated Hourly Weather Data. Solar Energy Laboratory, University of
Wisconsin-Madison. 12 November 2008 <http://sel.me.wisc.edu/trnsys/weather/generate.htm>.
[25] Ronnelid, M., Perers, B., Karlsson, B., On the factorization of incidence angle modifiers
for CPC collectors, S0038-092X(97)00016-9, Solar Energy, 1997.
16
Abstract:
A parabolic trough solar c ollector is improved the efficiency by a novel design of compound parabolic trough
solar collector where the aim is three-fold. Firstly, one aim is to achieve day-long collection efficiency without
the need for mechanical tracking of the sun. Secondly, the collector must be designed to operate efficiently
under diffuse solar irradiation as experienced for example in rainforest climate. Thirdly, one seeks to achieve
as a high an output temperature as possible. Newly developed system consists of multiple compound
parabolic troughs facing the sun at different angles. The s alient feature of this design is that the system can
collect the sunlight energy at every angle without any moving parts at the same time can receive the diffused
light, the maximum efficiency of the collector is 32% and has an ability to achieve high output temperature,
the maximum temperature at header of evacuated tube is 235 degrees Celsius, and is therefore suitable for
high temperature application such as industrial uses or cooling application.
Keywords:
solar energy, compound parabolic trough, non-tracking solar collector.
1. Introduction
A parabolic trough is a type of solar thermal energy collector which is generally used in solar power
plants. The solar collector is constructed as a long parabolic trough with a tube running its length at
the focal point. Sunlight is reflected by the trough and concentrated on the tube filled with synthetic
oil, which heats to 300-400 degrees Celsius [1-5]. The trough is usually aligned on a north-south
axis, and rotated to track the sun as it moves across the sky each day. Therefore it seems
unavoidable that there needs to be a tracking system that follows the position of the sun.
The disadvantage of the parabolic trough solar collector is that concentrating systems require sun
tracking to maintain sunlight focus at the collector. The tracking system increases the cost,
complexity and the maintenance cost due to the moving parts. This type of solar collector is not
preferred in a small residential house. Another problem is an inability to provide power in diffused
light conditions, which is due to the fact that the power output from concentrating systems drops in
cloudy conditions. As Thailand has a tropical rainforest climate, which causes the ratio of diffused
solar radiation to global solar radiation to be rather high (in the range of 31% to 58%) [8], one faces
a serious problem in utilizing such a solar collector to collect solar energy, especially in rainforest
climate.
A parabolic trough solar collector is improved the efficiency by a novel design of compound
parabolic trough solar collector which does not contain a solar tracking system and has an ability to
collect diffused sunlight by using compound parabolic troughs facing the sun at different angles [67]. The non-tracking parabolic trough solar collectors were presented in ref. [8-20]. The advantage
of this design is that there are no moving parts in the system, which leads to reductions in the cost
and maintenance. This collector yields higher temperatures than flat plate solar collector and could
be used in the residential house, the maximum temperature at header of evacuated tube is 235
degrees Celsius, and is therefore suitable for high temperature application such as industrial uses or
cooling application.
17
2. The Model
In order to design and develop the non-tracking solar collector, the mathematical model of
reflection of compound trough is calculated. Let the shape of a parabolic trough be described by the
curve y = f(x) on the x-y plane in Fig. 1. The law of reflection states that the angle of incidence is
equal to the angle of reflection relative to the tangent of the curve y = f(x) at any point (x,y). The
slope of this tangent line at point (x,y) is denote by mt = df(x)/dx, the slope of the incident ray by m0
and the slope of the reflected ray by m1.
Fig. 1. The reflection of a light ray by a curve y = f(x). is represented an angle of incidence and
an angle of reflection. m t, m0 and m1 are slope of a tangent line, an incident ray and a reflected ray
respectively.
From trigonometry [5], the relationship between the angle between two lines and their relative
slopes mt, m0 and m1 is given as
mt m0
m1 mt
tan
,
(1)
1 mt m0 1 mt m1
which yields a slope of the first reflected ray m1 as
m1
m0 mt 2 mt m0
.
mt2 2mt m0 1
Similarly, the ith reflected rays can be calculated by using the relation
mt mi 1
mi mt
,
1 mt mi 1 1 mt mi
(2)
(3)
where i are integers. From Eq. 1 and Eq. 2, the reflection of a parabolic trough can be simulated as
shown in Fig. 2.
Fig. 2. The reflection of parabolic trough solar collector at incident angle of 75 degrees where
blue and orange lines are incident and 1st reflected rays respectively. The circle is the position of
the focus point.
18
For the incident angle of 75 degrees, the conventional parabolic trough in Fig. 2 cannot receive the
reflected rays. Therefore it needs solar tracking system to maintain sunlight at the focus point. The
parabolic trough solar collector is designed to have an ability to achieve day- long collection
efficiency without the need for mechanical tracking of the sun by using 3 compound parabolic
troughs facing the sun at different angles. Using Eq.(1-3), the reflection of non-tracking solar
collector at various time are shown in Fig. 3.
12.00 am
1.00 p m
2.00 p m
3.00 p m
Fig. 3. The reflection of three-compound parabolic trough solar collector where blue, orange,
green and yellow lines are incident, 1st reflected, 2nd reflected and 3rd reflected rays respectively.
The circle in each trough is the position of evacuated tube.
The 3-compound parabolic trough shows that it has an ability to receive the sunlight at various time.
For 12.00 a.m., the solar collector can collect all reflected rays, the reflected rays in the middle
trough are concentrated at the lowest position of the tube and for both side of the middle trough, the
reflected rays are concentrated on the higher position inside the tube. When the time changes, the
reflected rays move up and down inside a tube. For this principle, this collector can collect the
sunlight in any time. However there are some ray losses when the time changes especially after 3.00
p.m. which could be ignored because of very low solar power.
The collector is designed to have an ability to collect diffused light. In Fig. 4, compound parabolic
trough can receive the incident rays in the period of 80 degrees. This implies that this collector has a
probability to collect incident rays from sunlight in both direct and diffused light in the period of 80
degrees at the same time while a conventional parabolic trough can collect the incident rays which
are nearly perpendicular to the trough. Although a parabolic trough could provide a high
concentration, the parabolic trough could not work effectively under diffused light conditions. The
experimental results have shown that the efficiency of the new design of solar collector is higher
than parabolic trough under diffuse solar irradiation as shown in Fig.10 and Fig. 11.
19
80
10
Fig. 4. The reflection of light rays at various angles of the incident rays. This design has an ability
to collect incident rays in the period of 80 degrees while the conventional parabolic trough can
receive the incident rays in the period of 10 degrees.
In this paper, SUNDA vacuum tubes (SEIDO1) are used to receive the concentrated light
from the trough. This tube is composed of flat plate absorber as shown in Fig. 5.
Fig. 5. The method to place an evacuated tube with flat plate absorber in compound parabolic
trough.
From Fig. 5, the flat plate absorber which is placed horizontally can receive reflected rays better
than the flat plate absorber which is placed vertically and cross shape absorber can collect all rays
but there are no cross shape absorber product at the moment. For this reason, flat plate absorber is
considered to place horizontally in each trough.
20
3. Experiment
The solar collector in Fig. 3 has been invented consisting of three compound parabolic troughs
made of stainless sheets, oriented at different angles. The solar collector has an overall width of 1 m
and a length of 1.9 m, and the evacuated tubes (SUNDA vacuum tube, (SEIDO1)) are placed along
its axis. These evacuated tubes are connected to a manifold header pipe and connected with the
pump to feed the oil. The flow rate is set at 5 lpm. The collectors are fixed on Earth and aligned
along the north-south direction as shown in fig (6-7).
Fig. 6. The novel non-tracking solar collector has an overall width of 1 m and a length of 1.9 m.
HE
AD
ER
W
S
N
E
Fig. 7. Diagram of test arrangement.
The experiment was performed in Bangkok, Thailand. The data was taken during the period of 9.00
a.m. to 4.00 p.m. on the 10th , 11th , 12th ,13th and 14th January 2012, The sky was not very clear
which lead the solar power is not smooth in any time. The diagram of test arrangement is shown in
fig. 7.
When the evacuated tubes absorb the sunlight from troughs, the heat from the tubes is transferred to
hot oil which flows in the system. The energy of the system can be calculated by [21]
QC t
mC Tout Tin
21
(4)
where t represents time, m and C are flow rate and the specific heat of the thermal oil respectively.
The efficiency of the system in any time is
t
QC
,
Qin
(5)
where Qin is the solar power. The evacuated tube is placed in the trough and measured the
temperature at the header. The maximum temperature at heat pipe is 235 degrees Celsius as shown
in Fig. 8 and the maximum temperature of hot oil is 180 degrees Celsius for 0.5 litres of oil as
shown in Fig. 9.
Fig. 8. The maximum temperature at the header of evacuated tube plotted against time from 8.00
a.m. to 5.00 p.m. on the 12th December 2011.
Fig. 9. The hot oil temperature plotted against time from 9.00 a.m. to 4.00 p.m. on the 14th
November 2011. The maximum temperature is 180 degrees Celsius for 0.5 litres of oil.
From the experiment, the solar power on the 11th ,12th ,13th and 14th of January 2012 in Bangkok
had been collected and its average is shown in Fig. 10. The results show that the efficiency of the
new-design solar collector at any time is fairly constant, which is similar to the parabolic trough
with solar tracking system, while the efficiency of a conventional parabolic trough at any time
distributes like a Gaussian curve having its maximum at around 11.30 a.m. as shown in Fig.11. The
three-compound parabolic trough solar collector yields higher temperature than flat plate or
evacuated tube solar collector. The average efficiency of solar collector is 25-32% .
22
50
800
40
600
30
Efficiency
Solar Power W m
1000
400
10
200
20
0
10 :00
12 :00
14 :00
16:00
10:00
12:00
14:00
16:00
Time
Time
Fig. 10. The average solar power and efficiency of 3-compound parabolic trough plotted against
times in the period of 9.00 a.m. to 4.00 p.m. on the 10th , 11th ,12th ,13th and 14th January 2012 in
Bangkok.
Fig. 11. The parabolic trough in Fig. 4 has been invented. The average efficiency of parabolic
trough plotted against time from 9.00 a.m. to 2.00 p.m. on the 4th, 6th and 8th January 2010[9]
4.Conclusions
The new-design of solar collector has an ability to collect the sunlight at every angle, similar to the
parabolic trough with a solar tracking system. This solar collector has an ability to receive the
diffused light, and this make it suitable for using in all kinds of climate. There are no moving parts
in the system, which results in the reductions in the cost of the system, the cost of maintenance and
complexity. This collector needs only 3 evacuated tubes while SUNDA collector (SEIDO1) needs 8
tubes at the same area. This collector yields higher temperatures than flat plate or evacuated tube
solar collector. The maximum temperature at heat pipe is 235 C and oil temperature is 180 C. It is,
therefore, suitable for high temperature application such as industrial uses or cooling application.
5.Acknowledgements
23
The authors would like to thank the National Research University Project of CHE, the
Ratchaphiseksomphot Endoment Fund (Project No. EN1180I), 2-V research program of National
Research Council of Thailand (NRCT) and Energy Research Institute of Chulalongkorn University
for the financial supports. We also would like to thank Mr. Narong Amornpitakpunt, AMP
METALWORKS [Thailand] Co.,Ltd for his help for inventing the 1st and 2nd prototype of solar
collector.
References
[1] Kearney, D., 1989. Solar Electric Generating Stations (SEGS). IEEE Power Engineering
Review (IEEE) 9, 4-8.
[2] Mills D., 2004. Advances in Solar Thermal Electricity Technology. Solar Energy 76, 19-31.
[3] Hodge B. K., 2010. Alternative Energy Systems and Application. John Wiley & Sons, Inc.
[4] Frank Kreith, D. Yogi Goswami, 2007. Handbook of Energy efficiency and Renewable
Energy, CRC Press.
[5] Milton Matos Rolim, Naum Fraidenraich and Chigueru Tiba, 2009. Analytic Modeling of Solar
Power Plant with Parabolic Linear Collectors. Solar Energy 83, 126-133.
[6] Wattana Ratismith, Parabolic Troughs Solar Collector with no Need of Solar Tracking System,
December, 9 2009 patent no.: 0901005526 ( patent pending)
[7] Wattana Ratismith, Compound Parabolic Troughs Solar Collector with no Need of Solar
Tracking System, July, 21 2011 patent no.: 1101001216 ( patent pending)
[8] Wattana Ratismith, Novel Parabolic Troughs without Solar Tracking System, Proc. Renewable
Energy 2010, Pacifico Yokohama, Japan, (2010)
[9] Wattana Ratismith and Urith Archakositt, Parabolic Troughs without Solar Tracking System,
Third International Conference on Applied Energy, Perugia, Italy (2011)
[10] Standard Mathematical Table, 25th Ed., CRC Press, Inc., 1978.
[11] Zambolin E., Del Col D., 2010. Experimental analysis of thermal performance of flat plate and
evacuated tube solar collectors in stationary standard and daily conditions, Solar Energy 84,
1382-1396.
[12] Grass C., Schoelkopf W., Staudacher L., Hacker Z., 2004. Comparison of the optics of nontracking and novel types of tracking solar thermal collectors for process heat applications up to
300 C. Solar Energy 76, 207-215.
[13] Winston R., and Hinterberger H., 1975. Principles of Cylindrical Concentrators for Solar
Energy. Solar Energy 17, 255-258.
[14] Winston R., 1974. Principles of Solar Concentrators of a Novel Design, Solar Energy 16, 89-95.
[15] Blanco J., Malato S. et al, 1999. Compound Parabolic Concentrator Technology Develop to
Commercial Solar Detoxification Applications. Solar Energy 67, 4-6.
[16] Rabl A., O'Gallagher J. and Winston R., 1980. Design and Test of Non-Evacuated solar
Collectors with Compound Parabolic Concentrators. Solar Energy 24, 335-351.
[17] Rabl A., Bendt P., and Gaul H. W., 1982. Optimization of Parabolic Trough Solar Collectors.
Solar Energy 29, 407-417.
[18] Pei Gang, Li Guiqiang, Zhou Xi, Ji Jie, Su Yuehong, 2012. Experimental study and exergetic analysis of a
CPC-type solar, Solar Energy 86, 12801286.
[19] M. Souliotis, P. Quinlan, M. Smyth, Y. Tripanagnostopoulos, A. Zacharopoulos, M. Ramirez, P. Yianoulis.
2011. Heat retaining integrated collector storage solar water heater with asymmetric CPC reflector, Solar
Energy 85, 24742487.
24
[20] O. Helal, B. Chaouachi, S. Gabsi, 2011, Design and thermal performance of an ICS solar water heater based
on three parabolic sections, Solar Energy 85, 24212432.
[21] Duffie J. A. and Beckmen W. A, 1991. Solar Engineering of Thermal Process, New York,
Wiley.
25
Abstract:
Exploiting low enthalpy geot hermal resources (hot water at 45 to 90 C) is increasing its attractiveness, due
to their higher widespread compared to high enthalpy ones. Anyway, at these temperature levels, their
application is almost solely related to district heating, due to the marginal efficiency achievable when
eventually converted to electricity by a power plant. The possibility of applying an A bsorption Heat
Trans former (AHT) to enhance low enthalpy geothermal resources available in the range from 45 t o 90 C
and produce electricity with an acceptable efficiency was investigated. A thermodynamic model of an AHT
working with a Lithium Bromide (LiBr) solution was developed. It showed the possibility of enhancing low
temperature heat, ranging from 45 to 90 C, to the higher value heat at 130 140 C with a COP variable
within 40 to 50%. The enhanced heat can be used as the hot source of an Organic Rankine Cycle (ORC),
which exploits the geothermal heat (available at 80 90 C) in summer time. Generally the geothermal
resource supplies district heating in winter, but when heat demand resets, it becomes attractive to convert a
fraction to electricity instead of leaving it waste. Even if the primary heat lost in the A HT (having COP in the
40 to 45% range) is less than compensated by the inc reas e in the ORC efficiency, leading to a lower
system efficiency, the adoption of an A HT is wort h up to about 80 100 C, because geothermal heat
source temperatures below 100C are not suitable for the current ORC technology When the temperature
of the heat source is higher t han 100 C, the direct use of the geothermal res ource into the ORC bypassing
AHT is more performing and convenient.
Keywords:
Absorption Heat Transformer (A HT), Organic Rankine Cycle (ORC), Low Enthalpy Geothermal.
Several analyses and optimization studies of AHT can be found in the technical literature [1-6; 1113]. In the present study, the work of Horuz and Kurt [3] is taken as a fundamental reference, and a
complete thermodynamic model was written in a simulation environment which includes the link to
common absorption fluid mixtures [14]. According to the existing literature [2-3;12-13] the main
assumptions made are the following:
The system is analyzed assuming steady state and thermodynamic equilibrium conditions.
Pressure drops and heat losses (or gains) are neglected.
The mechanical work consumed by the pumps is neglected.
The evaporator and generator temperatures are the same.
The refrigerant at the condenser and evaporator outlets are all saturated.
Steam
Evap orator
T [C]
140
130
120
110
100
90
80
70
60
50
40 3
30
20
10 2
0
0
1
Condenser
10
s [kJ/kg-K]
Figure 1a-1b: Absorber Heat Transformer layout (1a, left) and thermodynamic behavior of steam
at evaporator and condenser on T s diagram (1b, right).
It must be noticed that, as shown in [3], evaporator temperature higher than generator temperature
leads to a very slight increase of COP, especially for temperatures higher than 80C.
Furthermore, the temperature difference between the strong solution at the Absorber inlet (T10 ) and
the weak solution at the absorber outlet T5 is fixed at 10 C, as in [3]. Finally, the refrigerant mass
flow rate (m1 ) is fixed at 0,1139 kg/s, as in [3].
The thermodynamic analysis was carried out calculating the energy and the mass balance for each
AHT component. For the mass balance, the balance at the Generator must be highlighted:
(1)
(2)
Where x7 and x8 are the lithium bromide concentration in the weak and strong solution respectively.
An important design parameter is the flow ratio (f), which is defined as the ratio between the strong
28
solution mass flow rate (m8 ) and the refrigerant mass flow rate (m1 ). Hence, considering (1) and (2),
the flow ratio can be defined as:
(3)
The performance of the AHT is measured by the COP, which is the ratio between the heat delivered
from the absorber and the heat supplied to the evaporator and generator [14]:
(4)
Table 1 shows the thermodynamic conditions of the AHT cycle points (as referred on figure 1a), in
case of TAB=130C, TGE=TEVA= 80C, TCO=25C, while Table 2 reports the heat exchanged at the
absorber, evaporator, condenser, generator, the weak and strong solution concentration and the COP
in various cases.
Table 1. AHT thermodynamic states if TAB=130C, TGE=TEVA= 80C, TCO=25C.
1
2
3
4
5
6
7
8
9
10
T [C]
100
0
-100
1
0.5926
0.5926
0.5926
0.6565
0.6565
0.6565
m [kg/s]
0.1139
0.1139
0.1139
0.1139
1.171
1.171
1.171
1.057
1.057
1.057
P [bar]
0.03169
0.03169
0.4737
0.4737
0.4737
0.4737
0.03169
0.03169
0.4737
0.4737
T [C]
80
25
25
80
130
98
98
80
80
120
h [kJ/kg]
2650
104.8
104.8
2643
291.5
226.9
226.9
227.6
227.6
299.2
s [kJ/kg-K]
8.875
0.367
0.3669
7.611
0.7226
0.5564
0.5564
0.4204
0.4204
0.6126
29
Unit
C
C
C
C
C
kW
kW
kW
kW
kW
Case 1
65
25
45
45
55
30.25
0.4108
0.5314
196.9
282.2
282.4
197.1
77.84
0.4642
0.4795
Case 2
115
25
70
70
105
17.77
0.4667
0.5656
251.8
287.2
287.7
252.3
136.3
0.5779
0.6104
Case 3
155
25
90
90
145
11.5
0.4855
0.5895
275.6
291
292
276.7
120
0.6457
0.7018
0,64
0,56
270
260
COP
COPAHT, Horuz
QAB [ kW ]
280
0,6
250
Horuz
0,48
240
0,44
17
21
TCO [C] 25
29
33
present study
T CO [C]
230
18
21
24
27
30
Figure 2: COP(left) and Absorber heat (right) vs the variation of the condenser temperature.
In Fig. 2 the values for COP and Q AB obtained with the presented model are compared with the
values given in [3]. COP values show a good agreement, with an average relative error of 1%, while
for QAB the average relative error is slightly higher (3,5%). It is due to the strong influence that
small error on the Li- Br concentration in the strong solution value has on Q AB. However, taking into
account that our model is developed with a software having probably different (and unknown to the
reader) thermodynamic libraries of Li-Br than the one used in [3], the average error may be
considered acceptable.
On the other hand, Fig. 3 shows that if the evaporator - and thus the generator temperature raises,
also the COP and the absorber output heat increases. In fact, in this case the maximum pressure
increases and the LiBr concentration in the weak solution decreases, leading to a lower enthalpy
value at the absorber outlet (point 5, Fig. 1) and then to a higher absorption heat and COP.
290
0,5
0,495
QAB [kW]
0,49
COP
0,485
270
0,48
0,475
260
pr esent study
0,47
0,465
0,46
0,455
74
Horuz
250
COPAHT, Horuz
76
78
80
82
84
86
88
90
92
T GE = T EV [C]
240
74
76
78
80
82
84
86
88
90
Figure 3: COP(left) and Absorber heat (right) vs the variation of the generator temperature.
30
92
Also in this case our results are compared with [3]: COP values are approximately in agreement,
confirming an average error of 1%, while Q AB values show, once again, higher discrepancies , with
an average error of approximately 3,5%.
Finally, Figs 4 and 5 show, respectively, the variation of the COP and QAB with the absorber
temperature at different generator (i.e. evaporator) temperatures. The increase of the absorber
temperature leads, obviously, to decrease of COP and Absorber released heat. In fact, a higher TAB
implies a higher enthalpy value of the weak solution at absorber outlet (point 5) which means a
lower heat released at the absorber. At each fixed value of TGE, an absorber temperature which
optimizes COP and Q AB is found.
0,5
0,49
0,48
COPAHT
0,47
TCO=25 [C]
0,46
0,45
TGE=45 [C]
TGE=70 [C]
0,44
TGE=50 [C]
TGE=80 [C]
TGE=60 [C]
TGE=90 [C]
0,43
0,42
0,41
40
50
60
70
80
90
TAB [C]
100
110
120
130
140
150
160
290
QAB [kW]
270
250
TCO=25 [C]
T GE=45 [C]
230
T GE=50 [C]
T GE=70 [C]
T GE=80 [C]
210
T GE=60 [C]
190
40
50
60
70
80
90
100
110
120
130
T GE=90 [C]
140
150
160
T AB [C]
Figure 5: Absorber heat vs absorber temperature at different generator (evaporator) temperatures.
31
The results shown in Figs 4 and 5 are in a good agreement with those presented in [13]. It can also
be observed that the optimized values of COP and Q AB can be kept almost constant at relatively
high values, provided that the decrease of the heat source temperature is followed by an adequate
reduction of the temperature at the absorber.
4. ORC system
T [C]
The Organic Rankine Cycle (ORC) is a market proven technology [15-18]. In this work we
considered a standard ORC (Fig. 6), where a secondary, low boiling-point working uid (i.e.
R245fa) is vaporised (points 16-18), then it expands in a turbine, and it is subsequently condensed
in the condenser before being returned to the economiser and the evaporator by the cycle pump. An
internal heat exchanger (IHE) is added to recover the heat from the turbine outlet stream, in order to
improve the cycle efficiency. Thermodynamically, the power plant works with a pure Rankine
cycle, without superheater.
180
170
160
150
140
130
120
110
100
90
80
70
60
50
40
30
20
10
0
16
18
19
4,79 bar
15
14
1,1
20
21
1,2
1,3
1,4
1,5
1,6
1,7
1,8
1,9
s [kJ/kg-K]
Figure 6: layout of the proposed Organic Rankine Cycle (ORC, left) and the related thermodynamic
cycle (right)
released at the absorber of the AHT previously described. In the Geo-ORC system, for a given
temperature of the geothermal resource (Tsource, i.e. T11 ), the temperature of the organic fluid at
point 18 is calculated by (5), where the Approach temperature difference ( Tapp;ORC) is fixed by
technical constraints at the heat exchanger at 10C:
(5)
Furthermore, since the geothermal flow rate is given as an input data, the temperature difference at
the pinch point ( Tpp;ORC) is fixed at 5 C by the following equation:
(6)
The energy balance at the vaporizer and at the economizer are then given by the following
equations, where (7) provides the organic fluid flow rate (mORC), while (8) calculates the reinjection
temperature of the geothermal resource (Tgeo;out ):
(7)
(8)
In the second system, i.e. the ORC-AHT, the geothermal resource is used in the evaporator and
generator, which work at the same temperature. Since Q GE (or QEV) is fixed to the operating value
obtained by the AHT pure model, mgeo and TGE (or TE V) are given by the following equations:
(9)
(10)
The geothermal resource temperature at the outlet of the generator (or evaporator) is calculated by:
(11)
From (9) it can be stated that, as the geothermal resource temperature increases, T GE, and then Q GE,
rises, leading to an augmentation of mgeo . The geothermal mass flow rate calculated by (9) is used
as input in the GEO-ORC system, i.e. in (7) and (8). Approach temperature difference at the
generator ( Tapp;GE) is fixed at 10C, while the temperature difference at the geothermal resource
outlet from generator ( Tout;GE) is fixed at 5C. Finally, in the ORC-AHT system, the heat Q AB is
coupled with the vaporizer and economizer. A heat carrier fluid is used between the two systems.
However, as the temperature of the heat source and the evaporation temperature of the organic fluid
are constant and respectively equal to TAB and T18 , the temperature difference ( Tapp;ORC) between
TAB and T18 is fixed by technical constraints at the heat exchanger at 10C. Hence, the energy
balance at the vaporizer and economizer is calculated by the equation (12), where the temperature at
point 18 is fixed using (13):
(12)
(13)
33
GEO-ORC
80
115
25.74
90
135
26.48
100
155
26.95
110
160
27.61
120
160
26.32
60
23.6
80
25.74
100
26.95
120
26.32
539
556.1
568
583.6
558
493
539
568
558
105
125
145
150
150
50
70
90
110
14.16
1.478
21.27
1.478
30.84
1.478
33.81
1.478
33.81
1.478
3.432
1.478
6.097
1.478
10.09
1.478
15.74
1.478
48.7
52.7
51.2
47.4
47.4
32.2
38.3
44.5
49.9
75
85
95
105
115
55
75
95
115
1.11
14
23.2
6.5
15.6
0.467
0.672
35
1.147
16
27
7.7
17.9
0.479
0.664
42.5
1.17
17.2
30.4
8.4
20.1
0.486
0.662
47
1.229
17.2
31.2
8.5
22.2
0.50
0.712
49.6
1.11
17.2
31.2
8
24.2
0.469
0.775
45
2.39
5.5
10.5
5.5
10.5
27
2.51
9.1
15.6
9.1
15.6
49
2.56
12.1
20
12.1
20
68
2.45
14.5
24.2
14.5
24.2
81
84
99.54
114
129.5
135
52
84
114
135
252
25
217
288
31.6
266.3
30.3
223.7
289.9
90
276
29
228
292
40.6
289.4
26.2
239.8
294.2
110
262
24
217
296
40.4
16
466
11.4
32
490
19.5
47
499
26.8
58
477
33.1
34
This analysis suggests that, with the proposed ORC, the adoption of AHT to upgrade the
geothermal heat is worth for low temperature of geothermal resource, up to 85 100 C.
6. Conclusions
The possibility of using an AHT system with a LiBr water solution to upgrade the low
temperature heat available from geothermal resources has been discussed. The upgraded heat is
used to feed an ORC working with R245fa in a pure Rankine recuperative power cycle.
Both systems are modeled by calculation codes developed with the same calculation tool, which
includes the thermodynamic libraries of LiBr solution and ORC working fluid. The results showed
that the COP of the AHT and the amount of upgraded heat (Q AB, for a fixed value of m1 ) are
strongly dependent on the three main design temperatures, i.e. TCO (close to the environmental
value), TGE and TEV (close to the available geothermal source), and TAB (the value of upgraded
heat). Specifically, for temperatures of low grade heat variable between 45 and 90 C and the
corresponding upgraded heat variable into the 55 155 C range, the achieved COP varies between
38 and 50%. The condenser (i.e. environmental) temperature has a slight influence on AHT
performance: with fixed upgraded heat from 80 C (TGE=TEV) to 130 C (TAB ), the COP is reduced
from about 48% to about 45% for an increase of condenser temperature from 15 to 32 C. The
results agree with those found on literature [3].
The enhanced heat released by the absorber of the AHT can be used into the boiler of an ORC,
which has the possibility to use the geothermal heat source, which otherwise would be unexploited,
and thus wasted. However, the convenience of AHT ORC combined system over the simple GEO
ORC is strongly related to the temperature of the geothermal heat source. Generally, coupling
AHT ORC is convenient until the geothermal resource temperature is below 120C, which is the
lowest temperature suitable for current ORC technology.
Nomenclature
COP
T
LiBr
m:
p:
Q:
T:
x:
W:
[1]...[n]:
Suffixes
app;GE
app;ORC
source:
geo;pp:
geo;out:
geo:
out;GE
Coefficient of Performance
Temperature difference [C]
Lithium Bromide
Mass flow rate [kg/s]
Pressure [bar]
Heat rate [kW]
Temperature [K]
Lithium bromide mass fraction in the solution
Power [kW]
Thermodynamic point of the cycle
Approach difference to GE
Approach difference to ORC
Geothermal inlet to the system
Geothermal at the pinch point
Geothermal reinjection into the well
Geothermal
Geothermal outlet from GE
35
p:
pp;ORC
sp
t:
Acronyms
AB
AHT
CO
EVA
EX
GE
GEO-ORC
ORC
ORC-AHT
IHE
Pump
Pinch point ORC
Specific
Turbine
AHT Absorber
Absorption Heat Transformer
AHT Condenser
AHT Evaporator
AHT Liquid to liquid heat exchanger
AHT Generator
ORC directly fed by the Geothermal resource
Organic Rankine Cycle
ORC coupled with AHT
Internal Heat Exchanger
References
[1] Tufano, V., Heat recovery in distillation by means of absorption heat pumps and heat
transformers, Applied Thermal Engineering 1997; II(2): 171-178.
[2] Zhao, Z., Ma, Y., Chen, J, Thermodynamic performance of a new type of double absorptio n
heat transformer, Applied Thermal Engineering 2003;23: 24072414
[3] Horuz,I., Kurt, B., Absorption heat transformers and an industrial application, Renewable
Energy 2010;35: 2175-2181
[4] Xuehu Ma, Jiabin Chen, Songping L., et alii, Application of absorption heat transformer to
recover waste heat from a synthetic rubber plant, Applied Thermal Engineering 2003;23: 797
806
[5] Rivera, W., Romero, R.J, Thermodynamic design data for absorption heat transformers: Part 7 :
Operating on an aqueous ternary hydroxide, Applied Thermal Engineering, 1998;18(3-4): 147156
[6] Best, Rivera, W. Thermodynamic design data for absorption heat transformers: Part 6 :
Operating on water-carrol, Heat Recovery Systems and CHP, 1994;14(4): 427-436
[7] Lentz, A., Almantza, R., Solargeothermal hybrid system, Applied Thermal Engineering
2006;26: 15371544
[8] Fangtian Sun, Lin Fu, Shigang Zhang, Jian Sun, New waste heat district heating system with
combined heat and power based on absorption heat exchange cycle in China, Applied Therma l
Engineering 37 (2012) 136 144
[9] Pratt and Whitney Corp., Model
280 kW PureCycle Power System,
http://www.pw.utc.com/media_center/assets/model_280_purecycle_power_system.pdf
[10] ElectraTherm's
Heat
to
Power
Generation
Systems
available
at
http://www.electratherm.com/products.html last access 10 February 2012
[11] Yin J, Ming-Shan Z, Li- Zhong H. Performance analysis of an absorption heat transformer
with different working uid combinations. Applied Energy 2000;67(3):281-92.
[12] Sozen A, Yucesu HS. Performance improvement of absorption heat transformer. Renewable
Energy 2007;32:267-84.
36
[13] Scott M., Jernqvist A., Olsson J., Aly G. Experimental and theoretical study of an open
multi-compartment absorption heat transformer for different steam temperatures. Part I :
hydrodynamic and heat transfer characteristics. Applied Thermal Engineering 1999;19: 279298.
[14] EES Engineering Equation Solver, http://www.fchart.com/.
[15] Di Pippo R. Geothermal Power Plants: Principles, Applications and Case Studies. London,
UK: Elsevier Advanced Technology; 2006.
[16] Turboden Combined Heat And Power Orc Units For The Pellet Industries, 2008, (see also
http://www.genewscenter.com/content/detail.aspx?releaseid=7229&newsareaid=2http://www.t
urboden.eu/en/public/press/Turboden_ORC_for_pellets_english.pdf)
[17] Obernberger I, Hammerschmid A, Biomass fired CHP plant based on an ORC cycle
Project ORC-STIA-Admont. Final Report, Bios-energy systems; 2001 (see also
http://www.bios-bioenergy.at/uploads/media/Report-ORCAdmont-Thermie-2001-03-26.pdf)
[18] GE Energy Announces Industrial Waste-Heat Recovery Innovation for Onsite Power Plants,
http://www.genewscenter.com/content/detail.aspx?releaseid=7229&newsareaid=2, 6 July, 2009
(last access 4 January 2012).
37
Abstract:
A large share of the vegetable oil market is today shifting from food to energy sector. This trend led to
discussions about the chance of a confliction between these two chains and its possible social and
environmental consequences. Moreover the increasing interest in pure vegetable oils for energy and
biodiesel production promotes market volatilities and, in recent years, high and peak prices. The economics
of the biodiesel production is suffering the actual market situation.
In this framework, the present work presents the results of the OVEST project, which is supported by the
Tuscany region. Aim of the project is the investigation of the possibility to use waste oil streams of the food
industry to obtain an alternative feedstock for bioliquids and biofuels production. In the current food market, a
number of waste streams are available from the food industry: oleins and fatty acids, sludge oils and WCO
(Waste Cooking Oils) represent several examples of these waste streams from edible oil production. Most of
these feedstock oil streams are composed by distilled fractions obtained during vegetable oil refining, i.e.
free fatty acids, di- and mono-glyceride, etc. In the OVEST project, vegetable oil is obtained from waste
materials by the re-esterification process, in order to obtain a new feedstock for biodiesel production and for
power generation.
A market analysis has been carried out in OVEST, to estimate the economic viability and the potential of
each waste streams. PFAD (Palm Fatty Acid Distillate), fatty acids from vegetable oils and WCO are very
interesting feedstock to obtain re-esterified oils. The new vegetable oils obtained in OVEST show promising
chemical-physical characteristics for biodiesel production, such as the fatty acid composition and the
contamination from solids and inorganics.
Test on a modified Micro Gas Turbine (30 kWel) have been carried out. Performances and pollutant missions
have been monitored. The results of the experimentation confirmed the viability of these materials for the
energy production.
From the environmental point of view, the use of these feedstock allows to meet the requirements of the
European Directive EC/28/2009 (RED), that introduces minimum targets in terms of GHGs saving for
bioliquids and biofuels. According to the EC Directive, LCA must be carried out considering equal to zero the
wastes and residual products GHG emissions before processing.
The present work showed that most of the input materials to the esterification plant may be classified as
wastes or residues, and that assuring an accurate traceability for each treatment and transport phase is a
really complex issue. However the GHG savings indicated in the RED for bioliquids from wastes, was
confirmed by the work here carried out.
Therefore OVEST project demonstrates the viability of the use of re-esterified oils (reconstructed oils) for
biodiesel production.
Keywords:
Vegetable Oil, organic waste, conversion technology.
1.
INTRODUCTION
The European biofuels market is rapidly growing in recent years and it can be defined by three
main actors: biodiesel (78.5% on energy base), bioethanol (17.5%) and pure vegetable oil
(4.0%). Biogas is rapidly growing in EU market; the application to the transport sector is often
proposed but still represent a minor component.
38
Biodiesel is the most important biofuel in European region, both for consumption and for
production, representing about 78% of the total biofuels market in the transport sector. The EU27
production, in 2008, was 6,860 million liters, for a total consumption of 9,465 million liters
[USDA, 2008]. Germany, France and Italy are the leaders of this market and together accounted
for 80% of the EU biodiesel production.
The recent issued Renewable Energy Directive (RED [EU Parliament, 2009]) confirms the 10%
target, set by the EC, for energy from renewable sources in transport. This 10% is a minimum
target to be achieved by all Member States for the share of biofuels in transport petrol and diesel
consumption by 2020. Even if the EU directive poses a rigid framework for the fuel sustainability,
focusing on the GHGs reduction potential, the main reasons to set such ambitious target deals with
the EU fuels supply security.
From this point of view, it is necessary to underline that demand is not growing only in Europe.
China set a target of biodiesel penetration for 2020 and New Zealand, Australia, Japan, among the
others, have mandates for biofuels blends [New Zealand Parliament, 2008]. Biofuels worldwide
consumption accounted for 24.4 Mtoe in 2006, compared to only 10.3 in 2000. Despite this
considerable increase, they still represent only 1.5% of the total road-transport fuel demand in
2006 [IEA, 2008] and rooms are today present for an increase in the demand.
As regards the feedstock, in contrast with the United States, where about 50% of biodiesel is
produced from soybean oil, rapeseed oil and sunflower oils represent the major source of biodiesel
for the European industry (4.7 Mton [Pelkmans, 2009]).
The issue of the cost is today one of the critical issue for the biodiesel producers. During the entire
2007 and part of 2008 the price of the food oils rose up, constituting an example of the price
instability of this kind of market. Causes have been identified in the rapid growth of the Asian
market, in the temporary decrease in the yield of several crops, due to unfavorable climatic
conditions, in the increasing demand for biofuels and in the speculation on food and fuel
commodities, the crude oil price to 147 US$/barrel [US EIA, 2009]. Nevertheless price instabilities
still continue to appear cyclically and they can be still considered the major risk factor for transport
fuel market.
The total world vegetable oil production has increased in the last decades: the main growth has
been recorded for palm oil (+8% annually), followed by palm kernel oil (+7.6% annually) and
soybean oil (+5.7% annually) [Pelkmans, 2009]. Palm oil and soybean oil represent more than half
of the market, being 30% and 28% respectively. Rapeseed oil is also important, especially for
Europe [Lieberz, 2009].
The vegetable oil consumption is divided between food industry and other industrial applications,
among which, the biodiesel industry is rapidly rising in importance. Nonetheless more than 80% of
the vegetable oil produced worldwide is destined to the food market.
The sector of vegetable oil trade is today dramatically changing. In the framework of the
international scenario, European Union vegetable oil market is strongly affected by biodiesel
sector. Despite palm oil is often perceived as the major actor in the biofuel market, the production
of biodiesel from this feedstock is very limited due to the poor winter stability of the product. An
increasing use of palm oil is observed as bioliquid for co-generative systems (high displacement
reciprocating engine for power generation with heat recovery), or in substitution of rapeseed oil for
food industry.
Despite the intensive productions, European Union (EU-27) is a net importer of oil. The total
import of oil for food and industrial uses is estimated equal to 34% of the total domestic
consumption [USDA, 2009].
The OVEST project aims to investigate alternative feedstock for biodiesel and direct energy
production starting from food industry by-products and wastes. The main goal of the project are
related to the techno-economic demonstration of the viability of the use of this low quality
feedstock to produce energy in stationary systems. Moreover the lower costs of this materials
39
(figure 1), compared to the standard oils, are interesting as a potential answer for the biodiesel
sector.
2.
Many waste materials are today available from different markets. The possibility to consider these
waste streams as potential feedstock for energy production, it is very interesting topic. Feedstock
from waste materials are profitable in the actual context as an open discussion is on-going about
the land use and the competition between the food and the energy chains for bioenergy sector.
Supported by the Tuscany Region, the OVEST project focuses its attention on several possible
feedstock rich in fatty-acids content for oil production. There are several by-products of the
processing of edible vegetable oils; there are mainly fatty-acids and oleins, Waste Cooking Oils
(WCO) and sludge oils (the residues of tanks washings). The esterification process allows to obtain
tri-glycerides from these materials, creating a product useful for both energy of biodiesel
industries.
3.
NON-STANDARD FEEDSTOCK
Oleins and Fatty-Acids are by-products of the same food chain: the de-acidification of the edible
vegetable oils. This process is carried out to obtain a low acid-content edible oil, usually lower
than 1%. Oleins and Fatty-acids are the results of different techniques. Oleins are obtained by
treating a soapy paste, resulting from the chemical de-acidification of the oil, with sulfuric acid.
This kind of chemical de-acidification produces a material with over the 50% of FFA (Free FattyAcids) content [Dumont M-J et al. 2007]. Fatty-acids are secondary products of distillation
(physical de-acidification) of vegetable oils: their FFA content is very high (over 80%) [Hartman
L. 1978], with also a normally lower content of impurities and contaminant than oleins.
While the previous described feedstock may present a relatively little fluctuation of their FFA
content, different is the case of WCO and sludge oils. The properties of WCO are highly
influenced by the kind of heating treatment they undertook. In the case of the sladge oils, their
property are linked with the characteristics of the tanks they came from.
WCO can be a high quality feedstock if well pre-treated and in particular filtered for the removal
the solid impurities [Balat M. 2010].
The most interesting qualities of this kind of feedstock is related to their cost. This is generally
lower than the Pure Vegetable Oil. The price of this class of materials is subjected to volatility like
the price of the oil, nevertheless this it has been noticed that these fluctuations are smaller, which
make these substances interesting even in critical market situations.
Figure 1- Recent oil market price trend (CPO: crude palm oil; PFAD: palm fatty acids distilled).
40
In this framework, the sludge oils have to be considered for being very available and cheap,
compared to the other by-products. Nevertheless their utilization may be not easy in the actual
technologies, as their quality is poor. They can be used by a proper dilution with other purer
materials.
Particular is the case of WCOs; they represent a good-quality feedstock to start with, their cost is
similar to PFAD and they are available in large quantities even on regional scale. WCO has a high
potential which needs to be supported by a well-coordinated collection mechanism. An interesting
example is the experience carried out in Tuscany (Italy). In this Region, in the past years, the
collection of WCO was limited to the restaurants. Today it has been powered and WCO are also
available from a domestic collection system. The availability of this material has increased
strongly and it is still developing, so it may become possible the implementation of a highly
positive environmental short-chain embracing the attention to social and environmental
sustainability proposed by the European RED Directive (Renewable Energy Directive
EC/28/2009).
The actual market price is guiding the use of these non-standard feedstock. WCO is widely used as
feedstock for biodiesel production, even because the European directives allow to associate at this
material a double positive effect on the GHGs reduction. Sludge oils are more interested for the
use in engines for the direct power generation, as their quality is lower compared to the other
materials but also their costs.
4.
RE-ESTERIFICATION PROCESS
The esterification process allows to obtain a new vegetable oil from FFA and glycerol. This
process is well known in the biofuels area to be also a pre-treatment stage for the biodiesel
production.
Controlling the stoichiometric ratio between Free Fatty Acids and glycerol it is possible to re-built
an oil which is not edible (because of the reaction may alter the chemical residues from the
previous processing) [Bhosle B. M. 2004] that can be directed to the energy chain without
conflicting with the food sector. The feedstock required is largely available on international
markets (for example Palm Fatty Acid Distilled) and on regional scales.
The technology described in the present work is a non-catalyzed reaction, which will take place in
two separate reactors. In the first stage the glycerol and the FFA are introduced in the reactor and
then heated up at 200-250 C for two hours. The glycerol evaporates with the water produced by
the reaction itself, so it is necessary to discharge water outside while force the glycerol to
condensate back in the reactor, so to be re-circulated.
Depending on the kind of feedstock used, the esterification process needs some upstream
treatments on the feedstock, in order to obtain a high quality oil. To reduce the impurities and
contaminant: for example, on the WCOs, a severe filtration is required; de-phosphoration and
dehumidification are advised in any case.
The esterified oil could not immediately suitable for power generation or as feedstock for biodiesel
production. It is often necessary to provide further processes, like post-filtrations and
deodorization.
5.
Stationary power generation is an interesting market for the feedstock described. Standard
technology for energy conversion are the reciprocating engines, based on diesel cycle and the gas
turbine. The chemical and physical characteristics of the re-esterification process are different from
fossil diesel oil and rather far from the common technical specifications for engines. Major
41
differences are the kinematic viscosity and the Lower Heating Value, but also as regards
contaminant levels and composition, fuel cold properties, ignition behavior, etc.
A standard for the use of vegetable oils in engines is the DIN 51605 (table 1). It has been defined
for the use of rapeseed oil in transports engine. It is a reference to qualify the characteristics of the
vegetable oil for the power generation.
Table 1 DIN51605 technical specification of the use of vegetable oil in engines.
As previously stated, the sludge oils are an interesting class of materials for power generation. The
oil obtained by the sludge has presented the highest chemico-physical characteristics variation,
especially in terms of impurities and contaminants, because of the wide variation of the original
material. In general the quality of this oil is far from the technical specification defined in the
standard. In table 2 is reported the concentration of several important elements in the vegetable oil.
The high value shown confirm the difficulty to directly use these oils for engine application.
Their use is interesting when diluted to highest quality feedstock in order to reduce the cost of the
final product.
Table 2 - Metal content in oil from sludge .
Limits
max
Min
Mean
Max
Min
Mean
Sulfur
[ppm]
18,40
0,26
7,99
Calcium and magnesium
[ppm]
83,10
44,50
58,18
Vanadium
[ppm]
<1
Copper
[ppm]
1
0,1
-
For what concerns the other considered feedstock, the best results has been noticed for the fatty
acids and WCOs.
42
A
3
B
1.5
C
1.5
Humidity (%
d.b.)
Acidity (mg
KOH/g)
5.44
3.03
4.82
2.98
2.62
2.69
In OVEST project the re-esterified oil has been tested in a 30 kW Micro Gas Turbine. This
machine has been modified to be fed by this oils. In particular the fuel line has been adapted to the
characteristics of the oil in input. The functioning resulted very stable and the performances
substantially unmodified. After 250h of stable functioning many parts of the engine have been
examined. In particular analysis have been carried out on the deposits of the injectors. The
concentration of phosphorous in the oil appeared to be the most critical parameter as it was the
main constituent of the deposits.
The analysis of the pollutant concentration in the exhaust gases from the turbine confirmed the
good quality of the combustion. The concentration of NOx resulted to be higher when turbine
works with re-esterified vegetable oil compared with regular Diesel but CO concentration
appeared to be significantly lower. In figure 2 is shown a direct comparison with Diesel on the
pollutants emissions.
Figure 2: MGT C30 pollutant emission measured for pure vegetable oil feeding.
6.
The economical potential benefits in the use of these nonstandard products is not the only positive
effect. In the present work the LCA has been carried out starting with the definition and
classification of the feedstock. The classification as waste or by-product is a crucial point to assess
the environmental balance: according to the European RED Directive (EC/28/2009) definitions,
the emission of CO2 associated to the original chain are not taken into account if the product in
exam is a waste. This is an crucial starting point to meet the 35% minimum target of GHGs saving.
In order to verify the achieving of the RED target the CO 2eq calculation basing on the methodology
proposed by the directive has been implemented. Moreover, considering that the energetic
consumptions and therefore the fossil energy requirements are significant along the reesterification process also the CER parameter has been evaluated.
43
Besides the work provide to identify the production chain and the geographical origins of the
feedstock (especially when classified as by-product) and to define clearly the stages of
production and eventually the LCA methodology. The phases modeled (see, figure 3) in the LCA
carried out can be summarized in the following steps:
Transportation to the esterification reactor;
Re-esterification phase;
Use of oil for energy production.
Figure 3 - Scheme for the definition of the new oil in LCA analysis.
The LCA has been carried out through a spreadsheet built using the Biograce project assumptions:
particularly the Biograce emission factors have been considered (see table 4).
Table 4 - Emission factors used in the LCA (Source: Biograce).
GHG
g/g
CO2
1
CH4
23
N2O
296
Transport
Road transport
(liquid materials)
Ship transport
Train transport
MJ/tkm
1.01
gCO2/tkm MJ/tkm
88.34
1.1693
0.12
0.21
10.85
26.81
Re-esterification
Diesel
Natural gas
Grid electricity MV
Grid electricity LV
gCO2/MJ
87.64
67.58
127.65
MJ/MJ
1.16
1.13
2.70
129.19
2.73
0.1437
0.5660
The transport phase has been evaluated considering the geographical provenience of the current
feedstock treated in the Silo SpA, hypothesizing the means of transport and estimating the average
distance.
44
For the re-esterification process the energetic consumptions have been measured in the plant of the
Silo SpA (see table 5): the values measured do not highlight a difference due to the feedstocks
characteristics.
Finally, for the energy production a 250 kWel engine with 41% of electrical efficiency and 85% of
thermal efficiency has been considered.
As shown in table 6, the potential GHGs saving is near 90%. The important results allows to meet
the requirements of the directive EC/28/2009 on sustainability.
Moreover, the results obtained highlight that the transport phase have higher GHG emissions than
these expected by the RED because of the far distances to be covered for delivering the feedstocks
to the re-esterification plant. On the other hand the emissions are comparable to these indicated in
the directive for the biodiesel production from waste materials.
Table 5 - Energy consumption for the production phase
Power consumption
Methane consumption
47.81
48.72
kWh/t
kg/t
Than has been calculated the equivalent CO2 and the CER parameters, which consider the fossil
energy consumptions for the process.
Table 6 - Results of LCA
Phase
Trasport
Esterificaztion
Total
Saving for cogenation
gCO2/MJ_oil
2.54
4.53
7.08
92%
CER analysis
Process
Trasport
Esterification
Total
MJ/MJ_oil
0.0364
0.0783
0.11
As shown in table 5, the potential GHGs saving is near 90%. The important results allows to meet
the requirements of the directive EC/28/2009 on sustainability.
7.
CONCLUSIONS
B iodiesel
industry and power generation applications are today suffers for the market price and the
price volatility of the traditional feedstock. Alternative feedstock are available from food industry,
mainly from edible oils processing. By-products are oleins and the acid oils. Sludge oils and Waste
Cooking Oils (WCO) are also interesting materials, rich in free fatty acids, that can be converted
into re-esterified oils.
A new non-edible oil has been obtained thanks to the esterification process, adding glycerol to the
inlet materials. The characteristics of the re-esterified oil are variable, mainly based on the
characteristic of the feedstock. The oil could need further treatments but are mostly compatible
with the use in engines and for the biodiesel industry. The lower quality for engine feeding is an
aspect that can be improved, especially for very low cost feedstock, such as the sludge oils.
Test has been carried out in a 30 kW micro gas turbine. Good performances have been observed
for re-esterified feeding.
45
From technical and economical point of views the best results were given by Palm Fatty Acid
Distilled for the biofuel production. A good potential is for the Waste Cooking Oils, available also
at regional level.
For power production in engine the cost is a key factor and sludge oils are very interesting even if
the poor qua lity is today limiting their use.
The LCA resulted very interesting for the oils obtained from this process: the potential GHGs
saving resulted higher than 90% compared to the traditional fossil fuels, so the target (35%) set by
EU Directive can be reached.
ACKNOWLEDGMENT
The authors wish to acknowledge the Tuscany Region, for the support, project coordinator SILO
SpA and the project partner Il Trebbiolo.
The authors would also to acknowledge Eng. Lisa Bigozzi for her precious support.
BIBLIOGRAPHY
[1] Balat M, Potential alternatives to edible oils for biodiesel production A review
of current
work, Energy Conversion and Management vol. 52, Elsevier Journal, Ottobre 2010.
[2] Bhosle B. M, Subramanian R, New approaches in deacidification of edible oilsa review,
Journal of food engineering vol. 69, Elsevier Journal, November 2004
[3] Dumont M-J, Narine S, soapstock and deodorizer distillates from North American vegetable
oils: Review on their characterization, exrtation and utilization, Food Reserch International
n 40, Elsevier Journal, June 2007.
[4] EU Parliament. DIRECTIVE OF 2008/28/EC OF THE EUROPEAN PARLIAMENT AND
OF THE COUNCIL, 2009.
[5] Hartman L, Deacidification of edible oils by short path distillation, Rivista- Italiana-delleSostanze-Grasse Vol. 55, 1978.
[6] IEA, World Energy Outlook, 2008.
[7] Lieberz S.,EU27 Oilseeds annual, Global agricultural information network, NLE49039, 2009.
[8] New Zealand Parliament, Biofuel bill, www.parliament.nz, 2008.
[9] Pelkmans, F., Walter, F., Rosillo-Calle, C. (2009) A global overview on vegetable oils, with
reference to biodiesel, IEA Task 40, report.
[10] USDA. "United states department of agriculture foreign agricultural service".
www.fas.usda.gov, 2008.
[11] US Energy Information Administration, Policy research working pape, tonto.eia.doe.gov,
2009.
[12] USDA, EU27 biofuels annual, Global agricultural information network, NL9014, 2009.
46
University of Bremen, Faculty of Production Engineering & artec | Research Center for Sustainability
Studies, Bremen, Germany, tillz@uni-bremen.de
Abstract:
Wind energy converters are a major technology for the generation of power with even increasing relevance
in the future. The shares of wind energy in the grid mix es of many countries are growing steadily. This
growth is not only due to new c onverters on new sites but also due to repowering of sites already in use as
well as updating of old wind energy converters. There is a huge pot ential for repowering and updating but
since wind energy is supposed to reduce anthropogenic impacts on the environment, it is important to
consider environment al aspects, for repowering as well as for updating. The methodological approach
described in t his paper allows identifying the environmentally preferable option from a variety of product
scenarios using the specific energy demand as an exemplary indic ator, while other indic ators can be
included, too. The method facilitates life cycle data regarding the converters energy balance or depending
on the desired indicators- other figures from the converters life cycle inventory. The described assessment
can also be performed c onsidering case (and site) specific factors. However, it has t o be noted that data
demand and availability might occur as a burden.
Keywords:
Wind energy converters, Wind turbines, Energy demand, Repowering, Updating, Sustainability.
1. Introduction
Wind energy is a technology of increasing importance the installed capacity and the total number
of installed wind energy converters (WEC) are steadily growing, in Europe as well as globally [1,2].
As a consequence thereof, repowering and updating of old converters have an increasing relevance
within the wind energy sector. While various LCA studies have been performed for new WEC,
environmental assessments of specific repowering scenarios or of updated second-hand WEC are
rarely found. Still, the question about the environmental benefits from updating or repowering old
WEC should not be discounted for. Also, the question whether updating or repowering is the
environmentally favorable option for a particular old WEC might be of increasing relevance in the
future. To address these aspects a method for comparing different product scenarios has been
developed based on a previous work that focused exclusively on assessing repowering scenarios [3].
This method uses the specific energy demand as an exemplary indicator; including other indicators
is possible, too. To demonstrate the method two exemplary assessments are performed illustrating
use and possible outcomes of an assessment.
47
Germany as the leading European country in terms of installed wind power capacities has an
installed capacity of more than 27,000 Megawatts [4] contributing already a significant share to the
German grid mix. This share is constantly growing in Germany as well as in other European
countries. The European Wind Energy Association, for example, has the announced target of 190
GW onshore and 40 GW offshore installed in 2020 and 250 GW onshore and 150 GW offshore in
2030 respectively [1]. This is accompanied by individual goals in different European countries. The
German government for example has the goal to increase the share of renewable energies in power
supply to 35% in 2020 [5]. A steady growth can also be observed globally [2]. Here, the installed
capacity grew within the past years already by 22.9% (mid-2010 to mid-2011) and 23.6% (mid2009 to mid-2010) respectively [2]. This growth is not only due to newly installed wind energy
converters but it is also caused by repowering or updating of already existing converters and the
installation of second hand converters, respectively. Repowering means the replacement of older
WEC by new state-of-the-art WEC, hereby improving the utilization of existing sites and reducing
the total number of installed converters. Since in most countries the most suitable sites for onshore
wind energy are already in use, this is the main motivation for repowering. Also potential land use
conflicts are hereby reduced. Updating of WEC includes refurbishing of an existing WEC,
replacement of obsolete parts and possibly a slight increase of its rated power.
Germany, for example, has a potential for repowering of about 25,000 MW, which equals
approximately the rated power of WEC already installed in 2010 [6,7]. The worldwide development
of wind energy indicates that also in many regions outside of Germany repowering is going to be an
issue as well as updating of WEC.
As besides economic interests- the reasons for using wind energy are to a large extent
environmental aspects, it is important also for repowering and updating, that they results in an
overall improved environmental performance. Choosing the (ecologically and environmentally)
preferable option from a variety of available repowering and updating scenarios as well as finding
the (ecologically and environmentally) right moment in time for the respective measures is a
challenging task. The methodological approach, described in the next section addresses this task.
3. Methodological approach
To assess the respective product scenarios we used a method comparing the specific energy yield of
the different product scenarios to each other. Here, a product scenario is regarded as the
combination of an old WEC that reaches the end of its life span and a repowering WEC or an
updated version of the old WEC. If, for example, a 0.6 MW WEC reaches the end of its life span
and there are two repowering options (A and B) available with updating/ refurbishing being another
option (C), this results in three different product scenarios. Additionally, the repowering options
might become available at different points in time, resulting in different repowering points:
a) 0.6 MW WEC + repowering WEC A at point A
b) 0.6 MW WEC + repowering WEC B at point B
c) 0.6 MW WEC + updated/refurbished 0.6 MW WEC at point C
For the identified (generic) product scenarios the specific energy demand is compared in the next
step. This can be done calculative as well as graphically, which will be demonstrated in the next
section. If required, the method can be used for additional impact categories like the carbon
footprint (as demonstrated in [3]) or aggregated measures of several impact categories depending on
the availability of the required data. Using several different impact categories might lead, of course,
to inconclusive results (or at least to results that might need intensive discussion), since there will
not always be an option that minimizes all assessed impact categories. Focusing on one impact
category on the other hand will provide more conclusive results but discounts other relevant
aspects. Still, this relevance always depends on the goal of the study that also might give indications
on which impact categories should be assessed with priority.
48
cf x t
GWPold tr
Enet, old tr
GWPx (tx )
Enet, x tx
Other impact categories or result indicators like the abiotic depletion potential which is of growing
relevance can be included accordingly.
49
4. Exemplary assessments
Within this section, the previously described method is applied to two examples. In the first
example different repowering scenarios are compared to each other, while two different update
scenarios are assessed in the second example.
The underlying data represents generic converters of different rated powers, i.e. it is based on other
studies (see [3,8]) and does not consider any site specific factors that would need to be considered
in a case or site specific assessments of specific converter types. In some cases, site specific factors
may have a critical influence on the results. Therefore, the following examples basically
demonstrate how the described method can be used in order to compare different product scenarios
to each other, rather than providing decisive conclusions or recommendations on how the life cycle
of particular WEC should be designed.
Generally, the maximum life span of WEC is considered to be 20 years. This needs to be taken into
consideration when analyzing the results of the assessment.
CED [kWh]
2,800
3,100
3,100
3,600,000
3,053,000
3,100,000
1,880,000
Based on the method described in section 3 and the above data the comparison of the different
scenarios is carried out. The results of this comparison are shown in Fig. 1.
50
CED/Enet
[kWh/kWh]
1,00
0,10
0,01
1998
2008
WEC,old
2018
WEC, old
2028
Year
scenario 1
2038
scenario 2
2048
scenario 3
In the figure, additionally to the different repowering scenarios, a reference scenario representing no
repowering is shown by the dashed black graph. This reference scenario functions basically as an
orientation; no repowering is not a real option due to the limited life span of WEC but may still
function as a benchmark (however, updating/refurbishing could be an option, which is assessed in
the following section).
In the diagram, the x-axis marks the installation of the original WEC which is associated with a
high energy consumption (including production, transport and other factors influencing the CED)
while energy production starts right after the installation. The resulting graph showing the specific
energy demand falls with proceeding time as the produced net energy grows. The installation points
of the different repowering scenarios are also associated with energy consumption appearing as
peaks within the figure, falling with proceeding time, too.
It can be observed, that the repowering scenario 1 does not fall below the reference scenario within
the assessed time scope, while scenarios 2 and 3 show a better performance. Scenario 2 falls below
the reference line in 2023, which is 14 years after its installation. Scenario 3 shows an even better
performance, falling below the reference line already in 2022, i.e. eight years after its installation.
So, the graphical assessment (and the calculative assessment, too) shows scenario 3 as the favorable
option of the assessed product scenarios; i.e., here, repowering of the 0.6 MW WEC in 2014 with
the 3 MW WEC is the best option in terms of the specific energy demand.
For scenario 1 it has been assumed that rotor blades and tower cannot be reused and need to be
replaced. Rotor blades have a life span of maximum 20 years and usually after 12-15 years
refurbishing is not economically feasible anymore. Concrete towers are usually not suited for reuse,
either. So, it has been assumed that rotor blades and tower need to be re-produced, resulting in an
energy demand equal to the original primary production of these components. The same applies to
access roads, crane hard-standing and foundation. For the electrics it has been assumed that 90%
can be reused while 10% need replacement (cf. [9]). For the nacelle including the generator it has
been assumed that refurbishing and updating require an energy demand of 10% of the original CED
of these components. This can be considered as a rather conservative assumption based on [9] with
regard to the fact that around 80% of the production CED comes from material production, with
20% resulting from additional processing and transport (cf. [9]).
For scenario 2 with utilization of the same site it has been assumed that the existing infrastructure
(access roads, crane hard-standing, foundation) and the tower however requiring refurbishing
equal to 5% of the original CED can be used again; given that a life span of about 60 years can be
regarded as realistic for concrete towers (cf. [9]). For the other main components rotor blades,
generator, nacelle and electricity the same assumptions as for scenario 1 apply.
For both update scenarios 2,300 annual full hours have been assumed.
An overview of these parameters is shown in Table 3.
Table 3. Update scenarios
Component
Scenario 1
rotor blades
new
tower
new
generator
reuse, 10% of original CED
nacelle
reuse, 10% of original CED
electric components
reuse, 10% of original CED
infrastructure
new
full load hours
2,300
Scenario 2
new
reuse, 5% of original CED
reuse, 10% of original CED
reuse, 10% of original CED
reuse, 10% of original CED
reuse
2,300
The third scenario has been chosen as a benchmark for the update scenarios. It is identical with the
3 MW WEC that has been assessed in the previous section except for its earlier date of installation.
Within each product scenario, end-of- life credits are given 100% to the second product system. This
way, a consistent approach for each product system is assured and the comparability of the results is
increased. End-of- life credits are given according to the base end-of- life scenario described in [9]
using system expansion.
Using these assumptions and boundary conditions the graphical assessment of the update scenarios
has been carried out according to the described method, too, leading to the picture shown in Fig. 2.
52
CED/Enet
[kWh/kWh]
1,00
0,10
0,01
2005
2010
WEC, old
scenario 1
scenario 2
scenario 3
2015
2020
2025
2030
2035
2040
Year
In the diagram, the dashed black line shows the original WEC. It is continued by the alternative
scenarios for updating and repowering, respectively. Contrary to the assessment in section 4.1 this
assessment is not focused on (or limited to) one particular site but assessed different possible life
cycle or product options resulting from the original WEC reaching its end-of- life.
As it can be seen in the figure, scenario 2 not requiring infrastructure shows the best performance of
the assessed scenarios. The assessment of scenarios 1 and 3 leads to rather similar results; here
considering site specific factors could give a better understanding of the environmental burdens.
This shows that, if the utilization of an existing infrastructure is possible, updating appears to be a
preferable option with regard to optimizing the energy yield, even more attractive than repowering
assuming that a new infrastructure is required, here. For the use of updated WEC as second- hand
WEC at a different site, the difference to new WEC is not that significant.
Based on the exemplary assessment of update scenarios, it can be said that the use of updated
converters as second-hand WEC is definitely associated with environmental benefits but since
components like tower and rotor blades that usually do not have second-hand potential are
responsible for a significant share of the CED, the advantages compared to state-of-the-art WEC are
reduced- at least regarding the assessed indicator. If however re-using tower and infrastructure is
possible as well, updating shows a supreme performance also compared to state-of-the-art
repowering WEC. Still, if the favorable option needs to be identified in a specific situation a case
specific assessment needs to be performed. In general, however, it has to be said, that second- hand
WEC and new WEC serve different markets and do not compete against each other. Also, re-use
can be generally recommended; this also applies to wind energy converters. Giving WEC a second
life makes sense in terms of energy balance as well as other environmental impact categories and
economically (since selling and reuse is more profitable than disposal). Especially from a resource
point of view re-use of WEC can be generally recommended. Even though recycling of WEC is
relatively unproblematic with the rotor blades being an exception re-use still is the most
resource saving option.
While an assessment using the described method can provide valuable information and help to
better understand the environmental implications of different product scenarios, it requires a
significant amount of data. To assess the specific energy demand, for example, knowledge of
several figures is required: rated power, time of availability, number of annual full load hours and
cumulated energy demand. While the first three are usually known, the latter requires some effort to
be identified, especially when site specific aspects shall be included in the assessment. Here,
conducting a life cycle wide assessment of the energy balance for all identified product scenarios
might be necessary, or if other indicators shall be included a full life cycle assessment (LCA).
These assessments require some effort and might not always be feasible due to data availability.
The development and use of customized LCA tools (cf. [8]) can be of help, here, and significantly
reduce the effort related with such assessments (a onetime extensive data collection is still required,
though). Performing LCAs and additional case studies using the described methodology would
allow creating a data basis that could be used for future assessment. This data basis might also allow
drawing more specific conclusions on the environmental benefits and chances of repowering and
updating of WEC.
References
[1] European Wind Energy Association, Statistics and Targets. Factsheet, 2011. Available at
<http://www.ewea.org/fileadmin/swf/factsheet/1_statisticsandtargets.pdf>. [accessed
13.12.2011].
[2] World Wind Energy Association, Half year report 2011, Bonn, 2011.
[3] Zimmermann T., Gling-Reisemann S., Optimal repowering of wind energy converters:
Energy demand and CO2 intensity as indicators. Life Cycle Management Conference 2011,
2011 Aug 28-31; Berlin, Germany.
[4] European Wind Energy Association, Wind in power: 2010 European statistics, 2011.
Available
at
<http://ewea.org/fileadmin/ewea_documents/documents/statistics/EWEA_Annual_Statistics_2010.pdf> [accessed 13.12.2011].
[5] Bundesregierung, Erneuerbare Energien ausbauen, 2011.
Available at <http://www.bundesregierung.de/Content/DE/Artikel/2011/06/2011-06-06-eegnovelle.html> [accessed 13.12.2011].
[6] Bundesverband WindEnergie e.V., Zukunftsmarkt Repowering, 2010. Available at
<http://www.wind-energie.de/de/themen/repowering/> [accessed 22.10.2011].
54
[7] Windenergie Agentur, Repowering von Windenergieanlagen in der Metropolregion BremenOldenburg: Handlungsempfehlungen zur Steigerung der Akzeptanz von Repowering,
Bremerhaven, Germany, 2010.
[8] Zimmermann T., Fully parameterized LCA tool for wind energy converters. Life Cycle
Management Conference 2011, 2011 Aug 28-31; Berlin, Germany. Available at
<www.lcm2011.org/papers.html?file=tl_files/pdf/poster/day2/ZimmermannFully_parameterized_LCA_tool_for_wind_energy_converters-567_b.pdf>, [accessed
23.12.2011].
[9] Zimmermann T., Entwicklung eines Life Cycle Assessment Tools fr Windenergieanlagen.
Thesis, Bremen, Germany, 2011.
55
Division of Energy Processes, School of Chemical Science and Engineering, KTH Royal Institute of
Technology, Stockholm, Sweden, mimmim@kth.se (CA), alvfors@kth.se.
Abstract:
Biogas is a vehicle fuel of the first generation of biofuels with great potential for reducing the climate impact
from the transport sector. Today biogas is mainly produced by digestion in Sweden and the total amounts to
1.4 TWhLHV /year (2010) of which about 0.6 TWh LHV is upgraded and used in the transport sector. Using
industrial wastewater, e.g. from a pulp and paper mill, as substrate for production of biogas, the amount of
renewable fuel to the transport sector could be increased. In the pulping industry, substantial amounts of
organic matter are generated; this is commonly treated aerobically to reduce the chemical oxygen demand
(COD) in the effluent streams before discharge to a recipient. Treating these effluent streams mainly
anaerobically instead could contribute to the transport sectors energy supply. The aim of this study is to
investigate the potential for using effluent streams from the Swedish mechanical pulp and paper industry to
produce biogas. A typical Swedish mechanical pulp mill is considered for anaerobic treatment of the
wastewaters. This type of pulp mill presently uses conventional methods for wastewater treatment to reduce
COD, but converting most of this to anaerobic treatment would increase the amount of biogas produced.
When considering this conversion in a larger context, supposing that anaerobic treatment would be applied
to all Swedish mechanical pulp mills, which stand for about 30% of the total Swedish pulp production, it is
shown that the production could amount to as much as 0.5 TWh LHV /year of biogas. This represents about
one third of the biogas produced in Sweden today. The main conclusion of this study is that if anaerobic
treatment of effluent streams from the pulping industry were introduced, the biogas production in Sweden
could be significantly increased, thus moving one step further in reducing the transport sectors climate
impact.
Keywords:
Anaerobic digestion, Biofuel, Biogas, Mechanical Pulping, Vehicle Fuel, Wastewater Treatment.
1. Introduction
The transport sector is seen as one of the most problematic areas when it comes to reduction of
climate impact, due to its heavy dependence on fossil fuels. Biogas is a vehicle fuel of the first
generation of biofuels, meaning that it is presently available on a commercial scale and has a welldeveloped infrastructure for distribution and filling, in some areas via a gas grid. In total, the biogas
production in Sweden amounts to 1.4 TWhLHV/year (2010) of which about 0.6 TWhLHV/year is
upgraded and used in the transport sector [1]. The production of biogas is commonly performed by
anaerobic digestion, a technology that has been seen as one of the oldest to stabilise waste and
sludge [2]. The production today is mainly based on sludge digestion and takes place in sewage
treatment plants, but co-digestion plants, for instance with food wastes or agricultural wastes as
substrates also exist, and there is further anaerobic treatment as a method for wastewater treatment
in other types of industries.
The pulp and paper industry has been one of the most important base industries in Sweden for a
long time, and still holds a strong position domestically as well as internationally. The annual
production of pulp in Sweden is 12 million tonnes of the world total which is 178 million tonnes
[3]. Pulp and paper production requires large amounts of water for the process and the effluent
waters are rich in dissolved organic matter, measured as chemical oxygen demand (COD). Great
56
progress has been made to reduce the water use at the mills by closure of the water circuits and also
to improve the wastewater treatment of the effluents. Since the late 1970s the environmental impact
from pulp and paper industries has decreased substantially, for instance the emissions of COD in
Sweden has decreased from 1.6 to 0.2 million tonnes of COD per year, while the pulp production
has increased by about 4 million tonnes per year [3]. In Sweden the most common way to treat
wastewaters from the pulp and paper industry has been by aerobic methods, which are rather
electricity demanding: anaerobic wastewater treatment has so far not reached a break-through.
1.1 Aim
The aim of this study is to investigate the potential for using effluent streams from the Swedish pulp
and paper industry to produce biogas. A case study is made on a typical Swedish mechanical pulp
mill, the Rottneros mill, where a combination of anaerobic and aerobic treatment is considered an
alternative to conventional aerobic methods for wastewater treatment. In the case study, design
calculations are performed based on present effluent data from the mill. Based on the results from
the case study on the Rottneros mill a potential biogas production for all Swedish mechanical pulp
mills (producing groundwood, thermomechanical and/or chemi-thermomechanical pulp) is
estimated as a possible way of increasing the share of renewable fuels in the transport sector.
2. Background
Biogas is a vehicle fuel that is well suited for both light and heavy vehicles. It is commercially
available and can be distributed either in a gas grid, or to filling stations by trucks. The most
common way of production is by anaerobic digestion, one of the oldest processes for treating
organic wastes [4, 5]. Although used to a great extent in sewage treatments plants (for sludge
digestion), the anaerobic methods for wastewater treatment have not reached a break-through in the
pulp and paper industry in Sweden, where other wastewater treatment methods are more common.
In the following sections, a short introduction to anaerobic digestion and to the pulp and paper
industry is given.
(CCS) or, as described in a study by M ohseni et al [11], be used to produce additional methane by
reacting with hydrogen.
There are several parameters influencing the digestion and the rate of the different stages; among
these are, e.g. temperature, number of fermentation steps, pH, retention time, alkalinity and
moisture content of the substrate. These parameters are specific for each process and may be
adjusted to enhance the biogas production. The temperature for digestion, however, is often seen as
one of the most sensitive process parameters. There are three temperature intervals where the
digestion is performed optimally: psychrophilic/cryophilic (020 C), mesophilic (3042 C) and
thermophilic (4560 C): industrial digestion is most commonly performed with the two higher
intervals [6, 7]. Generally, a higher temperature gives a faster reaction rate, which is an advantage
since smaller reactor volumes are needed. However, the thermophilic digestion is more sensitive
when it comes to temperature and concentration variations and needs more process control, whereas
the mesophilic digestion is a more stable process [6]. Whether mesophilic or thermophilic, the
process requires heating to keep the reactor at the appropriate temperature.
The upgrading of the raw biogas, needed if the gas is to be used as vehicle fuel, is usually
performed adjacent to the digestion, mainly only at large facilities since upgrading is rather costly.
Common techniques for upgrading include water scrubbing, pressure swing adsorption (PSA) and
chemical absorption. After upgrading the biogas may be compressed or condensed to liquid biogas
(LBG) before being distributed to a gas grid, if available, or by trucks to filling stations [8].
of treatment method is chemical precipitation [12]. M ost WWT facilities at pulp and paper mills
today are external treatment plants handling effluent streams before discharge to the recipient.
Aerobically based WWT methods are generally rather energy demanding and also produce large
amounts of sludge. The solid wastes from the process include wood and bark residues, fibre rejects
(primary sludge) and biological sludge from the WWT [12]. Solids that are dry enough are usually
incinerated in the bark boiler or at external combustion plants, but the main fraction of solid wastes
primary sludge and biological sludge has to be dewatered, and possibly also dried, before
further treatment, for instance incineration, which is of course energy demanding [12, 13].
including aeration basin and secondary clarifier. Nutrients (phosphorus and nitrogen) are added at
the activated sludge stage. Part of the sludge from the secondary clarifier is returned to the aeration
basin. There is no withdrawal of excess sludge from the secondary treatment stage. Water and the
remaining sludge are led to tertiary treatment comprising chemical addition (of polyaluminum
chloride) and flotation. The pH of the effluent water from the flotation is measured and adjusted
before discharge to the recipient, lake Fryken. The sludge from the flotation is dewatered and the
reject water is returned to the secondary clarifier, whereas the biologic and chemical sludge,
together with the primary sludge, is taken care of by an external company. The total degree of COD
reduction in the present wastewater treatment is 83% [16].
Fig. 1. Simplified process diagram over the present WWT for Rottneros.
In the suggested WWT for the Rotteros mill, anaerobic treatment in an upflow anaerobic sludge
blanket (UASB) reactor is combined with aerobic treatment, see Fig. 2 for a simplified process
diagram. Before entering the anaerobic reactor the flow is treated by chemical precipitation with
FeSO4 in the primary clarifier to reduce the amount of organic matter that is difficult to degrade and
to make the wastewater less toxic to the micro-organisms present in the anaerobic process. The
COD reduction in this primary precipitation stage is approximately 15%. The water from the
primary clarifier is pH-adjusted with acid, and then taken to the UASB reactor. The degree of COD
reduction in this stage is 70%. The produced biogas, with methane content of 70%, is separated and
compressed. After the anaerobic treatment some further biological degradation is performed in the
present aerobic treatment at Rottneros. An aerobic polishing is needed to avoid odours from the
effluent. The flotation stage is used for final removal of suspended solids and if needed some
chemicals may also be used for further reduction of COD [16]. The total degree of COD reduction
for the entire WWT (including the primary clarifier, aeration and secondary clarifier) is 84%. The
data for the present and for the suggested anaerobic treatment for Rottneros are presented in section
4.1.
Fig. 2. Simplified process diagram over the suggested WWT for Rottneros.
60
400 000
2 400 000
740 000
3 600 000
26 000
In this study two scenarios concerning the potential for biogas production are presented: one
moderate and one optimistic. Three parameters are varied in the assumptions for the calculations for
these two scenarios; these parameters and other relevant data are summarised in Table 2, below.
The first one concerns the degree of COD reduction at the current wastewater treatments at the
mills, which is needed to be able to calculate the amount of COD from the process: the digestible
potential. As mentioned, the WWT methods are different at the Swedish mills and thus have
different degrees of COD reduction. According to [12], typical values for COD removal at
biological treatment plants are in the range of 80% to 90% of the COD. In this study, two levels of
reduction are assumed, the same for all mills: 85% in the moderate scenario and 90% in the
optimistic scenario. Secondly it is assumed that not all the effluent streams from the process are
suitable for anaerobic digestion, therefore it is assumed that 90% of the COD is possible to treat
anaerobically in the moderate scenario and in the optimistic scenario the corresponding figure is
95% [19]. The third parameter varied is the anaerobic degradability of COD in the digestion, which
depends on the material treated and the process conditions, such as the loading rate, for the
anaerobic system. According to [4] the anaerobic degradability is in the range of 60-87% for TMP
and somewhat lower for CTM P. In the case study, the anaerobic degradability was 70%, which is
reasonable for a mill with GWP and CTMP. The mechanical pulps produced in Sweden consist of a
large share of TMP though, and therefore the anaerobic degradability could be assumed to be
somewhat higher in average. In this study the degradability is assumed at two different levels, 75%
and 85% for the moderate and optimistic scenarios, respectively. The methane content in the
produced biogas is assumed to be 70% in both scenarios (common contents are in the range of 6575% according to [12]).
61
Table 2. Parameters used in calculations of the potential for biogas production from mechanical
pulp mills.
M oderate
Optimistic
scenario
scenario
a
Total pulp production, t/year
3 600 000
3 600 000a
COD emissions in effluent, t/year
26 000a
26 000a
Total degree of COD reduction in present WWT, %
85
90
Amount COD possible to treat anaerobically, %
90
95
Anaerobic degradability in digester, % of COD
75
85
Specific gas production, Nm3/kg COD removed
0.35b
0.35b
M ethane content of produced gas, %
70
70
3
c
Energy content of produced biogas, kWhLHV/Nm biogas
6.98
6.98c
a
From [18].
In [15, 20 and 21] speci fic gas productions in the range of 0.18 to 0.54 Nm3 /kg COD removed are reported.
c
The energy content of biogas varies with the methane content. The energy content for pure methane is 9.97
kWhLHV/Nm3 [8].
b
4. Results
In this section, the results from the design calculations are given, both for the case study on the
Rottneros mill and for the potential in the moderate and optimistic scenarios, respectively, if
implemented in all Swedish mechanical pulp mills.
The Rottneros mill today has a sufficient and adequate WWT that is well suited for its production.
If it, however, aims to expand the production at the mill, an anaerobic digestion treatment combined
with the existing one is a promising alternative.
62
5. Discussion
This study has first investigated the change to anaerobic wastewater treatment in a case study mill
by design calculations, partly as a way of decreasing the energy demand in the WWT system, partly
as a way of producing biogas. Secondly, the potential for biogas production from all Swedish
mechanical pulp mills was estimated, assuming that anaerobic treatment was introduced at these
sites. This estimation was made for two different scenarios, a moderate and an optimistic. The
difference between these scenarios can be found in the assumptions regarding the degree of COD
reduction in the present WWT, amount of COD possible to treat anaerobically and the degree of
COD reduction in the anaerobic reactor. The figures for the moderate scenario reflect the fact that
there might be disturbances in the production during the year, that full load is not always possible
and is overall a more careful estimation. The optimistic scenario assumes that production is more
stable and without disturbances; that a larger share of the streams is possible (or made possible) to
treat anaerobically and that the process improvements that may increase the biogas production are
implemented. These factors would give the possibility for the higher biogas production shown in
the optimistic scenario, however it should not be seen as easily achievable.
The focus of this study is biogas used as vehicle fuel in the transport sector. To be used for this
purpose the raw biogas has to be upgraded to vehicle standard. An upgrading facility might not be
an alternative for a smaller mill because of the rather large investment this would require. If not
being upgraded on-site, a possibility would be to transport the biogas to an external upgrading
63
facility, even though the raw biogas requires large volumes. The biogas could of course also be used
for other purposes, for instance internally in the pulp mill, replacing fossil and/or biomass-based
fuels needed for the production. In [16], the biogas was assumed to replace either oil or biomass
internally in the mill and, as expected, it was shown that it was more economic to replace the oil.
Biogas could also be used for heating, power or combined heat and power, especially use for
heating purposes is common for biogas produced in Sweden today. In 2010, about the same amount
of biogas was used for heating as for upgrading in the Swedish energy system (0.6 TWhLHV) [1].
The size of the anaerobic facility, producing 27 GWhLHV/year, is comparable to other digestion
facilities. The average production for industrial digestion plants and co-digestion plants in Sweden
is 19-23 GWhLHV/year. The average size for a wastewater treatment plant in Sweden is lower, about
5 GWhLHV/year, but the number of plants as well as range of their production is large, for instance
three of the wastewater treatment plants in the Stockholm area produce around 50 GWhLHV/year
respectively [1,22, 23]. It would thus be reasonable to believe that investing in anaerobic
wastewater treatment could be economic for the mill. Considering the production, as well as COD
emissions in effluent, of the Rottneros mill in comparison to other Swedish mechanical pulp mills,
it may be seen that the mill is within the range of an average sized mill [18] and thus possible
anaerobic wastewater treatment facilities at several of the mechanical pulp mills in Sweden should
also be in the same size.
The process for anaerobic treatment assumed to be used in this study is, as in most industrial
digestion facilities in Sweden, of the mesophilic type. However, as has been discussed, the higher
the temperature used, the faster the reaction rate and therefore it is possible to digest at thermophilic
conditions, which even may be preferable at TM P or CTM P mills because of the high temperature
(close to the thermophilic working interval) of the effluent streams from the process [14]. This was
also investigated in [15], where digestion in the thermophilic interval was assumed to be more
energetically favourable, compared to mesophilic digestion, since most process streams from the
pulp mill have rather high temperatures and cooling of the process streams would thus be avoided.
The WWT processes investigated in [15] have similarities to the one used in this study, however
these processes also use different set-ups of membrane filtration in combination with the anaerobic
digestion, as mentioned. Another difference between the processes in [15] and the one here is that
the sludge from the aerobic treatment following the anaerobic reactor is returned to this reactor to
be treated together with the rest of the effluent material. This is an efficient way of reducing the
sludge from the WWT and thus also the need for further sludge treatment. Depending on the
amount of sludge a separate sludge digestion stage might be needed.
The Swedish mechanical pulp mills all have their mill specific conditions and thus the biogas
potential estimated in this study should be considered carefully. To implement a change in the
wastewater treatment such as suggested in this study a thorough evaluation of the mill, and the
specific process used there, would be necessary. Depending on the process type, i.e. whether
bleaching is applied and, if so, which type of bleaching, only few of the effluent streams might be
suitable for anaerobic treatment. However, for a mill that currently has a satisfactory WWT but is
looking to increase their production, investing in an anaerobic treatment could be a promising
option for coping with the increased effluents.
This study only investigated pulp mills producing mechanical pulps. It is also possible to treat some
effluent streams from chemical pulp mills, thus investigating the potential for these mills is also
relevant. The production of chemical pulp in Sweden is about three times as large as the production
of mechanical pulps. Moreover, these types of mills give higher amounts of COD per tonne
produced pulp. However, the situation at the chemical mills is more complex than at mechanical
mills, for instance, the streams from the chemical mills are even more varying and may also contain
other types of chemicals that might disturb the anaerobic process. A potential assessment for biogas
production from the whole pulp and paper industry in Sweden will be made in a coming study.
64
6. Conclusions
Using effluent streams from the mechanical pulping industry in anaerobic digestion to produce
biogas could increase the share of biofuels to the transport sector. The main conclusions of this
study are summarised below:
Using anaerobic wastewater treatment combined with aerobic polishing stage as an alternative
to conventional aerobic wastewater treatment methods at mechanical pulp mills gives a
satisfactory wastewater treatment as well as a substantial contribution of biogas.
When using anaerobic digestion both electricity consumption and sludge production at the mill
are decreased, giving additional environmental advantages.
The production size of an anaerobic wastewater treatment facility at a typical Swedish
mechanical pulp mill is about the same as other industrial digestion plants and anaerobic
wastewater treatment plants and such an investment should therefore have the potential of being
be economic for a mill.
If implemented in the whole Swedish mechanical pulping industry the biogas produced could
amount to 0.3 TWhLHV (moderate scenario) or as much as 0.5 TWhLHV (optimistic scenario)
annually.
With a biogas production from the mechanical pulping industry according to the optimistic
scenario (0.5 TWhLHV/year), the total Swedish biogas production would increase by one third
compared to todays production.
Acknowledgments
This work has been carried out under the auspices of The Energy Systems Programme, which is
primarily financed by the Swedish Energy Agency. The authors would like to thank M ats
Westermark (prof. at the division of Energy Processes) for valuable comments, sa Sivard, Tomas
Ericsson (F) and Nippe Hylander (F and Adj. prof. at the division of Energy Processes) for
contributing with the basis for the potential estimation and for valuable comments throughout the
work and Yi Liu, author of the master thesis on which the case study in this paper is based.
Nomenclature
CCS Carbon capture and storage
CH4 M ethane
CO2 Carbon dioxide
COD Chemical oxygen demand
CTM P Chemi-thermomechanical pulp
DS
Dried substance
GWh Gigawatt hour
GWP Groundwood pulp
kWh Kilowatt hour
LBG Liquid biogas
LHV Lower heating value
Nm3 Normal cubic meter
PGWP Pressurises groundwood pulp
PSA Pressure swing adsorption
t
Tonne
TM P Thermomechanical pulp
TWh Terawatt hour
65
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
Swedish Energy Agency, Production and use of biogas 2010 (Produktion och anvndning av
biogas 2010) [in Swedish]. Eskilstuna, Sweden: Swedish Energy Agency; 2011 Nov. Report
No: ES 2011:07.
van Lier, J.B., Tilche, A., Ahring, B.K., M acarie, R., M oletta, M ., Dohanyos, L.W., Hulshoff
Pol, L.W., Lens, P., Verstraete, W., New perspectives in anaerobic digestion. Water Science
and Technology 2001;43(1):1-18.
Swedish Forestry Industry Federation. The Swedish Forest Industries 2010. Facts and figures.
Available at <http://www.skogsindustrierna.se/web/Utgivningsar_2011.aspx> [accessed
2011-10-03]
Rintala, J.A., Puhakka, J.A., Anaerobic treatment in pulp and paper-mill waste management: a
review. Bioresource Technology 1993;47:1-18.
Yadvika, Santosh, S., Sreekrishnan, T.R., Kohli, S., Rana, V., Enhancement of biogas
production from solid substrates using different techniques a review. Bioresource
Technology 2004;95:1-10.
Appels, L., Baeyens, J., Degrve, J., Dewil, R., Principles and potential of the anaerobic
digestion of waste-activated sludge. Progress in Energy and Combustion Science
2008;34:755-781.
Weiland, P., Biogas production: current state and perspectives. Applied M icrobiology and
Biotechnology 2010;85:849-860.
Swedish Gas Centre. Biogas folder 2011 Basic data on biogas (Biogas Basdata om biogas
2011) [in Swedish]. Available at
<http://www.sgc.se/uk/display.asp?ID=1352&Typ=Rapport&M enu=Rapporter> [accessed
2011-09-29].
Svensson, M ., Evaluation of Swedish biogas standard Basis for a future revision
(Utvrdering av svensk biogassatandard Underlag fr en framtida revision) [in Swedish].
M alm, Sweden: Swedish Gas Centre; 2011 April. Report No: SGC 229.
Danish Gas Technology Center. Adding gas from biomass to the gas grid. Final report of
Altener, Contract No: XVII/4.1030/Z/99-412. Available at <http://www.dgc.eu/publi2.htm>
[accessed 2012-04-02].
M ohseni, F., M agnusson, M ., Grling, M ., Alvfors, P., Biogas from renewable electricity
Increasing a climate neutral fuel supply. Applied Energy 2012;90:11-16.
EUROPA The official website of the European Union. European Commission, Joint
Research Center, Institute for Prospective Technological studies, Integrated Pollution
Prevention and Control. Draft reference document on best available techniques in the pulp
and paper industry, Draft April 2010. Available at: <http://eippcb.jrc.es/reference/> [accessed
2011-12-20]
Stoica, A., Sandberg, M ., Holby, O., Energy use and recovery strategies within wastewater
treatment and sludge handling at pulp and paper mills. Bioresource Technology 2009;100:
3497-3505.
Salkinoja-Salonen, M ., Apajalahti, J:, Silakoski, L:, Hakulinen, R., Anaerobic fluidised bed
for the purification of effluents from chemical and mechanical pulping. Biotechnology
Advances 1984;2:357-375.
Sivard, ., Ericsson, T., Process integration of wastewater treatment technique decrease of
the energy and resource use in the forestry industry (Processintegration av vattenreningsteknik
66
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
67
1. Introduction
Today, the global transport sector is heavily dependent on fossil fuels and fossil-originated fuels
accounts for 98 % worldwide [1] and 92 % in Sweden [2]. The demand for renewable energy
carriers is increasing in Stockholm as well as the rest of the world. A transition towards a
renewable-fuelled road transport system is desirable for a number of imperative factors, such as the
need to reduce anthropogenic green house gas emissions and security of supply. One measure to
increase the share of renewable fuels in a region, is to benefit from its generated organic waste,
which may be digested, form biogas and when purified, be used as fuel in automotive applications.
68
The European Commission distinguishes the organic compound in municipal solid waste as food
waste [3].
To investigate the biogas potential derived from food waste in the region, the Stockholm County
Administration Board carried out a regional inventory of available food waste in the Stockholm
County and the results was published in February 2011. The final estimations from this report
constitute the foundations for the calculations carried out in this paper. The analysis in this paper is
expanded to also consider utilisation of produced biogas as fuel for renewable electricity generation
aimed for automotive applications and tank-to-wheel (TTW) analysis including the actual vehicles,
in order to assess the overall energy efficiency of the two vehicle fuels.
69
Figure 1. Estimated amount of food waste available in the Stockholm County and the
corresponding share with biogas potential accounted as thousand tonnes.
The report [4] does not include any considerations regarding the biogas production process. All
process steps are therefore assumed to been accounted for as well as the produced biogas is
assumed to maintain Swedish automotive quality standard, SS 15 54 38.
The report [4] expresses its main result as litres of petrol possible to substitute with the food waste
originated biogas. The fossil fuel substitution potential concluded in the report [4] is complied in
Table 1. To reach these conclusions, assumptions such as energy content and conversion efficiency
has been made but is not included in the report. Relevant thermodynamic assumptions fundamental
for this paper will be presented in section 1.1.2.
Table 1. Fossil fuel substitution potential corresponding the available food waste
2009
2030
Food waste [tonnes]
122 000
152 000
Petrol substituted [litre]
15 000 000
19 000 000
1.1.2 Generated data
According to the EU Directive 2009/30/EC, also referred to as the EU Fuel Quality Directive, the
petrol sold in Sweden constitutes of 5 % ethanol [6]. The E5 petrol has the lower heating value
(LHV) of 8,94 kWh/l [7]. The production capacity of biogas is calculated from the reported amount
of petrol substituted. The produced amount of biogas corresponding to the substituted petrol, see
Table 1, is found in Table 2:
Table 2. Potential biogas production (LHV) originated from food waste in Stockholm County
2009
2030
Biogas [GWh]
134,1
169,9
To illustrate the production of biogas from food waste, Figure 2 is a simplified schematic
illustration of the conversion system.
70
Figure 3. System description of two possible vehicle fuel options utilising biogas as a renewable
resource
71
are different parameters that determine each routes energy efficiency, which influence the absolute
driving distance of the vehicle.
To ensure an accurate comparison, a model vehicle is designed using the shared classification
system of the European Commission and Euro NCAP [8] and the vehicle size is assumed to
correspond to a C-segment medium car/ small family car, for example Ford Focus or Opel Astra.
To obtain an additional measure of the total energy efficiency for each transportation fuel, an
annual driving distance for such model vehicle is determined. This paper will assume an annual
driving distance of 15 000 kilometres, which is the statistic distances a Swedish C-segment vehicle
owner drive every year [9]. The total number of each vehicle, respectively, function as an additional
comparison of the overall energy efficiency of the systems.
72
Table 5. Potential renewable electricity originated from food waste in Stockholm County
2009
2030
Electricity [GWh]
53,4
67,9
Within the category electric vehicles (EVs), the degree of powertrain electrification may vary from
a conventional mechanical powertrain equipped with a powerful machine working as start engine,
to a fully electrified powertrain. To utilise the generated electricity and charge from the grid, this
paper distinguish between two different types of EVs:
Battery Electric Vehicles (BEV) that have an electric engine and a battery as an energy
storage unit. No internal combustion engine (ICE).
Plug- in Hybrid Electric Vehicles (PHEV) that have both an ICE and an electric engine.
The ICE generates electricity, via a generator mounted upon the ICE, to the electric
powertrain. The energy storage in this case is also a battery.
This paper will only consider BEV since its sole fuel source is electricity, since the degree of
electrical mileage may vary for a PHEV.
The electrical powertrain, compared to a conventional powertrain, is characterised by a high overall
energy efficiency, about 80 % and includes charge/discharge cycle losses of the battery. The
improved energy efficiency is predominantly the result from replacing the ICE by an electric engine
(90 % energy efficiency compare to 30 % as for an ICE) and improvements due to electrical power
transmission instead of mechanical [18]. A C-segment electric vehicle utilise approximately 20
kWh/100 km [19]. Table 6 show the results of the total number of kilometres fuelled by generated
electricity and its corresponding numbers of electric vehicles with an annual driving distance of
15 000 kilometres.
Table 6. Electricity total propulsion and corresponding numbers of vehicles
2009
2030
Promoted propulsion [100 km]
2 682 000
4 529 600
Numbers of vehicles
16 111
27 209
73
Figure 4. Sankey diagram of energy flows if utilising available biogas as road transport fuel
The Sankey diagram illustrating the option of utilising renewable electricity as road transport fuel,
visualise the additional process steps associated with converting the energy carrier to electricity.
Data given in Table 4 is applied to each step, respectively, to express the coupled energy loss.
Figure 5. Sankey diagram of the energy flows if utilising the available biogas for power generation
in order to obtain renewable electricity for an electrified vehicle fleet.
This paper have made the comparison of the corresponding number of vehicles each fuel option
promote, found in Table 3 and Table 6.In 2009, the number of passenger vehicles in the County of
Stockholm was 800 534 [20]. Assessing the potential impact in fleet composition, the share of
electric vehicles would run up to 2,2 % compared to the corresponding share of biogas vehicles that
may be fuelled by the same resource that would be 2,0 %. Forecasting the vehicle fleet in the
Stockholm County by 2030, it may be assumed that the potential impact might increase hence to
intended governmental efforts to reduce the transport demand in the region [21], [22].
4. Discussion
The measure of energy efficiency used in this paper is total accumulated kilometres and
corresponding number of vehicles caused by each fuel option derived from the equal amount of
food waste originated biogas. In the case of biogas, produced amount of biogas is shown in Table 3.
This amount of biogas is available for the end-consumer. In the case of electricity, the same amount
74
5. Conclusion
Many factors may influence the utilisation of the limited renewable resources available, in this case
food waste. The two options considered in this paper was to digest the food waste to biogas, then
utilising the biogas either as a road transportation fuel or as fuel for renewable electricity
generation. Including the vehicle and its powertrain in the analysis showed that the most energy
efficient way of utilising the renewable resource was to fuel an electrified vehicle fleet. Converting
the biogas from the food waste to electricity adds approximately another 10 % of driving distance.
Assuming an annual driving distance of 15 000 kilometres, in 2030 this would imply either 27 450
biogas or 27 200 electric passenger cars in the county of Stockholm. Regarding food waste as a
renewable resource, of course most desirable would be to minimise this source and instead enjoy
the food before it becomes waste.
6. Acknowledgements
The work has been carried out under the auspices of The Energy Systems Programme primarily
financed by The Swedish Energy Agency.
In 2011, the Swed ish energy and carbon dio xide tax was 0,85 eurocents/kWh and 11,70 eurocents/kg CO2 ,
respectively [27].
75
7. References
[1] Kahn Ribeiro, S., S. Kobayashi, M. Beuthe, J. Gasca, D. Greene, D. S. Lee, Y. Muromachi, P. J.
Newton, S. Plotkin, D. Sperling, R. Wit, P. J. Zhou, 2007: Transport and its infrastructure. In
Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment
Report of the Intergovernmental Panel on Climate Change. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA
[2] Swedish Energy Agency. Energy Outlook 2010 (Energilget 2010). Report no. ER 2010:45.
[3] European Parliament. Directive 2006/12/EC of the European Parliament and of the Council of 5
April 2006 on waste
[4] Stockholm County Administrative Board. Present and future 2011:2 Food waste provides new
energy (Nutid och framtid 2011:2.Matrester ger ny energi).
Available at:
http://www.lansstyrelsen.se/stockholm/Sv/publikationer/2011/Pages/nof- matrester- ger-nyenergi.aspx accessed: 2011-09-15
[5] Swedish Waste Management Association. Report 2008:02. The Swedish biogas potential from
domestic raw materials (Den svenska biogaspotentialen frn inhemska rvaror). ISSN 1103-4092.
[6] European Parliament. Directive 2009/30/EC of the European Parliament and of the Council of
23 April 2009, amending Directive 98/70/EC as regards the specification of petrol, diesel and gasoil and introducing a mechanism to monitor and reduce greenhouse gas emissions and amending
Council Directive 1999/32/EC as regards the specification of fuel used by inland waterway vessels
and repealing Directive 93/12/EEC.
[7] SPBI. Swedish Petroleum and Biofuel Institute. Fact sheet available at:
http://spbi.se/faktadatabas/artiklar/berakningsmodeller accessed: 2011-09-28
[8] Euro NCAP. Technical report Development of the European new car assessment programme
(EURO NCAP). Available at: http://www.euroncap.com/files/Development-of-Euro-NCAP-1998--0-7498ffa2-36b6-4328-8b6e-f44fe5d35563.pdf accessed: 2011-12-27
[9] Sweden Statistics. Transports and Communications. Driving distance and fuel consumption
(Transporter och Kommunikationer. Krstrckor och brnslefrbrukning). 2009
[10] H Engerer, M Horn. Natural gas vehicles: An option for Europe. Energy Policy 38 (2010)
1017-1029
[11] Joint Research Centre of the European Commission (JRC). Tank-to-Wheel analysis of future
automotive fuels and powertrains in a European context. Version 2c. JRC, EUCAR, CONCAWE.
2006.
available
at:
http://iet.jrc.ec.europa.eu/sites/aboutjec/files/documents/TTW_Report_010307.pdf accessed: 2011-12-30
[12] Duval M. Advanced Batteries for Electric Drive Vehicles A technology and cost-effective
assessment for battery electric vehicles, power assist hybrid electric vehicles and plug- in hybrid
electric vehicles. EPRI. May 2004
[13] S E Plotkin. Examining fuel economy and carbon standards for light vehicles. Energy Policy.
37 (2009) 3843-3853
[14] A Poullikkas. An overview of current and future sustainable gas turbine technologies.
Renewable and Sustainable Energy Reviews 9 (2005) 409-443
[15] M Edstrm, Nordberg. JTI Swedish Institute of Agricultural and Environmental
Engineering. Report no. 107. Producing biogas on the farm (Producera biogas p garden). 2004.
[16] Swedish Energy Agency. Energy in Sweden facts and figures 2010. Report no. ET 2010.46.
2010
[17] Kgeson P. Nature Associates. The climate impact of electrified road transport. Klimateffekten
av elektrifierad vgtrafik. 2010-06-29.
76
[18] Helms H, Pehnt M, Lambrecht U, Liebich A. Electric vehicle and plug- in hybrid energy
efficiency and life cycle emissions. 18th International Symposium Transport and Air Pollution.
Dbendorf. Switzerland. 18-19 May 2010
[19] Duval, M. Advanced Batteries for Electric Drive Vehicles A technology and cost-effective
assessment for battery electric vehicles, power assist hybrid electric vehicles, and plug- in hybrid
electric vehicles. EPRI. May 2004
[20] Sweden Statistics. Vehicle statistics. Vehicles in counties and municipalities 2010/2011.
(Fordonsstatistik, Fordon i ln och kommuner vid rsskiftet 2010/2011). 2011
[21] Office of Regional Planning, Stockholm County Council. Report no. 5:2010. Regional
development plan for the Stockholm region (RUFS 2010 Regional Utvecklingsplan Fr
Stockholmsregionen). 2010
[22] Office of Regional Planning, Stockholm County Council. Energy future of the Stockholm
region 2010-2050: The way to reduce climate impact. 2010. Available at:
http://www.tmr.sll.se/Global/Dokument/publ/2010/2010_r_energy_future_of_the_stockholm_regio
n_2010-2050.pdf accessed: 2011-12-30
[23] Regulation (2010:178) on the taxation of energy (Frordning (2010:178) om skatt p energi).
Stockholm. 2010. (SFS 2010:178)
[24] Regulation (2003:564) on public grants to promote efficient and environmental- friendly energy
supply (Frordning (2003:564) om bidrag till tgrder fr en effektiv och miljanpassad
energifrsrjning). Stockholm. 2003. (SFS 2003:564)
[25] Regulation (2011:1480) on the electricity certificate system (Frordning 2011:1480 om
elcertifikat). Stockholm. 2011. (SFS 2011:1480)
[26] M. Gunnarsson, T Lygnegrd. Swedish Biogas International. stergtland biogas production
(stergtland
biogasring).
2008.
Available
at:
http://www.energiost.se/content/images/energiost/pages/biogasring_ostergotland_081201.pdf
accessed: 2011-12-30
[27] Swedenergy. Available at: http://www.svenskenergi.se/sv/Om-el/Miljo-och-klimat/Mal-ochstyrmedel/ Accessed: 2011-12-29
77
Abstract:
In the conversion of low temperature heat sources into electricity using an Organic Rankine Cycle system the
working fluid selection is a major design choice to maximize the overall performance. The placement of the
power cycle on a T-s diagram depends on the fluid critical temperature. Several studies have shown that the
power output can be maximized by using fluids with critical temperatures similar or lower than the inlet
temperature of the heat source, which allow a better temperature profile match between the heat source and
the working fluid. However, the choice of a fluid having a specific critical temperature also influences the fluid
properties in the expansion process over a given temperature interval, as shown by the generalized
compressibility chart. The aim of this paper is providing a better insight into the results of optimized ORCs
through the analysis of the compressibility factor in the expansion process.
To this purpose the real enthalpy change in the expansion process i s regarded as two separate terms
associated with temperature and pressure drops, re spectively. Starting from the analysis of different
expansion processes in optimized cycles a correlation is obtained between the compressibility factor and the
ratio between real enthalpy change and the enthalpy change term associated with temperature drop. Thus,
the ratio between the former and the latter can be directly evaluated from the simple knowledge of pressure
and temperature values along the expansion process and the observation of the compressibility chart.
Keywords:
Organic Rankine Cycle, Expansion Process, Real Gas, Compressibility Factor.
1. Introduction
In Organic Rankine Cycle systems low/medium temperature heat sources are used to evaporate a
secondary working fluid with a low boiling point that expands through a turbine to generate power.
The fluid is then discharged to the condenser where the condensing heat is transferred to a cooling
water or cooling air flow. The liquid condensate is pumped into the evaporator completing the
cycle. These systems, also known as binary cycle power plants [1], have found many applications in
the recovery of waste heat from industrial plants and are playing a very important role in the
modern geothermal electricity market [2].
The thermodynamic properties of the working fluid strongly influence the performance of these
systems. In particular, the critical temperature plays a primary role in fixing the placement of the
thermodynamic cycle on a T-s diagram [3]. The results of power-based performance comparisons
show that the best fluids have critical temperatures similar or lower than the inlet temperature of the
heat source ([4], [5]) so that the evaporation occurs at high subcritical reduced pressures (pR=p/pCR)
or even at supercritical pressures that imply a good match between thermal profiles of the heat
source and working fluid. At these conditions the ideal gas equation of state does not provide an
acceptable representation of the p-v-T relationship, since the compressibility factor (Z= pv/RT)
could assume values much lower than 1. An accurate knowledge of the p-v-T behavior is important
in the evaluation of the specific enthalpy change that can be expressed in terms of p,v,T and cP.
The favourable role of gas compressibility effects in reducing the work absorbed by the
compression process in vapor refrigeration and heat pump cycles was studied in [6]. The same real
gas effect is used to improve the conversion efficiency, through the minimization of the
78
cP T
(1)
dh c P T dT
On integration:
h T2
T2
h T1
T1
cP T dT
(3)
cP
T1
c P T dT
T2
T1
(5)
However, when the variation of cP over a given interval is slight, little error in normally introduced
by taking the specific heat required by (4) as the arithmetic average of the specific heat values at the
two end temperatures. Alternatively the specific heat at the average temperature over the interval
can be used.
79
dh
h
p
dT
p
(7)
dp
T
It can be shown [11] that (7) can be expressed in the following form:
dh cP dT
v
T
(8)
dp
p
where the right side of (8) depends solely on p,v,T and cP.
Changes in specific enthalpy between two states are found by integrating (8):
h2
h1
cP dT
v T
v
T
(9)
dp
p
Integration of the second term on the right of (9) requires knowledge of the p-v-T behavior at the
states of interest.
(10)
The compressibility factor Z tends to unity as pressure tends to zero at fixed temperature. From (10)
the specific volume can be expressed as:
v
ZRT
p
(11)
On differentiation:
v
T
RZ
p
RT
p
Z
T
(12)
p
Using the previous two expressions, the second integrand on the right of (9) becomes:
v T
v
T
ZRT
p
RZ
p
RT
p
Z
T
RT 2
p
Z
T
(13)
p
Thus the change in specific enthalpy between two states can be evaluated by:
h2
h1
2
1
c P dT
2
1
RT 2
p
Z
T
(14)
dp
p
dh cP dT
RT 2
p
Z
T
dp
(15)
dhdp
(16)
i.e., the real enthalpy change dh can be regarded as difference of two terms associated with the
temperature drop ( dhdT ) and pressure drop ( dhdp ), respectively. The first term depends on the
specific heat, the second one on pressure and temperature, and on the particular gas constant and
partial derivative of the compressibility factor Z with respect to temperature. Note that the term
dhdT is not the enthalpy change calculated under the ideal gas model, being cP dependent on both
pressure and temperature.
Table 1 shows the main working fluid properties and summarizes the optimal cycle parameters for
the four expansion processes analyzed in this study. Note that the fluid critical temperatures vary in
a large interval between 31C and 135C. The critical pressure of CO 2 is almost double the critical
pressure of the other working fluids. In addition, the reduced pressure (pR) at turbine inlet for CO 2 is
the highest (around 1.7) among the fluids considered, which implies much higher evaporating
pressure. The reduced temperature (TR) at turbine inlet varies in the interval between 0.9 for
isobutane and 1.2 for CO2. The reduced pressure at turbine outlet is still high for CO 2 (just below
1), whereas it is much lower for isobutane and R134a. Only CO 2 has a turbine outlet temperature
higher than the critical temperature.
Table 1. Fluid properties and optimal cycle parameters in the considered optimized ORCs.
1
2
3
4
Theat_in (C)
100
100
150
150
mheat_in (kg/s)
0.025
0.025
100
100
Working fluid
CO2
R125
R134
Isobutane
TCR (C)
30.98
66.02
101.06
134.66
pCR (bar)
73.77
36.18
40.59
36.29
Optimal turbine inlet and outlet states
pturb_in (bar)
123.32
44.56
47.97
18.85
tturb_in (C)
92.5
85.0
129.4
98.5
pturb_out (bar)
69.04
16.37
8.323
4.376
tturb_out (C)
47.72
35.95
50.8
49.2
Optimal turbine inlet and outlet reduced states
pR_turb_in
1.67
1.23
1.18
0.52
TR_turb_in
1.20
1.06
1.08
0.91
pR_turb_out
0.94
0.45
0.21
0.12
TR_turb_out
1.06
0.91
0.87
0.79
Power generated and absorbed
Pturb (kW)
0.5863
0.4587
4822.9
3863.1
Ppump (kW)
0.2965
0.128
757.0
311.2
PACC (kW)
/
/
567.8
556.6
Pnet (kW)
0.2898
0.3308
3498.1
2995.3
hIS
100
(17)
The real elementary enthalpy change is calculated using the isentropic efficiency of the overall
expansion process ( IS = 0.80 for CO2 and R125; IS = 0.85 for isobutane and R134a):
hel
hIS _ el IS
(18)
82
Equation (18) assumes that the polytropic efficiency ( p) is equal to the isentropic efficiency, that
is:
IS
hIS _ el
h IS
(19)
h
h T
(22)
RTin2
pin
Z in
T
(23)
p in
Real enthalpy change calculated integrating (15) over an elementary expansion process. This
value is an approximation of the real enthalpy change (20) because finite (and not infinitesimal),
although small, enthalpy changes are considered:
h approx
h T
h p
(24)
Percentage error between the enthalpy change calculated using (24) and the enthalpy change
calculated from (20):
Err %
happrox
h
h
100
(25)
Z
T
1, 30, 60, 90 and 100, respectively called I, II, III, IV, V. The thermodynamic conditions at the
outlet from the last interval (turbine outlet) may slightly differ from those in the original works
shown in Table 1 due to the different property methods used and the assumption that the polytropic
efficiency is equal to the isentropic efficiency. Table b shows, in the first three columns, the real
enthalpy change (20), the enthalpy change associated with the temperature drop (21) and their ratio
(22); in the last three columns the enthalpy change associated with the pressure drop calculated
using (23), the approximated real enthalpy change (24) and the percentage error (25).
83
84
- h p (kJ/kg)
0.1968
0.1969
0.1969
0.1969
0.1969
Err%
0.91
0.94
0.96
0.97
0.96
CarbonDio xide
125
140 bar
80 bar
100
Z=0.6630
Z=0.6574
T [C]
75
Z=0.6557
120 bar
Z=0.6575
100 bar
50
25
60 bar
40 bar
0
-1.75
-1.50
-1.25
-1.00
-0.75
-0.50
s [kJ/ kg-K]
Fig. 1. Turbine expansion of CO 2 in the optimized cycle represented on a T-s diagram. Note that
the values of the compressibility factor Z are almost constant along the expansion process.
1
2
30
31
60
61
90
91
100
101
44.56
44.06
32.56
32.22
24.16
23.93
18.28
18.12
16.71
16.56
85.0
84.39
67.84
67.28
52.66
52.20
40.17
39.79
36.47
36.12
360.7
360.6
357.6
357.4
354.3
354.2
351.0
350.9
349.9
349.8
85
2.231
2.223
1.817
1.801
1.431
1.421
1.207
1.202
1.156
1.151
0.4959
0.4978
0.5618
0.5642
0.6346
0.6370
0.6983
0.7002
0.7170
0.7188
0.01265
0.01267
0.01089
0.01078
0.007881
0.007795
0.005799
0.005744
0.005282
0.005234
- h p (kJ/kg)
0.1116
0.1030
0.1011
0.1021
0.1025
Err%
1.8
6.1
7.8
6.9
6.5
R125
125
50 bar
100
40 bar
Z=0.4959
T [C]
75
Z=0.5642
30 bar
Z=0.6370
50
20 bar
Z=0.7004
15 bar
25
10 bar
0
1.00
1.20
1.40
1.60
1.80
s [kJ/kg-K]
Fig. 2. Turbine expansion of R125 in the optimized cycle represented on a T-s diagram. Note the
low value of the compressibility factor Z at turbine inlet.
86
Table 4a. Thermodynamic conditions of R134a at inlet and outlet of intervals I, II, III, IV and V.
Interval
Points
p (bar)
t (C)
h (kJ/kg)
cP (kJ/kg-K)
Z
Z T p (1/K)
I
II
III
IV
V
1
2
30
31
60
61
90
91
100
101
47.97
47.05
28.05
27.56
16.76
16.48
10.21
10.04
8.669
8.529
129.4
128.3
100.2
99.28
76.62
75.92
57.44
56.85
51.67
51.11
318.3
318.0
309.2
308.9
299.8
299.4
290.3
290.0
287.2
286.8
1.868
1.852
1.440
1.429
1.197
1.192
1.064
1.061
1.032
1.029
0.5837
0.5871
0.6875
0.6909
0.7766
0.7791
0.8407
0.8425
0.8577
0.8593
0.006943
0.006891
0.004788
0.004717
0.003125
0.003084
0.002145
0.002120
0.001908
0.001886
- h p (kJ/kg)
0.3295
0.3190
0.3168
0.3161
0.3160
Err%
4.6%
1.3%
0.6%
0.4%
0.3%
R134a
150
55 bar
Z=0.5837
45 ba r
100
Z=0.6911
T [C]
35 bar
25 bar
15 bar
50
Z=0.7794
Z=0.8428
8 bar
0
0.25
0.50
0.75
1.00
1.25
1.50
s [kJ/kg-K]
Fig. 3. Turbine expansion of R134a in the optimized cycle represented on a T-s diagram. Note the
substantial increase of the compressibility factor Z from turbine inlet to turbine outlet.
percentage error introduced by using (24) is in all cases well below 1%. According to the procedure
followed in the previous cases, Fig. 4 shows on a T-s diagram the expansion process of isobutane
divided into 10 identical enthalpy intervals. The values of the compressibility factor are reported
near the inlet points of intervals 1, 4, 7 and 10. The points get closer toward the end of the
expansion process: the effect due to the increase of the ratio h h T is stronger than the decrease of
the specific heat. In the final stages the real enthalpy drop is more than 65% the enthalpy drop at
ideal gas conditions, which is the highest ratio among all cases considered in this study.
Table 5a. Thermodynamic conditions of isobutane at inlet and outlet of intervals I, II, III, IV and V.
Interval
Points
p (bar)
t (C)
h (kJ/kg)
cP (kJ/kg-K)
Z
Z T p (1/K)
I
II
III
IV
V
1
2
30
31
60
61
90
91
100
101
18.85
18.55
12.09
11.92
7.864
7.754
5.169
5.098
4.499
4.437
98.50
97.75
80.29
79.76
65.84
65.40
53.48
53.09
49.64
49.26
679.0
678.5
664.7
664.2
650.0
649.5
635.2
634.7
630.3
629.8
2.842
2.812
2.324
2.314
2.090
2.084
1.948
1.944
1.911
1.907
0.6754
0.6799
0.7790
0.7817
0.8461
0.8479
0.8909
0.8921
0.9025
0.9035
0.005170
0.005043
0.002874
0.002827
0.001839
0.001815
0.001260
0.001245
0.001121
0.001108
Err%
0.12
0.20
0.17
0.15
0.14
Isobutane
150
40 bar
30 bar
20 bar
100
T [C]
15 bar
10 bar
Z=0.6754
Z=0.7818
Z=0.8480
Z=0.8922
50
3 bar
0
1.0
1.5
2.0
2.5
3.0
3.5
s [kJ/kg-K]
Fig. 4. Turbine expansion of isobutane in the optimized cycle represented on a T-s diagram. The
values of the compressibility factor Z show that ideal gas conditions are nearly approached at
turbine outlet.
88
h/ h
Figure 5 shows the ratio (22) between real enthalpy change and the term of the enthalpy change
associated with temperature drop as a function of the compressibility factor Z. A good correlation is
found for each working fluid. As the compressibility factor increases the real gas approaches the
ideal gas behavior and the real enthalpy change approaches the enthalpy change calculated as the
product of the specific heat and temperature drop (15, 16). Thus, the real gas effects associated with
the second term on the right side of (15) can be easily correlated with the compressibility factor Z.
These effects result in a decrease of the enthalpy change from the value calculated as the product of
the specific heat and the temperature drop. Higher values of Z result in higher fractions of
temperature related enthalpy change converted into real enthalpy change.
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
CO2
R125
R134a
Isobutane
0.4
0.5
0.6
0.7
0.8
0.9
Compressibility factor Z
Fig. 5. Correlation obtained between compressibility factor Z and the ratio between real enthalpy
change and enthalpy change evaluated from the temperature drop.
According to these findings, the real enthalpy change in an elementary expansion interval can be
expressed as:
h cP _ in Tin Tout f ( Z )
(26)
where cP_in is the specific heat of the working fluid at the inlet of the temperature interval (Tin-Tout)
and f(Z) is a function that depends on the compressibility factor Z.
The functions h / h T obtained from the analysis of the considered expansion processes (see
Section 3) undergone by different fluids, are shown in Fig. 5. These functions are interpolated using
89
the second order polynomials shown in Table 6. The expression of f(Z) for CO2 is not shown since
the expansion process of CO 2 spans a narrow range (almost constant) of Z. It is noteworthy that
h / h T is a second order function of Z.
Table 6. Correlations between real enthalpy change and enthalpy change calculated from the
temperature drop.
Working fluid
f(Z)= h h T
R125
f ( Z ) 2 .4454 Z 2 2.1233 Z 0.5304
R134a
f ( Z ) 2 .9650 Z 2 2.8602 Z 0 .8115
Isobutane
f ( Z ) 3.2149 Z 2 3 .1110 Z 0.8682
The turbine power output is:
Wt
mwf
h
n
mwf
cP _ in Tin
Tout
f (Z )
(27)
where h is given by (26) and the sum is over all the temperature intervals (n).
The options to maximize the turbine power output consist in increasing all factors included in (27),
that is: working fluid flow rate, temperature drop between turbine inlet and outlet, specific heat and
Z along the expansion process. Accordingly:
A good match between the thermal profiles of the heat source and the working fluid leads to both
higher working fluid mass flow rates and to higher turbine inlet temperatures (close to the inlet
temperature of the heat source).
The use of a more effective cooling system allows the expansion to progress to lower
temperature levels, which are also more effective in generating power due to the higher values of
Z.
The reduced pressures (pR) and reduced temperatures (TR) involved in the expansion process
depend on the working fluid selection and fix the value of Z and f(Z).
Higher values of the specific heat increase the enthalpy drop ( h T ), however they might result
in lower working fluid mass flow rates evaporated by heat transfer with the heat source as well.
All these strategies contribute to the increase of turbine power output, but they do not affect the
above conclusion that operating conditions involving high values of the compressibility factor are
always suitable to maximize the enthalpy drop and, in turn, the turbine power output, according to
(27).
Fig. 6. Generalized compressibility chart with superimposed the turbine inlet and outlet conditions
of the analyzed expansion processes.
5. Conclusions
This study analyzes the influence of the real gas effects in the calculation of the enthalpy change in
an expansion process. The enthalpy change is regarded as the difference between two terms related
with temperature and pressure drops, respectively. The study shows that the ratio between the real
enthalpy change and the enthalpy change calculated as the product of the specific heat and the
temperature drop can be easily correlated with the compressibility factor Z using a second order
polynomial. As the compressibility factor increases, the real gas approaches the ideal gas behavior
and the enthalpy term related with pressure drop gradually decreases.
The choice of the working fluid in an Organic Rankine cycle fixes the thermodynamic states in the
expansion process in terms of reduced pressure and reduced temperature and, in turn, the values of
the compressibility factor from the generalized compressibility chart. Thus, working fluids and
operating conditions which lead to high values of the compressibility factor are always suitable to
maximize the turbine power output.
This condition does not guarantee by itself the achievement of the highest turbine power output,
which is the result of many concurring factors. Among them the optimal match between thermal
profiles of heat source and working fluid, the work minimization in the compression process and
features of the cooling system are the most important ones, and may force the expansion process
where the compressibility factor is relatively low. Nevertheless, the information that can be inferred
from this study may help both in the preliminary choice of the operating fluid and in the subsequent
analysis of the ORC system performance.
Nomenclature
cP
Err%
h
heat
m
ORC
p
R
R
s
t
T
v
Z
Greek symbols
efficiency
Subscripts and superscripts
ACC
air cooled condenser
approx approximated
CR
critical
dep
departure
t
related to temperature drop
p
related to pressure drop
el
elementary
idg
ideal gas
in
inlet
IS
isentropic
net
net
out
outlet
p
polytropic
pump
feed pump
R
reduced
rg
real gas
turb
turbine
wf
working fluid
References
[1] DiPippo R., Geothermal power plants: principles, applications, case studies and environmental
[2]
[3]
[4]
[5]
[6]
[7] Utamura M ., Tamaura Y., A solar gas turbine cycle with super-critical carbon dioxide as a
93
Abstract:
In this paper the methodology carried out to design a solar heating system (SHS) is shown, it applies in a
thermophilic process of anaerobic digestion for a new prototype of digester. The thermophilic process
required to keep the mix into the digester between temperatures of 40 to 60C. The optimal temperature to
this process is 55C. The SHS was designed with the f method and with the help of the selection program
Solar carried out at the Instituto Politcnico Nacional. To meet the application needed, the SHS is
integrated by flat plate solar collectors of parallel flow with a total collecting area of 12 m2. In this project a
new design of digester was needed. The most important parameters of design are rate and length of stirring
and hydraulic retention time (HRT). Taking in account these parameters, the digester designed has a volume
of 2 m3, the feedstock, which is the organic fraction (OF) sorted at the source of the municipal solid waste
(MSW) of Mexico City, is fed with centrifugal pumps and a heating exchange system was designed to
achieve a more homogenous heat transmission in the digester. As a calculation tool, a computer program
was developed in order to obtain the amount of energy required to several volumes of mix to heat into the
thermophilic range. Improving the design and operation of an anaerobic digester can be result of an organic
waste management more controlled and obtain the best design of anaerobic digester to applying it in rural
zones. The SHS reduces fuel consumption in the process of methane generation. The advantages of this
design are: reducing the time of methane generation and, thereby, improving the efficiency of the process,
also reducing the solid retention time inside the digester and reducing its volume. Among the important
achievements of this new design are: the deviation from landfilling of the OF since Mexico City, nowadays
generates about 5,000 ton/day of it.
Keywords:
Anaerobic digestion, Digester design, Methane generation, Solar heating system, Thermophilic
process.
1. Introduction
Among the most pressing global environmental challenges that faces humankind are the increase in
emission of greenhouse gases causing the climatic change and the exponential generation of waste.
In M exico the generation of solid municipal solid waste (M SW) is calculated in about 85,000 tons
per day. Nowadays, the problematic has advanced, overcoming the common methods of
recollection and its final disposal, in the specific case of M xico City (with a population of 8.9
millions of people), it generates about 11,000 tons per day of M SW [16]. It is known that 40% of
M SW correspond to the organic fraction (OF) of this, therefore only in M exico City, about 5,000
tons are generated on a daily basis, with a moisture content between 70% and 80%, which
potentially could produces about 385,000 Nm3 of methane (CH4) per day. For the above, and since
there is no treatment or method to avoid this methane to be released into the atmosphere, it is
contributing to pollution, as methane is a powerful greenhouse gas [17].
The trouble of handling the M SW in M exico has considerably increased due to the closure of an
important landfill (Bordo Poniente stage IV). The Bordo Poniente stage IV received 4872,300 tons
of M SW in 2009 and a slightly higher quantity in 2010. At the time of the closure (December 31,
94
2011), it ended up with 69 million tonnes buried under an area of 375 ha, its emissions to the
atmosphere are of about 1068,384 m3 of biogas per day [17].
Due to the dimensions of the problem, in this work an improvement to the method of anaerobic
digestion is proposed using solar energy to work under the thermophilic range. The anaerobic
digestion is a process to generate energy, which is carried out by managing the organic fraction of
the M SW to obtain a stabilized product and thus, produce biogas. The biogas is made of methane
and carbon dioxide (50 70% v/v), which can be used as fuel. The effectiveness in destructing
pathogens, as well as higher production in biogas production, depends on the temperature range, in
this case the thermophilic range (55 C) that is more efficient than the mesophilic range (35 C),
[14],[7].
The thermophilic range needs an external energy source that in many cases is a conventional fuel
which results in higher operation costs and does not avoid the problem of greenhouse gases emitted
to the atmosphere, then using solar energy to heat the digester is a good alternative, because these
sources are environmental friendly [18]. Therefore, the contribution of this work, is the design of a
solar heating system (SHS) to provide the energy needed to maintain the thermophilic range in the
process and to reduce costs. The obtained result was a new model of digester adapted with a heat
transfer system to use the SHS and the methane produced. It was designed as well; a computer
program to design digesters that calculates the energy needed to the process depending on the M SW
quantity to convert it in methane.
2. Proceeding
According to El-M ashad et al. (2003) the use of solar energy for heating anaerobic reactor represent
a kind of solar energy storage in the form of biogas. Hamed M . El-M ashad et al.(2002), present a
design of a stirred tank reactor for anaerobic treatment of liquid cow manure under thermophilic
conditions (50 C), using a solar heating system mounted on the reactor roof, in this case a
simulation models for two systems were developed.
In the present study, the solar thermophilic anaerobic digester (STAD) designed has been evaluated
for different configurations the dimensions used a computer program. The STAD designed to obtain
variable conditions which was used in the program is compound by 6 m2 of solar collector area. The
digester designed has a volume of 2 m3; the feedstock is the organic fraction (OF) at thermophilic
conditions (55 C). The main objective was to design a STAD whit high energy efficiency
including lower constructions costs, as well as a simple control system.
First, the design of solar heat transfer system was made. Second, the design of the digester was
modified to maintain the temperature of thermophilic process about 55 C. In the design all
parameter was took in a count, such as the feed, the stirring and heating. Therefore, a new design of
the digester was the result of analysis. Details of the procedure are below.
3. Design Methodology
3.1. Solar Heat Transfer System
The importance of the design is maintaining the mixture into the digester at 55 C, so that several
compounds are involved in the process to achieve it, [7]
Based on the energys requirements of the thermophilic process to generate biogas, the solar heating
system (SHS), with flat plate solar collectors was designed to apply to the Solar program [1]. This
program was designed for flat plate solar collectors; it gives the solar collection area and storage of
water per area of collection [10]. The solar collection area depends on the monthly radiation, on the
installation zone of the SHS and also on the monthly heating loads needed.
95
The available measured data for solar energy were the monthly average daily solar radiation on a
horizontal surface and the monthly average daily hours of bright sunshine [4].
The total solar radiation incident on a horizontal plane at the surface is the sum of direct and diffuse
radiation. Also, it is necessary to determine the solar radiation on tilted surfaces such as flat solar
collectors. Studies on the availability of solar resources in these FPSC have shown that the most
appropriate orientation is north - south, which means that, if the FPSC are to be installed in the
northern hemisphere, the collector should be oriented towards the geographical south and vice
versa, [12]. The radiation is calculated by (1):
cos
cos
Rb
(1)
Z
To determine the total solar radiation on a tilted surface, based on solar radiation data on horizontal
plane, it is necessary to define the coefficient R, which is the ratio of total radiation on the tilted
surfaces in relation to radiation total surfaces in the horizontal plane it is shown in (2):
R
HT
H
(2)
Therefore:
HT
R H
(3)
Hb
R
H b
Hd
R
H d
(4)
The total solar radiation on a horizontal plane (H) was measured for this study. Data measured are
shown in the table 1 (see Appendix A, table A.1). These values were plotted as it shows in the Fig.
1.
The method used to both design and selecting is the f design method.
The f chart-method or f design method is used to estimate the annual thermic performance of
the heating solar system activity by using a working fluid, by means of the calculus of the f energy
fraction, supplied by the sun to meet the required heating load. The main design variable is the
uptake area of solar energy. Among the secondary variables are: the type of collector, the storage
capacity, the flow rate and the size of the heat exchangers. The f fraction is the relation between the
useful energy delivered to the solar system (Qt), which is the difference between the energy of the
system only using conventional fuel (Laux ) and the heating load required by the system (L) [12]. For
a given month, the reduction of the f fraction of the supplied solar energy is shown in (5):
f=
Laux
L
Qt
L
(5)
The f fraction is the correlation obtained from results of hundreds of experiments in this area as
well as simulations with solar heating systems operating under different circumstances. This
fraction depends of two dimensionless parameters. The first one is related to the index of losses of
heating load, UL, of the collector and the second one with the solar energy absorbed, G, and the
heating load. M eteorological data of the zone are also required on a monthly basis.
The calculus of the f fraction is in function of two dimensionless parameters: X and Y. The
procedure which Klein describes to determine these two variables is shown by the equations (6) and
(7).
96
Fig. 1. Solar radiation on the study area from 2008 to 2010 [CINVESTAV, 2010]
FR AC
L
FR AC
L
IT
dt
FR AC
H
L
U L Tref
Ta dt
FR AC
U L Tref
L
(6)
Ta
(7)
The value f was obtained as a function of X and Y as it shows in the Fig. 2, [12].
The annual fraction of energy is provided for the solar heating system. This is the sum of
contribution monthly that brings the system, divided by the annual thermal load of heating as it
shown in equation (8):
12
f annual
i 1
Qu
(8)
Lannual
By means of experimental assays, design ranges of the variables used in the development of the f
method for air and liquid systems have been obtained, which are the transmittance and absorbtance
product (0.6 to 0.9), the collector-heat exchanger efficiency factor, (5 to 120 m2), and the overall
coefficient of heat losses, UL (2.1 to 8.3 W/m2-C) [12].
The collectors heat transference efficiency factor or removed heat factor, FR, is defined as the
relation between the actual energy available for the collector and the energy that would be available
if the whole absorber plate of the flat solar collectors, FSC, was at the same temperature as that of
the fluids entrance to the flat plate solar collectors. [4].
97
98
Fig. 3. Annual behaviour to the f energy fraction which is provided by the sun
The heat transfer in the heat exchanger involves convection in the fluids side and conduction
through the wall which separates it. The overall heat transfer coefficient (U), takes an account the
contribution of this effects, as follows [2]:
U=
1
1
int
(10)
ext
(11)
UAs Tlm
T1
ln
T2
T1
T2
(12)
In this evaluation the overall heat balance including all temperatures and heat fluxes, such as Tamb
ambient, Tpch air in pump chamber, Tb biogas and Td digester temperatures. The heat fluxes that was
considered are solar flux on a tilted surface IT, top losses from the solar collector Qtlc , sides losses
from the solar collector Qlc , bottom heat loss from the solar collector Qblc , electrical energy
consumption in agitation Qecp, heat losses from the liquid via biogas bubbles Qlbb, heat losses by
convection from the liquid to biogas Qlcb, useful heat gain rate from the collector Qu, auxiliary heat
add to the digester Qaux , rate of heat gain from the pumps Qrhp, heat losses from the pump chamber
to environment Qlche , heat losses from the pump chamber to the reactor Qlcd, heat recovered from
the effluent Qre .
The calculates was made more facility whit the development of a computer program in M ath lab, in
this program was considered all temperatures and heat fluxes, as well as the production of biogas.
(see Apendix B, Table B.1). This program is easier tool to release the calculus.
99
4. Results
4.1. Digester Designed with the Solar Heating System
The amount of energy needed to heat the digester mixture to 55 C was previously determined, a
value that is meet by means of solar energy. Therefore, the heating load for the SSCP is 7.2 x 109 J
/ month. The temperature of the exhaust manifold is 60 C. The digester is 1.8 m3 with a height of
1.45 m. The heat exchanger is composed by 22 aluminium tubes, an arrangement of 6 flat plat solar
collectors, a 150 l/m2 storage tank, a 7 m2 collection area. The design of this exchanger was made
thinking about temperature uniform and more heat transfer of a fluid to another. The mixed is for
two centrifugal pumps. The bottom on the digester has a conical shape to encourage the decantation
of the sludge and avoid the passage of these through of the mixtures recirculation system. The heat
exchanger is collocated in vertical position with its fixed outboard.
Into the digester the mixture is very corrosive due the corrosive components that are naturally
formed in the process.
Control system: the control system is responsible for monitoring of systems functions; it controls
the temperature into of the digester, the temperature of solar collectors, opening and closure of
valves to solar and auxiliary heating system and on-off pumps system.
This new digester design was designed to build and industrial scale, below is shows the integral
system to energy generation in Fig. 4 y Fig. 5.
100
Fig. 5. Biogas generation system with the solar and auxiliary heating system
The result more important of computer program that was calculated is the generation of methane
about the retention time. These values were determined taking in account standard conditions. In the
Fig. 6 is shown a chart of generation of methane.
Fig. 6. This chart shows the generating methane from the digester design
The digesters of M SW are designed below the base to minimum retention time from 20 to 30 days.
6. Conclusions
The calculation methodology to the design of integral system called Digester-SHS was developed.
Also the calculation tool was made which allows analysing of behaviour to volumes and loads
different. The contribution of this study was an economic and environmentally friendly digester
101
model, as well as this digester can be used in the industry. The list following show the advantages
of the digester designed:
Higher speed of reaction to generate methane
Lower solids retention time
Lower volume of the digester
Contribution to the final dispose of the SM W in the country
Contribution to the reduction of greenhouse effect to the atmosphere
Appendix A
This table show the data of solar radiation that were plotted of 2008 to 2010 to obtain the behaviour
of the radiation solar and can be consider whit the better option.
Table A.1. Data measurement of the total solar radiation on the zone
Year
2008
2009
2010
M onth
H
[J/m2-day]
H
[J/m2-day]
H
[J/m2-day]
M onthly average
temperature (C)
January
February
M archs
April
M ay
June
July
Augusth
September
October
November
December
1.55E+07
1.87E+07
2.00E+07
2.05E+07
2.02E+07
1.84E+07
1.72E+07
1.85E+07
1.47E+07
1.64E+07
1.63E+07
1.55E+07
1.54E+07
1.88E+07
2.21E+07
2.26E+07
2.21E+07
1.86E+07
1.82E+07
1.80E+07
1.42E+07
1.60E+07
1.58E+07
1.43E+07
1.38E+07
1.88E+07
2.42E+07
2.46E+07
2.02E+07
1.84E+07
1.72E+07
1.75E+07
1.47E+07
1.64E+07
1.63E+07
1.31E+07
14.3
15.7
17.6
19.8
20.1
19.8
18.5
19.3
18.3
18.4
15.6
15.4
102
Appendix B
In this appendix an example of data calculated for the compute program is shown.
103
Nomenclature
.
c
h
T
Q
U
X
load
Y
load
References
[1]
Barbosa J. G. M todo de diseo de sistemas solares para calentamiento de agua. Tesis de
M aestra, 1999. Instituto Politcnico Nacional, SEPI. ESIM E. 151p.
[2]
Cengel Y. A. Transferencia de calo y masa. Tercera edicin, 2007, 970p.
[3]
Couper J. R., Fair J. R., Penney W. R. (2010). Chemical process equipment. ButterworthHeinemann is an imprint of Elsevier, 166-164, 121, 134. 730 p.
[4]
Duffie J. A., Beckman W. A. (2006). Solar engineering of thermal processes. Ed. Wiley
Intersciencie publication, USA. 908 p.
[5]
El-M ashad H. M ., Van Loon WKP, Zeeman G.. A model of solar energy utilization in the
anaerobic digestin of cattle manure. Biosystems Engineering, 2003. 84(2), 231-238.
[6]
Foster C. T. (2005). Digestin anaerobia termoflica seca de residuos slidos urbanos:
estudio de las variables del proceso en el arranque y estabilizacin del bio-reactor. Tesis Doctoral.
Universidad de Cdiz. 421 p.
[7]
Gutin S., M arinek-Logar R. Effect of PH, temperature and air flow rate on the continuos
ammonia stripping of the anaerobic digestion effluent. Process Safety and Enviromental Protection
89 (2011) 61-66.
[8]
Hamed M . El-M ashad; Wilko K.P. van Loon; Grietje Zeeman (2002). A model of solar
energy utilization in the anaerobic digestion of cattle manure. Science Direct, 84 (2), 231238.
[10] Hottel H. C. (1976). A simple model for estimating the transmitance of direct solar radiation
trough clear atmospheres. Solar energy, 18, 129-134.
[11] Kaparaju P., Buenda I., Ellegoard L., Angelidakia I. Effects of mixing on methane
production during thermophilic anaerobic digestion of manure: Lab-sacale and pilot-scale studies.
Bioresurce Technology, 2008. 4919-4928.
104
[12] Klein S. A., Beckman W. A., Duffie J. A (1977). Solar heating design by the f chart method.
Ed., Willey Intersciencie Publication, USA. 636 p.
[13] M ata-lvarez, J., M ac, S., Llabrs, P. (2000). Anaerobic of organic solid wastes. An
overview of research achievements and perspectives. Biores. Techn., 74:3-16.
[14] Romero L. I., Sales D., Galn M . A. Thermophilic anaerobic digestin of winery waste
(vinasses): Kinetics and process optimization. Process Biochem 1988. 23 (4), 119-125.
[15] Soteris A. K. (2009). Solar energy engineering; processes and systems. Elsevier, USA,
760p.
[16] SEDESOL-INE. Informe de la situacin general en materia de equilibrio ecolgico y
proteccin al ambiente. Mxico, 1994. 220p.
[17] Tovar L. R., Castro R. A., Gutirrez M . E., Estrada R. What if the organic fraction of the
wholesale central market of Mexico city (Central de abastos del D.F.) is anaerobically digested?.
Proceedings Venice 2010, Third International Symposium on Energy from Biomass and Waste
Venice, Italy.
[18] Van Lier J. B. Thermophilic anaerobic waster water treatment; temperature aspects and
process stability. PhD Thesis, 1995. Wageningen A gricultural University. Wageningen, The
Netherlands.
105
Abstract:
In this study, the hotels located in Mexico, with energy demand over 100 kW are analyzed to determine the
conditions for which it is profitable to produce their own electricity through photovoltaic panels. The electric
tariff paid by hotels, has different prices for electricity during the day, i.e., is consider the hour of day in which
electrical energy is consumed and the maximum demand that is taken in the month.
By using HOMER software, economic profitability is analyzed under different levels of solar radiation and
demand curve to energy electricity in hotel.
The result shows that only for the hotels located in regions with higher solar radiation in Mexico the use of
photovoltaic panels is profitable. During the study, one of the most important factors to profitability is the time
of day that peak demand comes. However, when making a future projection of the electricity tariff, if they
keep raising the price of electricity as in recent years, in early 2014 photovoltaic panels could be the best
choice for hotels in most parts of the country.
Keywords:
Hotel, photovoltaic, Mexico
1. Introduction
M exico has very favorable natural conditions for the application of photovoltaic systems. In many
parts of its vast territory, the average solar radiation is twice that of European countries like
Germany [1], which is now one of the largest PV markets in the world.
Given the limited information currently available, regarding the financial feasibility of PV systems
connected to the grid in M exico, in this paper examines whether the use of photovoltaic systems in
hotels, with energy demand greater than 100 kW (kilo-watt), enables cost savings from the
perspective of an investor, compared with the purchase of all electricity to the national grid.
To perform the analysis is used the Homer software.In this study, is limited to analysis of
photovoltaic systems as an alternative to the purchase of electricity to the national grid.Other
measures and technologies, such as energy efficiency or other renewable energy technologies are
not considered.
2. Regulatory framework
In 2001 is approved by the Regulatory Commission for Energy (CRE - ComisinReguladora de
Energa) the "interconnection agreement for renewable energy source or system on a small scale
cogeneration"that allows users to general use, install a maximum power generation of 30 kW
produced by renewable energy source under the regime of Net Metering energy. In 2004 comes into
force amendments to the Law on Income Tax (Ley del ImpuestoSobre la Renta), which states that
taxpayers who invest in machinery and equipment for power generation from renewable sources,
can deduct 100% of the investment in a single year [2].
The Tariff H-M (Horaria-M edia Tensin, Time-M V), applies to energy services intended for any
use, supplied at medium voltage, with a demand of 100 kW or greater [3]. This tariff is applied to
hotels that are studied in this work.
The tariff has a charge by energy maximum demand (refers to the maximum power demanded,
measured in kilowatt, it is considered from keeping it for 15 minutes) and charges for energy
consumed in time: peak, intermediate and base; therefore, the energy does not have the same cost
during the day.Table 1 shows the different fees for this tariff.
The annual increase in maximum demand charges in the last three years is on average 4.12%,for
energy in top is 8.33%, for energy in middle 12.61%, for energy in base 12.14%.
Figure 1 show the different periods of peak, intermediate and base for a long of the year, summer
and not summer. The base period is from 00:00 hours until 06:00 throughout the year; the peak
period, in not summer, is from 18:00 to 22:00 hours, while for summer season from 20:00 to 22:00;
the rest are intermediate period.
1
2
Mexican pesos/kilo-watt
Mexican pesos/kilo-watt-hour
107
108
191.79
207.92
2.0206
2.3711
Scenario 1
Scenario 2
5. Analysis of PV systems
We analyze two hotels which have different demand curves of electricity.Figure 2 shows the energy
demand curves for each hotel, in the property a the peak demand of energy occurs from 16:00 to
18:00 hours, while for the property b the maximum demand of energy occurs from 17:00 to 19:00
hours, therefore the hotel a has no maximum demand in the peak hours, while for bthere is
maximum demand in peak hours.
The energy demand curves for each hotel are formed from data measured on the property for a
month every 20 minutes.The hotels are located in M exico and were measured by the National
Autonomous University of M exico.
5.1. Scenario 1
Figure 3 shows the results of the simulation carried out in Homer,for hotela,the right axis shows the
maximum price that can be paid by the photovoltaic system to make it profitable to install,as seen
with increasing solar radiation the price you can pay for the system is greater.
Figure 4 shows the results of the simulation carried out for the hotelb, the right axis shows the
maximum price you can pay for the system for this is even profitable. As with the hotela, with
increasing solar radiation can be paid a higher price for the system, can also see that both hotels the
consumption level of energy is not relevant to the study, because, for different energy consumption
and the same solar radiation, the maximum price arrangement is the same.
Contrary to expectations, for these two hotels the level of energy consumption (x-axis) does not
affect the financial viability of PV systems, i.e., the financial viability of PV systems is not a
function of energy consumption, is depends mainly on solar radiation and the total price of the
array.
3
4
Mexican pesos/kilo-watt
Mexacan pesos/kilo-watt-hour
109
The results for the two hotels are very similar, although it is a bit more profitable system in the
hotela, is due the peak demand of energy occurs earlierand may cover some of this demand with the
PV system.
At current prices of PV systems,installation is profitable, if you subtract part of initial investment of
the Income Tax (ISR) and has a solar radiation in the area greater than 4.65 kWh/m2/day.
110
5.1. Scenario 2
M aking a future projection is repeated the study for both hotelswith priced electrical tariffs that are
expected for 2014.
Figures 5 and 6 show the results obtained in the simulation to the hotel aandb respectively.As in the
previous scenario the results are similar for both hotels, does not influence the energy consumption
in the project's profitabilityand as solar radiation increases the maximum price you can pay for the
system increases.
If there is no change in the prices of the system and have no change in the Law of Income Tax,for
hotels that subtract part of the investment with the Income Tax, the installation of systems
photovoltaic will be profitable, area regardless of the country where they are located.
Comparing the two scenarios is clear that the financial viability of PV systems is affected by the
price of energy in the electricity tariff, i.e., with increasing prices of energy also increases the
maximum price of the PV system with which is profitable yet.
Fig. 5.Analysis of the photovoltaic system for the hotel a with the expected electricity tariff in
January 2014.
111
Figure 5.Analysis of the photovoltaic system for the hotel b with the expected electricity tariff in
January 2014.
6. Conclusions
The financial viability of PV systems is not affected by the level of energy consumption in hotels,
i.e., the financial viability of PV systems depends mainly on solar radiation in the area and the total
system price.As seen in Figures 3 to 5, a given solar radiation has a maximum price of the system
which remains constant for all levels of power consumption.This is true only for hotels with tariff
H-M .
Comparing the two scenarios can be seen that as the price of energy from the grid grows the
maximum price of the photovoltaic system can be greater and therefore becomes more viable
economically, i.e., the price of electricity from the grid Electricity is directly proportional to the
financial viability of PV systems.
At current prices of PV systemsis profitable installation in hotels that cover the tariff H-Mthat found
in areas where average solar radiation in the year is higher than 4.65 kWh/m2/day, regardless of the
level of energy consumption, is a little more profitablefor those hotels that can meet peak demand
of energy with the photovoltaic system.
If you continue the current trend in the electrical tariff increase,by January 2014, will be
economically feasible to install photovoltaic systems in hotels located in any area of M exico.
References
[1] CONUEE, Nichos de mercado para sistemas fotovoltaicos en la conexin con la red elctrica
en Mxico, 2009
[2] Diario Oficial de la Federacin(2004). Fracc. XII, Art. 40, Ley del Impuesto sobre la Renta,
M xico.
112
[3] CFE - The official web site of the ComisinNacional de Energa. Available at:
<http://www.cfe.gob.mx/Paginas/Home.aspx>.[accessed 4.1.2012].
[4] ECOTEC ECOTECSA S.A. de C.V. Available at: <http://www.ecotec2000.de/espanol/sun7.htm> . [accessed 4.1.2012].
[5] Quezada S. Anlisis de la rentabilidad econmica de producir energa elctrica en el sector
residencial y comercial por medio de paneles solares y aerogeneradores en M xico. Tesis, Facultad
de ingeniera. 2011.
113
Universit degli Studi di Firenze, Dipartimento di Energetica Sergio Stecco, Florence, Italy,
riccardo.secchi@unifi.it
b
Institute of Thermal Technology, Silesian University of Technology, Gliwice, Poland,
jacek.smolka@polsl.pl
Abstract:
In this paper, a study of increasing a photovoltaic (PV ) module efficiency via natural/ forced cooling of t he PV
cell is presented. The PV module is cooled by the air flowing in a duct placed under a back surface of the PV
panel. The cooling air is moved either by fan or naturally by the temperature difference generated from the
heat trans fer with a panel. The system could be applied on t he roof of industrial facilities, with electrical and
thermal energy needs. The pres ented 1-D model examines an already published thermal and electrical PV/ T
approach supplement ed with a radiative heat transfer in the panel duct. In addition, an effect of the
roughness of the panel back surface is also added to the model. The model coded in an Engineering
Equation Software (EES) is capable of evaluating the PV module efficiency and other thermal parameters as
the outlet air temperature and back surface temperature. As a result, the performance of the system in a
typical winter and summer day is discussed. Furthermore, an annual simulation of the system is also
analysed coupling the EES model with the Trnsys software. The results of the annual simulation show that
the raise of the relative roughness of t he panel back surface from 0 to 0,05 leads to an increase of 0.25% of
the average efficiency of the PV module in cas e of air moved by a fan. In terms of overall efficiency, this
result means a relative improvement of 1.73%. Annual average thermal efficiency of the PV/ T system of
28.6% is achieved for relative roughness of the panel back surface set to 0.05.
Keywords:
Solar photovoltaic thermal (PV/ T) air collector, heat trans fer, system optimization.
1. Introduction
In Europe, the solar photovoltaic (PV) electricity generation capacity has increased 160 times from
185 MW in 2000 to 29.5 GW in 2010 [1]. This is mainly due to the public support programs for
renewable energies, which usually consist of feed- in tariffs for the new installed PV systems. Since
only a small fraction of the incident solar radiation is used by the PV cells to generate electricity,
the PV module efficiency is in the range of 4-17% depending on the type of solar cells and the
working conditions. This means that at least 50% of the incident solar energy is mainly turned into
heat in the cells and substrate, leading to two consequences: (a) a drop in cell efficiency (usually
0.4% per C increase for c-Si cells) and (b) a permanent structural damage of the module if the
thermal stress persists for extended period [2]. In solar photovoltaic thermal (PV/T) collector, the
PV module is cooled by air or water, producing thermal and electrical energy simultaneously. The
higher efficiency leads to a higher production of the system, which causes an increase in the
economic incomes from the feed- in tariff for the PV system holder. In addition, the heated cooling
flow rate (air or water) can be used both for domestic heating and production of hot water. For
these reasons, in the last 40 years a lot of studies (theoretical as well as experimental) on the PV/T
systems has been carried out. Kern and Russel [4] presented the design and the performance of
water and air cooled PV/T systems, while Florschuetz [5] included the PV/T modelling in his
114
works. Garg and Adhikari [6] presented a variety of results regarding the effect of design and
operational parameters on the performance of the air type PV/T systems. Hegazy [7] investigated
glazed photovoltaic/thermal air system for a single and a double pass air heater for the space heating
and drying purposes. Kalogirou [8] modelled the hybrid photovoltaic/thermal (water) solar energy
system with TRNSYS and the simulations carried out for Nicosia (Cyprus) showed an increase of
the mean annual efficiency of the PV solar system from 2.8% to 7.7% with the thermal efficiency of
49%. Zondag et al. [9] developed a range of steadystate and dynamic simulation models for the
PV/T (water) energy performance analysis that included 1-D, 2-D and 3-D models of the serpentine
PV/T collector (their accuracy was verified by experimental data). Tiwari et al. [10] validated the
theoretical and experimental results for the photovoltaic (PV) module integrated with an air duct for
a composite climate of India and concluded that the overall thermal efficiency of PV/T system is
significantly increased due to utilisation of thermal energy from the PV module. Dubey et al. [14]
studied different configurations of the glass-to-glass and the glass-to-tedlar PV modules. Analytical
expressions for electrical efficiency with and without airflow were developed as a function of
climatic and collector design parameters. Experiments that were performed at the Indian Institute of
Technology, Delhi showed that the glass-to- glass type achieves higher supply air temperature and
electrical efficiency. By the use of validated theoretical models, Tonui and Tripanagnostopoulos
[16] studied a degree of improvement by adding suspended metal sheet at the middle of the air
channel and the finned arrangements at the opposite wall of the air channel. It was found that these
low-cost improvements are more effective at small collector length, and can be readily applied to
PV/T (air) installations. On the other hand, an effect of the channel depth, mass flow rate or system
length on the fan power consumption was found small.
In this study, on the basis of the PV models developed by Tiwari et al [10,15], and
Tripanagnostopoulos and Tonui [16], a theoretical steady state 1-D model including a convective
and radiative heat transfer in a panel duct is developed with the EES software [19]. In addition, the
annual gain in energy production of the investigated system is modeled integrating the EES model
with TRNSYS that is a transient simulation program, and using typical meteorological year (TMY)
conditions data for the central Italy. Furthermore, an effect of the roughness of the back surface on
the PV module effectiveness is also investigated. Finally, the results of the cooling of the PV
module through air moved by fan or naturally flowing are presented.
Air outlet
PV module
Insulating
panel
LD
Air inlet
b
Fig.1 Perspective view of the hybrid PV/T system studied.
115
T SKY
h rad
h conv,upp
Ut
Glass
T CELL
Silicon Cell
UT
Tedlar
Airflow m
TBS
hf
TAI R
Airflow m
h rad_duct
hf
Insulation
Ti
Ub0
h conv, und
T AMB
116
CELL
p I
1 p
U t TCELL TAMB
I b dx
U T TCELL TBS
b dx
CELL
CELL
(1)
I b dx
2 - An overall heat
loss from top
surface of cell
to ambient
3 - An overall heat
transfer from cell
to backsurfac e of tedlar
4 - The rate of
electrical energy
produced
The second energy balance equation describes the heat transfer for back surface of tedlar:
UT
TCELL
TBS
b dx
hf
TBS
TAIR b dx
hRAD _ DUCT
TBS
TI
b dx
(2)
3 - An overall heat
TBS
m AIR C AIR
TI b dx
(3)
TAMB b dx
7 The mass
flowrate of
flowing fluid
8 An overall heat
transfer from upper
surface of insulation
to ambient
Finally, the fourth energy balance equation for the upper surface of insulation can be written as:
h RAD _ DUCT TBS
TI
b dx
U b 0 TI
TAMB b dx
hf
TI
T AIR b dx
(4)
8 An overall heat
The mathematical transformations to obtain the explicit formulation for all temperatures are
described in Appendix A. In addition, the relations used in Eqs (1)-(4) for the heat transfer
117
coefficients, and the assumptions adopted for the calculation of the radiative heat transfer
coefficient are given in Appendix B.
In order to calculate the temperature-dependent electrical efficiency of the PV module, the
following expression has been used [17,24]:
TCELL TREF
(5)
CELL
REF 1
The rate of thermal energy obtained from the hybrid system is:
QU
TAIR _ IN
(6)
The thermal efficiency of the PV/T system has been calculated using the following equation:
TH
QU
b LD I
(7)
Value
2m
1.143 m
0.003 m
1 W/(mK)
0.95
0.88
0.85
300 X 10-6 m
0.036 W/(mK)
0.5
0.5 X 10-3 m
0.033 W/(mK)
0.87
0.1
0.05 m
0.035 W/(mK)
0.0782 m
0.83
0.1446
298 K
1000 W/m2
In winter reference conditions, the lower ambient temperature leads to a higher electrical efficiency.
Furthermore, for RR=0.05, an air flow speed of 2 m/s is sufficient to obtain the maximum
efficiency.
0.15
0.145
0.14
0.135
0.13
0.125
0.12
0.115
0.11
0.105
0.1
0.095
Summer,
RR=0
Summer,
RR=0.05
Winter,
RR=0
Winter,
RR=0.05
Fig. 3. Electrical efficiency for RR=0 and RR=0.05, for summer and winter weather reference
conditions.
119
The effect of the roughness cannot be determined for average air flow speed lower than 0.2 m/s,
because the Reynolds number is too low at that speed (see overlapped curves in Fig.3).
In Fig.4, the power gains between the configuration with RR=0 and RR=0.05 for both reference
conditions are shown. The gap of power gains between summer and winter shown in Fig.4 include
the power costs for ventilation, given by the total pressure drop in the duct, in accordance with the
following equation:
1 m AIR
0.6 AIR
W FAN
(8)
Power (W)
20
18
16
Summer
14
Winter
12
10
8
6
4
2
0
0
0.2
1.8
2.2 2.4
2.6 2.8
Fig. 4. Net power gains (in Watt) between RR=0 and RR=0.05 for summer and winter weather
reference conditions.
An airflow velocity equal to 1.5 m/s, which is a mean value between the optimum of the single
reference conditions, is considered the optimum value for the annual simulation.
The convective (hf) and radiative (hRAD_DUCT) heat transfer coefficients are the key parameters that
affect the heat transfer inside the duct. For this reason, they are extracted from the results and
shown in Table 2.
Table 2. Convective (hf) and radiative (hRAD_DUCT) heat transfer coefficients for RR=0 and RR=0.05
obtained for summer and winter weather reference conditions.
hf , W/(m2 K)
hRAD_DUCT , W/(m2 K)
Summer
Winter
RR=0 RR=0.05 RR=0 RR=0.05
7.797 22.170 8.190 24.110
0.790
0.733
0.578
0.546
fraction of solar radiation, which impacts the PV cells and the one that strikes the spaces between
the cells. Furthermore, a consequence of this assumption concerns the calculation of
temperatures in the different layers of the model: no distinction is made between the temperature
of the PV cells and that of the back surface of Tedlar. As a result, the model one equation.
Another important assumption concerns the airflow temperature, which is assumed to vary
linearly in the flow direction. Tan and Charters [18] correlation is used to compute Nusselt
number and the resulting the forced convection heat transfer coefficient in the air channel. The
friction factor that is needed for the calculation of the pressure drop along the duct, is calculated
from the equations given by Incropera and DeWitt [23].
Tiwari at al. - Radiative heat transfer in the air duct is neglected. In consequence, the temperature
of the insulation upper surface is not calculated and the model is reduced by one equation. The
convection heat transfer coefficient inside the duct has been originally assumed to be a constant
value, but in recent works [11] is calculated according to flow regime and its Nusselt number (no
information is given in [11] about the formulas used).
The three models have been tested with the design parameters of the present study in order to
compare the values obtained for layer temperatures and other basic parameters of the PV/T system.
The comparison is shown in Tables 4 and 5 for winter and summer reference conditions,
respectively.
Table 4. Comparison of simulations results obtained with the three different models, for winter
reference condition (TAMB=TAIR IN=283 K, I=600 W/m2 , Vw=0 m/s), Relative Roughness=0, and
forced circulation with the airflow velocity of 1.5 m/s.
MODEL
Tripanagnostopoulos
Tiwari et al.
Present work
TAIR OUT
T AIR
TCELL
TBS
TI
hf
EL
TH
(K)
(K)
(K)
(K)
(K)
(W/(m2 K))
285.5
285.5
285.5
284.2 302.4
285.5
284.2 310.8 306.5
284.3 310.1 305.7 285.1
9.49
8.19*
P
(Pa)
Table 5. Comparison of simulations results obtained with the three different models, for summer
reference condition (TAMB=TAIR IN=308 K, I=1000 W/m 2 , Vw=0 m/s), Relative Roughness=0, and
forced circulation with the airflow velocity of 1.5 m/s.
TAIR OUT
MODEL
Tripanagnostopoulos
Tiwari et al.
(K)
312.7
312.0
312.1
T AIR
(K)
310.4
310.0
310.1
TCELL
TBS
TI
hf
EL
TH
(K)
(K)
(K)
(W/(m2 K))
343.1
312.6
350.9 344.6
349.8 343.1 312.8
9.08
7.79*
P
(Pa)
Present work
* Since Tiwari et al. [11] gives no information about the formulas used for convective heat transfer coefficient and
pressure drop inside the duct, these values have been calculated by means of the EES function ductflow like in the
present work.
121
the software database. As an example, the weather data for Rome have been chosen in the present
study.
In an annual simulation, the incident radiation on the PV module varies considerably depending on
the position of the panel (azimuth, tilt to the horizontal) and the incoming radiation direction.
In this simulation, the available radiation on the panel was calculated as a function of the angle of
incidence of the three radiation components (direct, diffuse and reflected from the ground) in each
time step. Incidence Angle Modifiers (IAM) of each radiation component was determined as a
function of the Incidence Angle of the component [20,25].
The PV/T system has been considered oriented to the south with a slope of 30.
Annual simulations have been performed for the following cases:
1. Natural circulation: the model has been modified so that the flow of air is generated by the
temperature difference between the air inside the duct and the surrounding environment. For this
configuration, only the simulation with RR=0 is considered.
2. Forced circulation (air average velocity of 1.5 m/s) with RR=0 and RR=0.05.
It is important to point out that in the annual simulations the wind speed values from the database
have been used, while in the simulations in steady-state reference conditions the wind speed was
fixed to zero. This leads to higher values of electrical efficiency due to the increased heat removal
on the upper surface of a PV module. Moreover, the effects of wind are accentuated by the fact that
the incoming direction is not accounted for.
The monthly net electrical energy generated by the PV module for all three cases considered is
shown in Fig.5. According to this figure, it can be observed that the effects of the forced circulation
and non- zero roughness lead to the maximum increase in energy produced that occurs in the months
from May to September. Furthermore, Fig.5 shows a difference of the energy produced above 2
kWh between natural convection and convection with RR=0.05, and above 1 kWh between natural
convection and forced convection with RR=0.
Forced Circulation,
RR=0
65.00
60.00
Forced Circulation,
RR=0.05
55.00
50.00
45.00
Natural Circulation,
RR=0
40.00
35.00
O
CT
N
O
V
D
EC
SE
P
L
A
U
G
JU
A
PR
M
A
Y
JU
N
R
M
A
JA
FE
B
30.00
25.00
N
70.00
Month
Fig. 5. Monthly Net Energy in kWh generated for natural circulation (RR=0) and forced circulation
(RR=0 and RR=0.05).
Fig.6 presents the monthly averaged electrical energy efficiency for all three cases simulated. In this
figure, it can be observed that an increase in the electrical efficiency obtained with a forced
122
0.1450
0.1430
0.1410
0.1390
0.1370
0.1350
0.1330
0.1310
0.1290
0.1270
0.1250
Forced Circulation,
RR=0
Forced Circulation,
RR=0.05
JU
L
A
UG
SE
P
OC
T
NO
V
D
EC
AP
R
M
A
Y
JU
N
M
AR
FE
B
Natural Circulation,
RR=0
JA
N
circulation is higher in the months from May to September. The difference in the monthly averaged
electrical efficiency between the forced convection cases and the natural convection case is shown
in Table 3.
M onth
Fig. 6. Monthly averaged electrical efficiency for natural circulation (RR=0) and forced circulation
(RR=0 and RR=0.05).
As already shown for the steady-state weather reference conditions, in winter months the increase in
electrical efficiency of the PV module is limited, due to the low ambient temperature, which enables
the module to naturally work close to its maximum efficiency.
The resulting annual average electrical efficiency are the following:
1. 13.35% for Natural Circulation
2. 13.60% for Forced Circulation and RR=0
3. 13.85% for Forced Circulation and RR=0.05
Table 3. Difference in monthly averaged electrical efficiency between the forced circulation (FC)
cases with RR=0 and RR=0.05, and natural circulation (NC) case.
(FC,
RR=0)
- (NC)
(FC,
RR=0.05)
-(NC)
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT NOV
DEC
0.16
0.18
0.25
0.25
0.29
0.28
0.24
0.28
0.31
0.26
0.21
0.22
0.29
0.33
0.46
0.49
0.57
0.56
0.50
0.57
0.61
0.52
0.40
0.39
The monthly averaged thermal efficiency for all simulated cases is shown in Fig.7. The air flow
velocity under natural circulation reaches the maximum value of 0.4 m/s. Then the thermal energy
removed by the air flow results in the annual average thermal efficiency of about 8.3%. The annual
average thermal efficiency of about 18.5% is achieved for the forced circulation simulations with
123
RR=0, due to the air flow velocity of 1.5 m/s. The third configuration (forced circulation,
RR=0.05), leads to the maximum increase in annual average thermal efficiency of 28.6%.
Forced Circulation,
RR=0
0.2500
Forced Circulation,
RR=0.05
0.2000
0.1500
Natural Circulation,
RR=0
0.1000
0.0500
JA
FE
B
M
A
R
AP
R
M
A
Y
JU
N
JU
L
A
U
G
SE
P
O
CT
N
O
V
D
EC
0.0000
N
0.3000
Month
Fig. 7. Monthly averaged thermal efficiency for natural circulation (RR=0) and forced circulation
(RR=0 and RR=0.05).
4. Conclusion
In this paper, the electrical and thermal performance of three different configurations of a hybrid
PV/T system has been presented. A detailed one-dimensional thermal model was developed to
calculate the thermal parameters of the system. The photovoltaic electrical efficiency was calculated
using a linear correlation of the solar cell temperature. Numerical simulations concerning the
evaluation of the optimal air flow velocity varying the roughness of the back surface of the module
were carried out.
From the study the following conclusions can be drawn:
The airflow velocity that maximizes the net power generated gains between RR=0 and RR=0.05
results is 1.5 m/s;
For annual simulation, the raise of the relative roughness of the panel back surface from 0 to 0.05
leads to an increase of 0.25% of the annual average efficiency of the PV module in case of air
moved by fan. In terms of overall efficiency, this result corresponds to a relative improvement of
1.73%.
For RR=0, there is a difference of 0.25% in terms of the annual average electrical efficiency
between forced and natural circulation configurations.
The annual average thermal efficiency of the PV/T system increases by 10%, while a natural
circulation is replaced with a forced circulation (both with RR=0). With a roughness of 0.05,
there is an additional increase of 10%, which leads to an overall annual average thermal
efficiency of 28.6%.
This preliminary assessment of the performance of the PV/T system here proposed will be used as a
reference solution for the next planned experimental evaluation. However, in terms of power gain
(and, finally, in terms of cash earnings from the feed in tariff) the PV/T system proposed is very
interesting for industrial facilities with large roof surface, where it can easily be installed.
124
Appendix A
The mathematical steps for the explicit formulation of all model parameters are reported in this
Appendix.
From Eq. (1), the expression for solar cell temperature is:
EFF
TCELL
I U t T AMB U T TBS
Ut UT
(A.1)
where
EFF
CELL
1 p
CELL
(A.2)
Using Eq. (1) and Eq. (2), the expression for the temperature of tedlar back surface is:
h P1
TBS
EFF
I U tT T AMB
U tT
hf
h f T AIR
hRAD _ DUCT TI
(A.3)
hRAD _ DUCT
where
UT
U
U t U T , tT
h P1
U t UT
U t UT
Combining Eqs (1), (2) and (3), the following linear differential equation is obtained:
dTAIR
dx
TAIR
(A.4)
where
hP 3
b,
TAIR U tf
U b0
h P1 hP 2
I TI
EFF
hP 4
mAIR C AIR
h P2
U tf
hf
hf
hRAD _ DUCT
hf
, h P3
U tT
h RAD _ DUCT
h RAD _ DUCT
U tT
U tT h f
hf
hRAD _ DUCT
U tT
, h P4
TAIR _ IN
U tT hRAD _ DUCT
hf
hRAD _ DUCT
U tT
m AIR C AIR
hRAD _ DUCT
U tT h f
U b0
T AIR _ IN
(A.5)
The temperature of the air leaving the duct (at a distance LD form the entrance) can be obtained
from Eq. (A.5) for x=LD.
A fluid temperature can be averaged over the considered length and can be calculated as:
L
TAIR
1 D
T x dx
LD 0 AIR
1
LD
T AIR _ IN
1 e
125
LD
(A.6)
The expressions obtained in Eqs (A.5) and (A.6) for air temperature involves the value of TI
(temperature of the insulation upper surface), therefore, they can be solved using Eq. (4).
Appendix B
The heat transfer coefficients used in the modeling equations, are defined as follows:
hCONV ,UPP
2.8 3 Vw
(B.1)
hCONV ,UND
2.8 3 VA _ UNDER _ I
(B.2)
h RAD
TSKY
TCELL
TSKY
TCELL
(B.3)
The radiative heat transfer coefficient inside the duct was calculated considering the formula for
two infinitely long, gray, opaque, directly opposed parallel plates of the same finite width [22]. The
emissivity of the back surface of PV module was set at 0.87 [26] and that of the insulation upper
surface made of aluminum sheet was defined to be 0.1 [26].
1
h RAD _ DUCT
BS
TSKY
Ub0
Ut
UT
TAMB
TBS
TI
TBS
(B.4)
(B.5)
1
(B.6)
1
hCONV ,UND
1
LG
KG
(B.7)
1
hCONV ,UPP
hRAD
1
LCELL
K CELL
(B.8)
LT
KT
Nomenclature
b
LD
C
I
L
LI
KI
TI
width of PV module, m
length of PV module, m
specific heat, J/(kgK)
incident solar intensity, W/m2
thickness, m
126
hCONV_UPP
hCONV_UND
hf
hRAD
hRAD_DUCT
K
QU
T
p
P
Ub0
Ut
UT
UtT
V
Greek symbols
)EFF
absorptivity
product of effective absorptivity and transmittivity
efficiency reduction coefficient
emissivity
StefanBoltzmann constant, W/(m2 K4 )
efficiency
density, kg/m3
transmittivity
References
Journals :
[1] Jger-Waldau A. The Photovoltaic Market Appendix D. In: McEvoy A., Markvart T. and
Castaner L., Practical Handbook of Photovoltaics (Second Edition) Fundamentals and
Applications. Elsevier 2012, p. 1145-1151.
127
[2] Chow T.T. A review on photovoltaic/thermal hybrid solar technology. Applied Energy
2010;87:36579.
[3] Wolf M. Performance analysis of combined heating and photovoltaic power systems for
residences. Energy Conversion Management 1976;16:7990.
[4] Kern Jr. E.C., Russell M.C. Combined photovoltaic and thermal hybrid collector systems. In:
Proceedings of 13th IEEE photovoltaic specialists, 1978, Washington, DC,USA; 11537.
[5] Florschuetz L.W. Extension of the HottelWhiller model to the analysis of combined
photovoltaic/thermal at plate collectors. Solar Energy 1979;22:3616.
[6] Garg H.P., Adhikari R.S. Conventional hybrid photovoltaic/thermal (PV/T) air heating
collectors: steady-state simulation. Renewable Energy 1997;11:36385.
[7] Hegazy A.A. Comparative study of the performances of four photovoltaic/thermal solar air
collectors. Energy Conversion Management 2000;41:86181.
[8] Kalogirou S.A. Use of TRYNSYS for modelling and simulation of a hybrid PV thermal solar
system for Cyprus. Renewable Energy 2001;23:24760.
[9] Zondag H.A., De Vries D.W., Van Helden W.G.J., Van Zolengen R.J.C., Van Steenhoven A.A.
The thermal and electrical yield of a PV-thermal collector. Solar Energy 2002;72(2):11328.
[10] Tiwari A., Sodha M.S., Chandra A., Joshi J.C. Performance evaluation of photovoltaic
thermal solar air collector for composite climate of India. Solar Energy Material Solar Cells
2006;90(2):17589.
[11] Sarhaddi F., Farahat S., Ajam H., Behzadmehr A., Adeli M.M. An improved thermal and
electrical model for a solar photovoltaic thermal (PV/T) air collector. Applied Energy
2010;87:232839.
[12] Kumar R., Rosen M.A. A critical review of photovoltaicthermal solar collectors for air
heating, Applied Energy 2011;88:3603-3614.
[13] Dubey S., Tiwari G.N. Thermal modeling of a combined system of photovoltaic thermal
(PV/T) solar water heater. Solar Energy 2008;82:60212.
[14] Dubey S., Sandhu G.S., Tiwari G.N. Analytical expression for electrical efficiency of PV/T
hybrid air collector. Applied Energy 2009;86(5):697705.
[15] Joshi A.S., Tiwari A., Tiwari G.N., Dincer I., Reddy B.V. Performance evaluation of a
hybrid photovoltaic thermal (PV/T) (glass-to- glass) system. Int J Therm Sci 2009;48:15464.
[16] Tripanagnostopoulos Y., Tonui J.K. Air-cooled PV/T solar collectors with low cost
performance improvements Solar Energy 2007;81:498511.
[17] Evans D.L. Simplified method for predicting PV array output. Solar Energy 1981;27:555
60.
[18] Tan H.M., Charters W.W.S. Effect of thermal entrance region on turbulent forcedconvective heat transfer for an asymmetrically rectangular duct with uniform heat flux. Solar
Energy 1969;12,513516.
Books:
[19] Klein S.A., Engineering Equation Solver, Academic Version V8.603, www.fchart.com
[20] Duf e J.A., Beckman W.A. Solar engineering of thermal processes. 3rd ed. New York,
USA: Wiley; 2006.
[21] Nellis G.F, Klein S.A. Heat Transfer. Cambridge (UK): Cambridge University Press; 2009
[22] Cengel Y.A., Termodinamica e trasmissione del calore, McGraw-Hill,1998.
[23] Incropera F.B., DeWitt D.P.. Fundamentals of Heat and Mass Transfer. John Wiley & Sons,
New York, 1996.
Conference Papers :
128
[24] Schott T. Operational temperatures of PV modules. In: Proceedings of 6th PV solar energy
conference; 1985. p. 3926.
[25] King, David L., Jay A. Kratochvil, and William E. Boyson. Measuring the Solar Spectral
and Angle-of-Incidence Effects on Photovoltaic Modules and Irradiance Sensors. Proceedings
of the 1994 IEEE Photovoltaics Specialists Conference. Sept 30-Oct 3, 1997. pp. 1113-1116.
Web references:
[26] Material Emissivity Properties : http://snap.fnal.gov/crshield/crs- mech/emissivity-eoi.htm
129
Abstract:
The first concentrating solar power plant in Algeria began to produce electricty in the middle of the last year.
The hybridiz ation solar-gas is realized by the integrating of parabolic trough collectors into a c ombined cycle
power plant. The thermodynaic evaluation of the power plant performance at design running conditions is
based on the exergy analysis. In this analysis, the exergy flow diagram is aideed by the value diagram which
is used as a visualisation tool for identifying ex ergy degradation in heat trans fer processes involved in the
HRS G and the solar steam generator. The exergy destruction and the functional exergy efficiency are the
key paramet ers in assessing the performance of every power plant component. In regard to the electrical
power capacity of t he power plant, 160 MW, the solar energy s hare is 14% (22 MW), while the solar exergy
share is 12% (18.4 MW). The combustors and the solar field are the less efficient systems.The combustors
destruct about one third of natural gas exergy and the solar field consumes about three quarters of solar
exergy.
Keywords:
Integrated solar combined cycle, Concentrating solar power, Parabolic
Thermodynamic performance, Exergy analysis, Value diagram, Cycle-Tempo.
trough
collector,
1. Introduction
In Algeria the total installed power generating capacity is over 9 GW, 98% of which is provided by
gas- fired plants [1]. Algeria, located in the MENA region, has impressive solar resources [2-4]. The
Algerian desert is exposed yearly to a direct sun irradiation higher than 2000 kWh/m2 gained from
3500 hours of sunshine. These solar potential associated to huge land resources are suitable for the
implementation of concentrating solar power plants (CSPPs) [5]. The Algerian authorities are
planning to produce 6% of electricity from CSPPs by 2020 [1].This target should be reached
through the building of four CSPPs totaling a solar installed capacity of 240 MW [1]. The first
CSPP, at Hassi RMel, is already running since 2011. Since Algeria is a major producer and
exporter of natural gas, hybrid solar-gas power plants are more appropriate; besides this option
avoids the issues of storage systems. This choice is supported by the observation that in the last
years the peak of electrical consumption in Algeria was recorded in summer and in day times [1].
That meets with the times of the day and the year when solar energy has the advantage to produce
energy. The planned hybrid plants will contribute to save about 2.7 billion m3 of natural gas.
Similar projects are in progress in Morocco, Tunisia, Egypt, Jordan, and Iran [6-11] and elsewhere
of MENA region [12].
The solar thermal power generated by CSPPs can be supplied at different temperature levels. Puresolar power tower plants work at high temperature level (>500C) [13-16]. CSP systems at low or
medium temperature levels (250 to 400C), based on trough or Fresnel collectors, are more suitable
as options for solar repowering of new or existing fossil- fired power plants [17-22]. The parabolic
trough collector (PTC) is the most common medium-temperature solar technology. It represents the
most mature technology, from both commercial and technical viewpoints, for mid-to- large scale
130
grid connected power plants [23, 25]. Based on this technology a total installed electric capacity of
354 MW is generated from nine steam power plants in the California Mojave Desert [23].
In terms of cost and efficiency, integrated solar combined cycle systems (ISCCS) is the more
appropriate cycle design for MENAs countries [26]. An ISCC consists on the integrating of the
parabolic trough into a gas fired combined cycle composed of a gas topping cycle (GTC) and a
steam bottoming cycle (SBC). The PTCs can supply a hot heat transfer fluid (HTF) at a temperature
of about 400 C. Thus, the thermal energy carried by the HTF becomes the hot source of generating
additional steam to be sending to the SBC.
In last years, many studies had demonstrated the usefulness of exergy analysis in assessing the
performance of power plants [27-30]. The methodology has been proven in analyzing gas turbine
[31-33], steam turbine [34, 35] and combined cycle power plants [36-38].
In regard to CSPPs subjected to exergy analysis, the study [17] showed the benefits of a scenario
proposed for the solar repowering of a coal fired steam turbine. In another work [19], the best
options for solar boosting and fuel saving of two existing coal- fired power plants were determined.
In analyzing the hybridization solar-gas in ISCCs [22], a detailed examination was paid to the solar
collector, i.e. the linear Fresnel reflecting solar concentrator. The global optimal design of an
ISCCS was carried out based on the exergoeconomic approach [20]. Reference [18] performed an
exergy analysis of the first ISCCS in Iran. The plant is with full capacity 467 MW and the solar
power is about 17 MW. The main results are that the solar collectors are the least efficient
components in the plant and the combustors are the major destructors of exergy.
The present study presents the exergy analysis of the first ISCCS in Algeria. The Hassi RMels
plant has been lunched in operation few months ago. The analysis is based on the exergy flow
diagram and the value diagram, the later is an unusual visualization tool for identifying exergy loss
in heat transfer processes.
The thermodynamic simulations are performed by the flow-sheet program, Cycle-Tempo. This
software is a freeware advanced tool for the analysis and optimization of energy systems, developed
at the Delft University of Technology [39].
because the analysis is performed with the power plant running in design mode, with favourable
sunny conditions.
The GTPP is fuelled with Hassi RMels field natural gas. The gas is rich in CH4 by about 85%, it
has a LHV=45778 kJ/kg. The GTPP is supported by an inlet air cooling system to counter the
adverse effect of air temperature during hot times on its performance. Therefore, 4 Sadinter chiller
(CH) unites contribute to boost the GTPP by reducing intake air temperature to 15C. Although,
higher is the air temperature, higher is the power consumption of the chiller, this is usually
synchronous with higher solar irradiation (higher solar power output). An air cooled condenser
(ACC) is adopted as an option for condensing steam. This technology is preferable to the traditional
water-cooled condenser when water availability is limited, notably in Hassi Rmel, characterized by
an arid climate. The use of ACCs can reduce the plant water consumption by 90%; however, the
plant efficiency suffers with the higher condensing pressure. The SPX cooling system with heat
transfer duty of 147600 kW guaranteed by regrouping 15 fans, provides to the Hassi RMel power
plant statured water at 52 C and at about 0.14 bars. The design parameters of the power block are
summarized in Table 1.
At 183,120 m2 mirrors, the solar field comprises 224 parabolic collectors assembled in 56 loops, 4
collectors per loop. The collector is of ET-150 technology [24], aligned on a north-south line, it
tracks sun from east to west by a single axis tracking system. The HTF circulating in loop in the
solar field is synthetic oil; Therminol PV-1, its thermophysical properties versus temperature can be
found in Reference [42]. Cycle-Tempo assumes 3rd degree polynomial approximation for
calculating heat capacity of the HTF at any temperature. The supplying of thermal energy from the
solar field to the power block is performed when water/steam recovers the thermal energy from the
HTF through the SSG.
The SSG is the assembling of an economizer (ECO), an evaporator (EVA) with a drum (DR) and a
super heater (SHE).
The net output power of the plant is proportional to steam flow rate expanding in the ST, it is the
sum of the flow rate of steam generated in the HRSG and that generated in the SSG. The operation
of the hybrid plant is under the compulsory condition that the HTF circulates across the SSG at
constant inlet temperature, 393C, and at constant outlet temperature, 293C, but it can be with
variable mass flow rate. The high limit of temperature is imposed because the long-term exposure
of the organic HTF beyond 400C temperatures can lead to thermal decomposition of the fluid.
Thermal fluid decomposition occurs when enough heat is applied to the fluid to cause the breaking
of molecular bonds, which results degradation in the HTFs physical properties [43].
The solar steam flow rate is proportional to the HTF mass flow rate. The HTF mass flow rate varies
following the DNI intensity, in other words, according to time during day and to climate conditions.
The design value of the HTF mass flow rate is 200 kg/s, it is the resultant of DNI assumed at 751
W/m2 . This value of DNI is able to generate 22.6 kg/s of solar steam. Then, at the exit of the
economizer, 22.6 kg/s of pressurized water is withdrawn from the HRSG and is sent to the SSG.
After preheating and evaporating, the resultant saturated steam is superheated and resent to the
HRSG at 372C. The solar steam mixes with the steam exiting the SHE1, the whole steam passes
through the SHE2 before expanding in the ST. At design load the ST, with about 70 kg/s of
superheated steam at 560C and 80 bars, delivers 80 MW of electric power.
3. Exergy analysis
The exergy analysis of the power plant is based on the results of the thermodynamic simulations
performed by the flow-sheet program, Cycle-Tempo. The calculations considered the
simultaneous resolutions of mass equation and energy equation applied for each power plant
component. Cycle- Tempo also calculates exergy values of all fluid flows of the cycle by using the
ambient air (35C and 0.928 bars) as the reference state. Furthermore, the program calculates
exergy losses and efficiencies of all apparatuses. In the present study is used the functional exergy
132
definition where the exergy efficiency of any power plant component is determined as the ratio of
the exergy flow rate considered to be the product of the power plant component and the exergy flow
rate considered necessary for making this product [36].
Cycle-Tempo is able to draw Q-T diagrams and value diagrams.
Table 1. Design parameters of the power block.
Value
GT
Model
SGT-800
Ambient pressure
0.928
Ambient temperature
35C
Intake compressor air temperature
15 C
Compressor pressure ratio
20.2
Compressor isentropic efficiency
0.88
Inlet turbine temperature
1200C
Turbine isentropic efficiency
0.88
Exhaust mass flow rate
120,20 Kg/S
Exhaust temperature
550 C
LHV of natural gas
45778 kJ/kg
Net output power
40 MW
Thermal efficiency
35%
HRSG
Type
Fuel mass flow rate in the DBs
Approach temperature
Pinch temperature
Pressure losses in flue gas side
Pressure losses in water/steam side
Inlet water temperature
Exit stack temperature
Thermal efficiency
ST
Model
Inlet steam temperature
Inlet steam pressure
Steam mass flow rate
Condensate temperature
Isentropic efficiency
Full output capacity
SSG
Inlet water temperature
Inlet water pressure
Exit steam temperature
Water/steam mass flow rate
Inlet HTF temperature
Exit HTF temperature
HTF mass flow rate
Pressure losses in water/steam side
Pressure losses in HTF side
Thermal efficiency
SST-900
560C
83 bars
70 kg/s
52C
0.9
80 MW
195C
93 bars
372C
22.60 kg/s
392C
292C
205 kg/s
5.8 bars
2 bars
98 %
133
(1)
(2)
the symbols T0 and Ts are, respectively, the ambient temperature and the temperature of the Sun
(5777 K), and f is the dilution factor ( 1.3 10 5 ).
The useful heat and the associated exergy transported by the HTF to the SSG are, respectively:
(3)
and
.
(4)
The global energy and global exergy efficiencies of the solar field are defined, respectively, as
follows:
(5)
and
(6)
The global efficiency includes the optical and thermal efficiencies of the solar field, in other words,
the efficiency of the global conversion of solar irradiation to heat to be injected into the SSG. The
optical efficiency depends upon incident angle effects, solar field availability, collector tracking
error and twist, geometric accuracy of the mirrors, mirror reflectivity, cleanliness of the mirrors,
shadowing of the receiver, transmittance of the receiver glass envelope, cleanliness of the glass
envelope, absorption of solar energy by the receiver, end losses, and row-to-row shadowing [44].
The thermal efficiency is function of receiver thermal losses and piping thermal losses. Receiver
thermal losses are caused mainly by the thermal radiation of the receivers selective coating. Piping
thermal losses corresponds to the thermal losses from the solar field header piping and the HTF
system piping.
diagram, the hybrid power plant is able to converting 397.4 MW of natural resources exergy, of gas
and sun, in electrical power, 160 MW. The rest, 234.5 MW (60%), is lost.
The percentage of the solar exergy input is 24% (96.9 MW), that of natural gas is 76% (300.5 MW),
from which 60% (237MW) are injected through the CCs and 16% (63.5 MW) are injected through
the DBs.
Gas
5
33 .00
17 .74
453 .44
1 17. 735
320 .65
232 .21
P el = 400 00. 32 kW
AC
0. 928 0
GT
35. 00
48
-1 16.5 6
0 .87 80
AIR
Ai r
1 200 .00
CC
117 .735
17 .54
120 .200
25. 00
0.9 530
15 .00
5 50. 00
FLU EGAS
CH
5
Gas
33 .00
32
H TF
14. 00
392 .00
1 02.6 61
499 .59
DB
25.0 0
39
92. 10
3 04. 99
11. 310
27 39.3 6
SHE
31
51
87. 20
372 .00
11 .310
30 39.7 5
0.95 30
36
PMP
35
41
GAS
2 80
GTP P1
87.2 0
3 72. 00
34. 704
303 9.7 5
H
44
37
37
11. 310
13 16.1 5
92.1 0
3 04.9 9
11.7 82
273 9.36
379 .71
49
1 02. 661
466 .49
37 2.00
1 02. 661
309 .72
94. 10
19 5.00
11. 310
83 3.47
47
88 .20
301 .97
2 92.0 0
ECO
27
19 5.0 5
11
V alve
26
52
8 8.2 0
27 6.9 7
2 3.3 94
121 9.68
0 .95 30
25 9.4 1
12
16
h
= Ma ssf low [kg/ s]
94.1 0
34. 704
14 7.91
0.3 21
62 3.28
34. 704
63 4.66
0 .953 0
DEVA
13
P MP
27
2 01.9 1
7.0 00
60 .00
37 .880
25 1.7 3
CD
34
14 7.91
0.3 21
62 3.22
18
DECO
9
22
4 .500
4.5 00
14 0.00
34. 485
58 9.26
12
1 40. 00
3 7.8 80 5 89. 26
21
24
0.9 280
13
7.0 00
14 0.0 3
3.3 95
58 9.5 6
PM P
20
11
HRSG#1
38
23
39
19
35. 00
-1 16.5 6
ACC
31
WATER
P MP
0. 928 0
16
-1 11.4 9
22
FAN
28
14
- 116 .56
40 .00
14 536 .204
AIR
4.5 00
35 .00
14 536 .20 4
0.9 280
25
Stack
0. 928 0
24
33
147 .91
623 .22
20 7.12
22 0
0.95 30
30
4.50 0
35.0 25
DEA
15
26
P MP
29
WATER
el
14 9.23
25
83 3.47
5.0 00
STEAM
23
96. 10
19 5.00
2 743 .39
55
ECO1
14 7.91
0 .321
17
20
ECO2
4 .500
46
28
23 .39 4 83 3.4 7
h = En tha p
l y [kJ/kg ]
P = Electr ical Pow er [kW]
P MP
54
32 6.9 7
SSG
p = Pr es sur e [ bar ]
T = Te mpe rat ure [C]
EV A
19
89 .20
52. 00
DR
0 .95 30
43
0 .136 3
301 .90
10
SF
2 862 .38
18
2 51.6 7
20 1.85
0 .219
H
12.0 0
4 .500
14
53
10 2.66 1
17
FLU EGAS
35
89 .70
29
ST
301 .90
58 3.4 9
36
33
88 .20
23 .39 4 27 45. 84
10
0 .95 30
12 .50
42
P MP
21
DR
45
87 .20
E VA
30
34
13 .50
40
6 31. 20
S HE1
9
2 95. 06
Pe l = 4 000 0.04 kW
50
STEAM
0.9 530
560 .00
H
92. 10
83 .00
S HE2
38
7 50.3 8
10 0.00
32
7 .000
FLU EGAS
5 2.04
3 4.48 5 2 18. 47
0 .13 63
52.0 0
3 4.4 85
217 .69
Since the ACC has only the function to discharge heat to the environment, similarly to the stack,
evaluating its performance does not have a significant importance.
temperature difference at the cold end of the ECO should be the pertinent location for increasing the
efficiency. The potential of increasing the performance of the EVA is limited because its efficiency
is yet relatively high and because it is common to heat transfers associated to phase change to be at
lower efficiency.
SSG
SHE
300
EVA
ECO
200
100
35
0
0
3.4
19.5
25
Temperature [C]
1 - T0 / T [-]
600
500
400
Conclusion
The purpose of the present study is the thermodynamic analysis of the first hybrid solar- gas power
plant in Algeria. The Hassi RMels power plant has been lunched in operation few months ago.
The power block is composed of two SGT-800 gas turbines and a SST-900 steam turbine. The
HRSG is of single-pressure with supplementary firing and no reheat. The solar field gathers ET-150
parabolic trough collectors, in which circulates the heat transfer fluid Therminol PV-1.
The exergy analysis is adopted as a methodology for evaluating the performance of every power
plant component. The thermodynamic modelling and simulation, and also the post-processing of the
results are performed by the flow-sheet program Cycle-Tempo.
At design load the capacity of the plant is 160 MW of electrical power, 80 MW from the two gas
turbines and 80 MW from the steam turbine. Of this capacity the exergy solar share is lower than
the solar energy one, 12% (18.4 MW) against 14% (22 MW). The reason is that the solar steam
generator supplies about 50 MW of thermal power to the HRSG at 372 C, which is lower the
temperature of 560 C, at which the HRSG sends superheated steam to the steam turbine.
The thermal efficiency of the plant is 56% whereas the exergy efficiency is 53%. The combustors,
with efficiencies lower than 68%, are the least efficient components in the power block. They are
the major destructors of exergy; they consume about 32% of gas natural exergy. The turbomachines
(compressors, gas turbines and steam turbine) are the most efficient; their efficiencies are in the
range of 90% to 94%. The HRSG and the solar steam generator, equipped together by 17 heat
exchangers, have, receptively, the efficiencies 79% and 89%. With global exergy efficiency (optical
and thermal) limited to 25%, the solar field loses 74% of solar exergy.
Acknowledgments
The author is grateful to the management and staff of New Energy Algeria (NEAL SpA) and he
thanks especially Mr. Nordine Hamrene, Design Engineer, for providing the Hassi RMels power
plant data and for fruitful discussions. The author also thanks Mr. Teuss van der Stelt, from Delft
University of Technology, for giving help about Cycle-Tempo.
139
Nomenclature
A
e
h
mirror area
mass exergy, J/kg
mass enthalpy, J/kg
LHV
Greek symbols
efficiency
Subscripts and superscripts
0
Ambient
s
Solar
Acronyms
AC
Air Compressor
ACC Air Cooled Condenser
CC
Combustion Chamber
CH
Chiller
CSPP Concentrating Solar Power Plant
DB
Duct Burner
DEA De-aerator
DECO low pressure Economizer
DEVA low pressure Evaporator
DNI Direct Normal Irradiation
DR
Drum
ECO Economizer
EVA Evaporator
G
Generator
GT
Gas Turbine
GTC Gas Topping Cycle
GTPP Gas Turbine Power Plant
HRSG Heat Recovery Steam Generator
HTF Heat Transfer Fluid
HX
Heat Exchanger
ISCCS Integrated Solar Combined Cycle System
MENA Middle East and North Africa
PMP Pump
PTC Parabolic Trough Collector
SBC Steam Bottoming Cycle
SF
Solar Field
SHE Super Heater
SSG Solar Steam Generator
ST
Steam Turbine
140
References
[1] Commission de Regulation de lElectricit et du Gaz (Electricity and gas authority of Algeria).
Annual Report 2010 (in French). Available at:<http://www.creg.org.dz> [accessed 12.6.2011].
[2] Himri Y., Arif S. Malik, Boudghene Stambouli A., Himri S., Draoui B., Review and use of the
Algerian renewable energy for sustainable development, Renewable and Sustainable Energy
Reviews 2009;13:1584-591.
[3] Boudghene Stambouli A., Promotion of renewable energies in Algeria: Strategies and
perspectives. Renewable and Sustainable Energy Reviews 2011;15:1169181.
[4] Boudghene Stambouli A., Algerian renewable energy assessment: The challenge of
sustainability. Energy Policy 2011;39:4507-519.
[5] Quaschning V., Technical and economical system comparison of photovoltaic and
concentrating solar thermal power systems depending on annual global irradiation. Solar
Energy 2004;44:171-78.
[6] Kane M., Favrat D., Ziegler K., Allani Y., Thermoeconomic analysis of advanced solar-fossil
combined power plants. Int. J. Applied Thermodynamics 2000;3(4):191-98.
[7] Horn M., Fhring H., Rheinlnder J., Economic analysis of integrated solar combined cycle
power plants. A sample case: The economic feasibility of an ISCCS power plant in Egypt.
Energy 2004;29:935-45.
[8] Dersch J., Geyer M., Herrmann U., Jones S. A., Kelly B., Kistner R., Ortmanns W., Pitz-Paal
R., Price H., Trough integration into power plants-a study on the performance and economy of
integrated solar combined cycle systems. Energy 2004;29:947-59.
[9] Hosseini R., Soltani M., Valizadeh G., Technical and economic assessment of the integrated
solar combined cycle power plants in Iran. Renewable Energy 2005;30:1551-555.
[10] Al-Soud M. S., Hrayshat E. S., A 50 MW concentrating solar power plant for Jordan. Journal of
Cleaner Production 2009;14:625-35.
[11] Tsikalakis A., Tomtsi T., Hatziargyriou N.D., Poullikkas A., Malamatenios Ch., Giakoumelos
E., Cherkaoui Jaouad O., Chenak A., Fayek A., Matar T., Yasin A., Review of best practices of
solar electricity resources applications in selected Middle East and North Africa (MENA)
countries. Renewable and Sustainable Energy Reviews 2011;15:2838849.
[12] Assessment of the World Bank/GEF strategy for the market development of concentrating solar
thermal Power. Global Environment Facility Program. The World Bank; 2006. Available at
<http://www.worldbank.org/gef > [accessed 12.6.2011].
[13] Heller P., Pfnder M., Denk T., Tellez F., Valverde A., Fernandez J., Ring A., Test and
evaluation of a solar powered gas turbine system. Solar Energy 2006;80:1225-230.
[14] Schwarzbzl P., Buck R., Sugarmen C., Ring A., Crespo, M. J. M., Altwegg P., Enrile J., Solar
gas turbine systems: Design, cost and perspectives. Solar Energy 2006;80:1231-240.
[15] Xu C., Wang Z., Li X., Sun F., Energy and exergy analysis of solar power plants. Applied
Thermal Engineering 2011;31:3904-913.
[16] Spelling J., Favrat D., Martin A., Augsburger G., Thermoeconomic optimization of a
combined-cycle solar tower power plant. Energy; Available online 29 April 2011.
[17] Hu E., Yang Y., Nishimura A., Yilmaz F., Kouzani A., Solar thermal aided power generation.
Applied Energy 2009;87 (9):2881-885.
[18] Baghernejad A., Yaghoubi M., Exergy analysis of an integrated solar combined cycle system.
Renewable Energy 2010;35:2157-164.
[19] Suresh M. V. J. J., Reddy K. S., Kolar A. K., 4-E (Energy, Exergy, Environment, and
Economic) analysis of solar thermal aided coal- fired power plants. Energy for Sustainable
Development, 2010;14 (4):267-79.
141
[20] Baghernejad A., Yaghoubi M., Exergoeconomic analysis and optimization of an Integrated
Solar Combined Cycle System (ISCCS) using genetic algorithm. Energy Conversion and
Management 2011;52:2193-203.
[21] Popov D., An option for solar thermal repowering of fossil fuel fired power plants. Solar
Energy 2011;85:344-49.
[22] Siva Reddy V., Kaushik S. C., Tyagi S. K., Exergetic analysis of solar concentrator aided
natural gas fired combined cycle power plant. Renewable Energy 2012;39:114-25.
[23] Price H., Lpfert E., Zarza E., Cohen G., Gee R., Mahoney R., Advances in parabolic trough
solar power technology. Journal of Solar Energy Engineering, 2002;124:109-25.
[24] Geyer M., Lpfert E., Osuna R., Esteban A., Schiel W., Schweitzer A., Zarza E., Nava P.,
Langenkamp J., Mandelberg E., EUROTROUGH - Parabolic trough collector developed for
cost efficient solar power generation. In: Steinfeld A., editor. Proceeding of 11th Int.
Symposium on Concentrating Solar Power and Chemical Energy Technologies; 2002
September 4-6; Zurich, Switzerland. Published by Paul Scherrer Institut:1-7.
[25] Fernndez-Garca A., Zarza E., Valenzuela L., Prez M., Parabolic-trough solar collectors and
their applications. Renewable and Sustainable Energy Reviews 2010;14:1695-721.
[26] Dersch J., Geyer M., Herrmann U., Jones S. A., Kelly B., Kistner R., Ortmanns W., Pitz-Paal
R., Price H., Trough integration into power plantsa study on the performance and economy
of integrated solar combined cycle systems. Energy 2004;29:947-59.
[27] Szargut J., Morris D. R., Steward F. R., Exergy analysis of thermal, chemical, and
metallurgical processes. Springer ; 1988.
[28] Kotas T.J., The exergy method of thermal plant analysis. Reprint edition, Krieger, Malabar, FL,
1995.
[29] Bejan A, Tsatsaronis G, Moran M, Thermal design and optimization. New York: Wiley; 1996.
[30] Kaushik S.C., Siva Reddy V., Tyagi S.K., Energy and exergy analyses of thermal power plants:
A review. Renewable and Sustainable Energy Reviews 2011;15:1857-872.
[31] Reddy B. V., Butcher C., Second law analysis of a natural gas- fired gas turbine cogeneration
system. International Journal of Energy 2009;33(8):728-36.
[32] Ehyaei M.A., Mozafari A., Alibiglou M.H., Exergy, economic & environmental (3E) analysis
of inlet fogging for gas turbine, power plant. Energy 2011;36:6851-861.
[33] Khaldi F., Adouane B., Energy and exergy analysis of a gas turbine power plant in Algeria. Int.
J. Exergy, 2011;9(4):399-13.
[34] Al jundi I. H., Energy and exergy analysis of a steam power plant in Jordan. Applied Thermal
Engineering 2009;29:324-28.
[35] Regulagadda P., Dincer I., Naterer G.F., Exergy analysis of a thermal power plant with
measured boiler and turbine losses. Applied Thermal Engineering 2010;30:970-76.
[36] Woudstra N., Woudstra T., Pirone A., van der Stelt T., Thermodynamic evaluation of combined
cycle plants. Energy Conversion and Management 2010;51:1099-110.
[37] Ahmadi P., Dincer I., Rosen M. A., Exergy, exergoeconomic and environmental analyses and
evolutionary algorithm based multi-objective optimization of combined cycle power plants.
Energy 2011;36:5886-898.
[38] Srinivas T., Reddy B. V., Gupta A. V. S. S. K. S., Parametric simulation of combined cycle
power plant: A Case Study. Int. J. of Thermodynamics 2011;14 (1):29-36.
[39] Cycle-Tempo, Release 5.0., Delft University of Technology. 2007. Available at
<http://www.cycle-tempo.nl.>
142
[40] Shukin S., Annerfeldt M., Bjorkman M., Siemens SGT-800 industrial gas turbine enhanced to
47MW. Design modifications and operation experience. ASME Turbo Expo 2008: Power for
Land, Sea and Air GT2008; 2008 Jun 9-13; Berlin, Germany. Proceedings of ASME:65-70.
[41] Steam turbines for solar thermal power plants. Siemens AG; 2008. Order No. E50001-W410A105-V1-4A00. Available at <http://www.siemens.com/ energy> [accessed 12.6.2011].
[42] Therminol-VP1. Technical Bulletin. Available at <http://www.therminol.com> [accessed
12.6.2011].
[43] Moens L., Blake D., Mechanism of hydrogen formation in solar parabolic trough receivers.
National Renewable Energy Laboratory; 2008 Feb. Technical Report NREL/TP-510-42468.
Available at<http://www.osti.gov/bridge> [accessed 12.4.2012].
[44] Assessment of parabolic trough and power tower solar technology cost and performance
forecasts. National Renewable Energy Laboratory; 2003 Oct. Subcontractor Report No.
NREL/SR-550-34440. Available at<http://www.osti.gov/bridge> [accessed 12.6.2011].
143
Abstract:
The increase of waste production, joined to the difficulties concerning both the identification of new disposal
sites and the construction of big conventional incinerators, hardly accepted by the communities, led in recent
years to the development of new technologies for waste management.
The waste gasification and melting treatments, if compared with conventional incinerating methods, allow
reducing significantly the burdens on final disposal sites. Therefore gasifying and melting technologies are
attracting more and more the attention of academia and market operators.
Consequently, the possibility to introduce in the Italian context the Direct Melting System (DMS)
technology, designed and manufactured by Nippon Steel Engineering Co. Ltd., has been taken into account
for the scope of proposed work.
DMS technology consists in MSW gasification, slags melting and combustion of the syngas produced, with
the consequent generation of electric energy through a steam cycle. The system does not simply treat
wastes; in fact it minimizes environmental impact because there is an effective recycling of useful resources
since the melted slags are vitrified and inert and then still reusable, particularly for the most demanding tasks
in terms of leachability such as road pavement. DMS technology has also the real advantage of owning a
modular design. This aspect allows greater degree of flexibility in terms of matching communities needs,
bringing therefore a significant advantage during the phase of site identification.
The aim of this article is to consider different plant configurations in order to optimize the energy recovery
downstream the DMS module.
As a case study also landfill gas exploitation integrated in the DMS plant will be considered as a typical
situation that could occur in the Italian scenario. The energetic input provided by the biogas is generally
interesting because it allows improving the thermo-economic performances also thanks to market incentives.
Keywords:
DMS, gasification, landfill gas, melting, municipal solid waste, recycling, slag.
1. Introduction
Disposal of municipal solid waste is a problem that every local government has to face. Waste
volume annually produced has been increasing almost constantly in the various European States [1].
In densely populated areas proper disposal of a considerable mass of waste and the need of large
areas dedicated to the disposal are two important problems.
The EU has ruled issuing directives under which the legislative authorities of each M ember State
should act with regard to waste management [2]. These principles address to the so-called
integrated waste management that can be pursued by: reducing the amount of waste produced,
increasing the recycling and reuse, recovering energy from the not recyclable fraction and reducing
the volume of waste conferred to landfill without having undergone any treatment [3, 4]. In fact
landfilling of unsorted waste represents both a waste of resources, and a potential source of
environmental contamination.
144
On the one hand exists the energy recovery maximisation goal, which leads to the hypothesis of a
few large size plants network able to reach an highly efficient electric conversion, on the other hand
the ethical-economic-environmental principle of zero kilometer waste leads to the hypothesis of a
smaller size distributed waste-treatment plants network.
In the field of the innovative technologies for the municipal waste treatment, DM S technology has
attracted strong interest. The technology has been designed by Japanese Nippon Steel Engineering
Co. Ltd. and DM S means Direct M elting System.
DM S technology is very competitive because, due to modularization, allows small and medium size
treatment plants construction, performing energy recovery with overall efficiencies in line with
similar size conventional technology plants [5].
This article is reporting the preliminary study carried out on some technical improvements, starting
from the DM S original solution designed by Nippon Steel Engineering, aiming to the increase of
electricity production in order to suit better on a specific Italian case, suggested by Paul Wurth Italia
S.p.A.
The case study represents a potentially typical plant scheme for the Italian scenario and therefore
proposed solutions are expected to be generally implementable in order to match Italian electric
market requirements.
3. DMS technology
The technology has been derived from a thermal process used so far in the metallurgical industry,
which has been extended to the urban waste treatment. The process (Fig.1) consists in M SW
gasification and slags melting in a shaft furnace. The system provides a high degree of flexibility in
treating different types of wastes both combustible and incombustible. This includes wastes that
cannot be recycled at the collection level and those unsuitable for conventional technology based
waste to energy plants, such as incineration residues, sludge, automobile shredded residues, landfill
reclamation waste, CFC gas and asbestos [8, 9]. This technology can be defined safe and reliable
because 35 plants are nowadays in commercial operation (since the first installation in 1979) [10].
The process consists in a combined treatment of waste gasification, combustion of the syngas
produced and slags melting. The gasification and melting occur within the same equipment that
consists in a shaft furnace where waste is introduced from the top (after size reduction) together
145
with coke and limestone in percentage of around 4-5% of the waste input quantity (depending on
the characteristics of the waste) [11]. The slags are discharged through the bottom of the furnace
and granulated by a cooling system.
The syngas generated in the DM S is emitted from the top of the furnace and completely oxidized in
the post-combustion chamber, by controlling the temperature, the turbulence and the residence time.
Thanks to the homogeneous conditions of combustion which is taking place between gaseous
phases, the pollutants generation is limited in comparison to conventional mass burn incineration
and the plant, equipped with state-of-the-art flue gas treatment line, can achieve emissions into
atmosphere that are by far lower than regulatory limits and fully in compliance with BREF both in
terms of pollutants concentration and volume of flue gases [12].
The exhaust flue gas heat can be used to produce electricity by a recovery boiler and a steam cycle
[13].
The system does not simply treat wastes; in fact it minimizes environmental impact because there is
an effective recycling of useful resources since the melted slags are vitrified and inert and then still
reusable, particularly for the most demanding tasks in terms of leachability such as road pavement.
Even metal produced as final output of DM S can be reused as a resource [14]. At the end of the
process, the only residue to landfill is made up from the fly ashes caught by the filters and generally
it amounts at about the 2-3% of the waste input [14].
The system has generally a modular design, with multiple melting furnaces working in parallel,
which allows flexibility and reliability, while the auxiliary systems can be grouped as a single
module. Such an aspect allows greater degree of flexibility also in terms of matching communities
needs, bringing therefore a significant advantage during the phase of site identification.
146
4. Case of study
The plant will have to treat 60,000 t/y and the average composition of wastes is reported in
Table 1. The mean lower calorific value (LCV) of the feeding mixture is 10,691 kJ/kg and the heat
flow diagram for the baseline solution is reported in Fig. 2. The area identified for the construction
of the plant is located near to a landfill that provides a flow of biogas which should not be dispersed
into the atmosphere and therefore has to be used for power generation. The average biogas mass
composition consists of approximately 49.9% of methane, 27% carbon dioxide, 10.8% nitrogen,
9.5% moisture, 1.86% oxygen and argon, plus traces of other compounds.
Table 1. MSW average composition.
It can be noticed that the availability of the biogas will decrease from a value of 1,000 Nm3 /h till
406 Nm3/h (Fig. 3). Both mean biogas and waste composition values are the result of experimental
and statistical data processing.
In order to use the energy released by the combustion of biogas at high temperature, it was chosen
to apply only external components outside DM S to avoid the redesign of the module and to
maintain independent from the actual biogas availability the superheated steam output parameters
outgoing the module.
5. Plant solutions
The steam cycle shown in Fig. 4 is the basic configuration, designed for heat recovery from the flue
gas coming out from the DM S. The boiler consists of an evaporator (EVA) a superheater (SH1) and
an economizer (ECO). The steam generation does not occur only in the recovery boiler, but also in
the bottom of the combustion chamber covered by boiler pipes.
The flow rate coming from the DM S is about 48,400 Nm3/h and it is available at a temperature of
1,100C and it exchanges heat with steam till 200C.
In the superheater the steam reaches a maximum temperature of 428C. The steam cycles subjected
to DM S technology in Japan, reach a maximum temperature of 400C [15] and this choice is mainly
due to two reasons:
in Japan there is only moderate interest in producing electricity from waste to energy plants
because there are not incentives, hence the lack of need to reach high performances;
steam temperatures above 430C may cause problems of corrosion of pipe bundles [16].
In Italy electric generation efficiency must be maximized to exploit the beneficial contribution of
market incentives.
Considering on the one hand minimum steam quality (in terms of water to vapour ratio) accepted by
steam turbine manufacturer at the end of the expansion and on the other hand maximum
temperature allowed by material resistance to corrosion, the implemented thermodynamic cycle
optimization process has led to the selection of the steam condition equal to 428C and 65 bar at
turbine inlet.
The search for a further performance increase would lead to reach higher values of pressure, but this
would necessarily require to adopt a reheater which is not justified given the small size of the plant
[17].
148
In the gas treatment line, there are two heat exchangers which aim to preheat the feed water and at
the same time to cool the flue gases coming from the DM S, before they are ejected from the stack.
With regard to the condenser, the choice falls on the air-cooled one for two main reasons:
the use of water as coolant needs complicated path authorization;
the air-cooled condenser is not tied to the availability of water.
M oreover the benefits, in terms of cycle efficiency, obtained by using water instead of air are
negligible because, in each of the two cases, it is necessary to opt for a different cycle pressure. The
maximum steam cycle pressure allowed when using water as coolant is 45 bar because the increase
of this pressure would compromise the steam quality at the end of the expansion.
The second plant configuration (Fig. 5) aims to exploit the biogas produced in the landfill. The fluegases generated by the biogas combustion are not critical with regard to pipes bundle corrosion as
DM S ones, so it is possible to realize an external superheater (SH2) which can further increase the
steam temperature above 428C. Thus in an external component, i.e. separated from the main fluegas stream proceeding from the DM S, it is possible to reach higher temperatures without any
problems due to the pipe bundles corrosion.
Flue gases generated by biogas combustion are available at a temperature of about 680C and they
are cooled to 460C. With this energetic input it is possible to reach a temperature of about 470C
(with an average biogas flow rate of 700 Nm3/h); this value is far from the typical temperature of
about 540C which characterizes the steam cycle of fossil fuelled power plants, but that is because
the biogas flue-gas flow rate is small if compared to the flow rate coming out from the DM S.
149
Heat remaining in the biogas flue gases is then used in a heat exchanger to preheat the feed water.
The increase in temperature, both of the feed water and of the superheated steam, can increase the
overall electric efficiency and the power production.
The possibility to exploit the biogas directly into the DM S module was discarded in principle
because the post-combustion chamber and the boiler redesign and optimization would not be
justified given the uncertainty about the availability of biogas, hence the need for the use of a
configuration with an external component.
Another plant solution with an external component (Fig. 6) has been taken into account. The biogas
can be used in an external evaporator, to produce more steam. The flue gases coming from the DM S
module are cooled till about 310C and the maximum temperature reached from the steam is 428C
like the base configuration. Also with this configuration, is possible to use the heat remaining in the
biogas flue gas coming from the external evaporator, to preheat the feed water.
150
The last plant solution (Fig. 7), considered in present work, consists in coupling DM S plant with a
biogas internal combustion engine. Part of the biogas (about 430 Nm3/h) is burned in the engine
with an electric efficiency of 33% and it is possible to produce about 700kW. The biogas flue gases,
at a temperature of 500C are still usable in a heat exchanger with the task to further preheat the
feed water. The remaining part (about 280 Nm3/h) can be recovered in the external component,
evaporator or superheater. In this case, the plant solution is very flexible and also competitive from
an economic point of view, because the electric power generated from the engine awards
considerable incentives.
In most plants, the feed water degassing is made by means of a steam bled from the turbine. On the
contrary, the plant here described proposes a solution that does not require to bleed any steam from
the turbine, in order not to penalize the work obtained from the expansion with the aim to maximize
the energy production. In this solution (Fig. 8), a portion of the feed water flow rate, is bled at a
pressure of 65 bar (cycle pressure) and sent to a heat exchanger where it can exploit the heat of the
flue gases coming from the recovery boiler. The water reaches a temperature T x with an enthalpy
hx=hL(65 bar, T x); then it expands maintaining the same enthalpy. The steam produced in this way,
at an enthalpy condition hx can degas and preheat the water coming from the condenser.
151
Fig. 7. Steam cycle Configuration with external superheater coupled with ICE
6. Results
Results obtained from analysis are reported in Table 2. All results are referred only to the DM S
plant, so in the case of presence of the internal combustion engine (ICE) (with a biogas flow rate of
430 Nm3/h), the electrical power output has to be increased by 700 kWhel.
Table 2.
Configurations
Basic configuration
Cycle with external SH
Cycle with external SH and ICE not coupled
Cycle with external SH and ICE coupled
Cycle with external EVA
Cycle with external EVA and ICE not coupled
Cycle with external EVA and ICE coupled
700
270
270
700
270
270
DM S net
electric
power output
[kWhel/t MSW ]
626
714
663
686
701
657
679
DM S net
overall
electric
efficiency 1[%]
18.52
18.78
18.66
18.91
18.43
18.49
18.71
In the configurations with DM S and ICE coupled, the DM S net overall electric efficiency is better
than the case of all biogas recovered in the external SH/EVA. In fact the largest part of biogas
power input is exploited in a higher efficiency converter like ICE and DM S receives only as direct
energy input the ICE waste heat and a little part of biogas flow.
Basic configuration results are included in the typical performance range of 400700 kWhel/t MSW
for commercial gasification technologies [18]. Other solutions overcome this range up to +10% due
to the landfill biogas combustion contribute (directly or indirectly). To note that steam cycles
subjected to the DM S technology of Japanese operating plants, with steam at a pressure of 40 bar
and a temperature of 400C reach a net overall electric efficiency of approximately 16% (194,000
t MSW /y) [12].
7. CHP option
Coupled DM S and biogas engine solutions can achieve the best electric performances, so it was
interesting for these solutions to analyze the performance sensitivity versus the heat cogeneration. It
was examined a solution using a steam bleed from the turbine at a pressure of 4 bar and a
temperature of 146C to heat a pressurized water flow from 80C to 120C which can feed a district
heating network.
These coupled solutions allow to maximise the electric production even exploiting a thermal load.
In fact, owning to the electric production incentive frame that induce not to regulate, if the ICE
works independently, it will dissipate part of the exhaust gas heat when the district-heating load is
under the effective ICE production capacity. In this sense coupled solutions offer a more flexible
ratio between electrical power output and thermal power output than not coupled ones by reason of
the variable steam bleed.
DMS net overall electric effi ciency is obtained by the following ratio:
153
Table 3.
Configurations
Cycle with external SH
and ICE coupled
Cycle with external EVA
and ICE coupled
DM S electric
production
[%]
100
90
80
100
90
80
Net electric
power
[kWhel/t MSW ]
Thermal power
[kWhel/t MSW ]
686
617
549
679
612
544
0
278
552
0
274
548
DM S net
overall electric
efficiency [%]
18.91
17.01
15.14
18.71
16.86
14.98
20%
660
[kWh /t
18%
620
16%
580
14%
540
12%
500
10%
0
100
200
300
40 0
500
[kWh th/t M SW ]
DMS Net Electric Power vs. Thermal Powe r
DMS Net Overall Electric Efficiency vs. Thermal Power
Fig. 9. DMS net electric power production and DMS net overall electric efficiency vs. Thermal
power generation diagram for Cycle with external SH and ICE coupled
20%
660
18%
[kWh /t
620
16%
580
14%
540
12%
500
10%
0
100
200
30 0
400
500
[kWhth /t M S W ]
DMS Net E lec tric P ower vs. Th ermal Power
DMS Net Overall E le ctric E fficiency vs. Thermal P ower
Fig. 10. DMS net electric power production and DMS net overall electric efficiency vs. Thermal
power generation diagram for Cycle with external EVA and ICE coupled
According to the steam bleed progressive increase, electric performances decrease linearly with the
increase of the heat production.
DM S CHP option is particularly suitable because if necessary it achieves up to 550 kWh/t MSW of
thermal power in respect of a reduction of overall electrical efficiency of only 4 percentage points.
154
CHP option is generally very suitable to improve environmental and economical performance of
waste to energy plants compensating for their lower overall electric efficiency, caused by technical
constraints, and their higher specific capital costs [19].
Furthermore, a significant aspect is that with a bleed of steam corresponding to a thermal power of
about 440 kWh/t MSW it is possible to produce approx. 26 GWhth/y. This thermal energy corresponds
to the heating requirements of 5,000 inhabitants in climatic zone E2 with 2850 degree-days [20]. It
thus demonstrates the versatility of the DM S in particular to serve small communities needs on the
territory, while maintaining the electric standard efficiency usually accepted in Japan.
8. Conclusions
Italian and European laws prescribe to increase recycling and reuse and to decrease waste
production and landfilling. The share of non-recyclable waste must be treated in plants for energy
recovery in order to reduce volumes landfilled and reduce, as much as possible, danger of
environmental contamination.
In Italy, the construction of these facilities is supported by an incentive that rewards the production
of electricity.
The techniques to reach high conversion efficiencies to electricity require the realization of large
size plants. These need during their life cycle to be fed with a waste flow that often comes from
areas very far from the plant. This factor, together with the largest impact on the surrounding
environment, often causes opposition of local people to their implementation.
M oreover recycling and reducing policy should ensure that the amount of unsorted waste decreases
over the years, making difficult to operate the plant at design treatment capacity.
The present study considered the application of DM S technology in a particular Italian scenario
requiring landfill biogas exploitation.
The results lead to say:
DM S technology reaches high value of overall net efficiency (all solutions exceed 18%),
considering a small plant size (60,000 t MSW /y);
coupling DM S with external SH or EVA and ICE can maximize revenue from electric energy
production incentives even in CHP configuration.
M ore generally DM S technology energy recovery performances result aligned with other non
conventional waste to energy technologies, while at the same time DM S features the peculiar
advantage of minimizing the volume of material landfilled thanks to the slag and metal reuse. This
fact represents a significant improvement in economical and in environmental terms. If the site
chosen for construction offers a biogas flow rate, DM S technology can improve its performances by
external components and if coupled with an ICE it can guarantee a very flexible electrical power vs.
thermal power ratio in CHP configuration.
In particular, concerning environmental performances, the basic configuration examined in present
work generates emissions into atmosphere that are by far lower than regulatory limits and analogous
to those of modern conventional incinerators equipped with efficient flue gas cleaning systems
according to BREF, both in terms of pollutants concentration and volume of flue gases [12].
At this stage of the study, as regards both the external evaporator configuration and the external
superheater configuration, the pollutants incremental contribution to the DM S exhaust flow gases
was not deemed significant, considering both biogas combustion gases small volume and low
2
According to Law Decree D.P.R. N. 412 (1993) the Italian territory is divided into 6 climatic zones with the purpos e
of referen ce for the estimation of space heating systems average en ergy consumption. These climatic zones are named
with letters from A to F in function of number of degree-days. Zone A includes the municipalities featuring a number
of degree-days up to 600 whilst Zone F includes the municipalities featuring a number of degree-days greater than
3,000.
155
temperature due to cooling effect provided by high combustion air excess. In addition, these two
conditions of combustion and heat exchange seem to have no more critical elements than ICE
combustion that is deemed fully environmental acceptable in similar applications.
Finally, concerning the investment, previous studies have shown how, although the insertion of
components such as external SH and EVA in addition to DM S baseline configuration or CHP
option increase the initial investment amount, the Net Present Value and the Internal Rate of Return
of the project are improved by the increase of revenue items [21].
Nevertheless economically best solution, among the ones discussed in this preliminary study,
depends strongly on specific scenario and in particular on available biogas flow amount, heat
demand and site characteristics.
Nomenclature
BREF
CHP
CV
DMS
d.b.
ECO
EGR
ESP
EVA
EVA1
EVA2
GSE
h
hL
IAFR
ICE
LCV
mi
mr
ms
MSW
mv
SCR
SH1
SH2
Tamb
Tcond
Tfb in
Tfb out
Tsat
TSH1
Reference Document on the Best Available Techniques for Waste Incineration (European
Commission Integrated Pollution Prevention and Control)
Combined Heat and Power
Certificato verde (Green Certificate)
Direct M elting System
dry basis
Internal economizer
Exhaust gas recirculation
Electrostatic Precipitator
Internal evaporator
Internal evaporator
External evaporator
Gestore Servizi Energetici, Italian state owned energy service company
enthalpy, kJ/kg
enthalpy in liquid phase conditions, kJ/kg
Impianto Alimentato da Fonti Rinnovabili (Plant Powered by Renewable Sources)
Internal Combustion Engine
Lower Calorific Value
incondensable mass flow rate, equal to mr, kg/s
reintegration mass flow rate, 2% of mv, kg/s
feed water mass flow rate bled, kg/s
M unicipal Solid Waste
mass flow rate at turbine inlet, kg/s
Selective Catalytic Reactor
Internal superheater
External superheater
ambient temperature, C
steam condensing temperature, C
biogas flue gases temperature at external superheater inlet, C
biogas flue gases temperature at external superheater outlet, C
steam saturation temperature, C
superheated steam temperature at boiler outlet, C
156
TSH2
Tx
Ty
w.b.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
157
[16] Berlo M . V., Simoes P, High efficiency waste-to-energy: Amsterdams experiences after 4
years of operation, International waste management and landfill symposium, Sardinia, Italy;
2011.
[17] Genon G., Tedesco V., Urso P., Verifica della fattibilit di un impianto di trattamento termico
dei rifiuti a tecnologia innovativa nella Provincia di Torino. M ilan: CIPA Editore, 2010 (in
Italian).
[18] Arena U., A critical assessment of municipal solid waste gasification, International waste
management and landfill symposium, Sardinia, Italy; 2011.
[19] Consonni S., Giugliano M ., Grosso M ., Alternative strategies for energy recovery from
municipal solid waste Part A: M ass and energy balances. Waste Management 2005; 25: 123135.
[20] Pini Prato A., Strobino F., Broccardo M ., Parodi Giusino L., Integrated management of
cogeneration plants and district heating networks. Applied Energy 2012, article in press doi:
10.1016/j.apenergy.2012.02.038.
[21] Damonte M ., Ventura D., Ottimizzazione del recupero energetico da RSU mediante
tecnologia DM S. M aster thesis. Universit di Genova, Italy, M arch 2011 (in Italian).
158
Centre of Engineering, Modelling and Social Sciences, Federal University of ABC (CECS/UFABC),
Santo Andr, SP, Brazil, reynaldo.palacios@ufabc.edu.br
b
Centre of Engineering, Modelling and Social Sciences, Federal University of ABC (CECS/UFABC),
Santo Andr, SP, Brazil, adriano.ensinas@ufabc.edu.br
c
Centre of Engineering, Modelling and Social Sciences, Federal University of ABC (CECS/UFABC),
Santo Andr, SP, Brazil, marcelo.modesto@ufabc.edu.br
d
Centre of Engineering, Modelling and Social Sciences, Federal University of ABC (CECS/UFABC),
Santo Andr, SP, Brazil, and Interdisciplinary Centre of Energy Planning, University of Campinas
(NIPE/UNICAMP), Campinas, SP, Brazil, silvia.nebra@pq.cnpq.br
Abstract:
The aim of this study is to make an evaluation of the possibilities of ethanol production increas e through the
introduction of bagasse hydrolysis process in conventional distilleries, considering the limiting situation of
bagasse use: it is the major by-product in sugar and ethanol production and is burnt in boilers t o satisfy the
steam and power requirements of the process. Simulations in ASPEN PLUS soft ware were performed, in
order to evaluate the mass and energy balances, for the integrated process, considering the pre-treatment of
sugarcane bagasse by steam explosion. The cogeneration system was also modelled and integrated with
the ethanol production process. It consists of a steam cycle with backpressure steam turbines and
parameters of live steam of 67 bar and 480C. In all the cases studied it was considered that the steam flow
used in the system was just that necessary to fulfil the process thermal needs, so, it was assumed that the
surplus of bagasse was used to produce ethanol. The use of sugarcane trash was considered in order to
accomplish the energetic needs of the overall process as well as lignin cake, which is a hydrolysis process
residue. Several cases were evaluat ed, which include: the conventional ethanol production plant without
hydrolysis (Case I), the conventional plant joint with hydrolysis process without thermal integration
considering different solid contents in the hydrolysis reactor (Cases II, III and IV ), and the conventional plant
joint with the hydrolysis process considering t hermal int egration through Pinch method (Case V). The res ults
shown a modest ethanol production increase of 9.7% for the situation without thermal integration and low
solid content in the hydrolysis reactor, on the other hand, the c ase where t hermal integration was applied
presented an ethanol production inc rease of 17.4%.
Keywords:
Ethanol, sugarcane, enzymatic hydrolysis, thermal int egration.
1. Introduction
Ethanol is produced in Brazil in large scale using sugarcane as raw material by fermentation of
sugars and distillation. World consumption of ethanol tends to grow in the next years because of the
growing interest of many countries by biofuels use, due to factors such as: environmental damage
(avoided emissions of greenhouse gases) energy security (diversification of energy sources and
reducing dependence on oil) and support to farmers [1]. The sugarcane bagasse is the major byproduct in sugar and ethanol production and it is burnt in boilers to satisfy the steam and power
requirements of the process. Sugarcane bagasse, as well as other lignocellulosic materials, can be
also used for ethanol production but, the introduction of the bagasse hydrolysis process in the
current ethanol production system is a real challenge, being bagasse the fuel of the current process
and at the same time, raw material for the new one. Thus, the aim of this study is to make an
evaluation of the possibilities of ethanol production increase through the bagasse hydrolysis
159
process, considering the limiting situation of bagasse use. Simulations in ASPEN PLUS software
were performed, in order to evaluate the mass and energy balances and Pinch Analysis was used to
determine the minimum hot and cold utilities required by the integrated process in order to increase
the ethanol production. The characteristics of the cogeneration system were adopted considering
devices currently used in the new industrial plants, and prioritising the increase in ethanol
production, and not electricity cogeneration.
% Mass
13.85
13.15
0.59
0.20
1.79
69.35
1.07
In sugarcane factory the process begins with the cleaning operation: the sugarcane that arrives to the
factory contains some amount of soil that is carried in the harvesting operation; so, it must be
cleaned upon reception. Dry cleaning was considered in the simulation. After that, cleaned
sugarcane goes to the extraction system where sugarcane juice and bagasse are obtained; an
extraction system with mills was considered. From specifications in the extraction system, the
composition of bagasse obtained was (dry basis, wt.): cellulose 36.8% wt., hemicelluloses 35% wt.,
lignin 20.3% wt. and ashes 2.3%. The moisture content of bagasse (wet basis, wt.) was 50%.
Raw sugarcane juice goes to the physical-chemical treatment while bagasse goes to the
cogeneration system. For juice treatment, the following operations were considered: screening,
heating, liming, decantation and mud filtration. After that, the clarified juice goes to the
concentration stage.
In conventional autonomous distilleries, the must for ethanol production is prepared from sugarcane
juice, its concentration is necessary in order to obtain an adequate must sugar concentration for the
fermentation process. Thus, a part of the clarified juice, with its original sucrose content of 11.8%,
160
Steam turbines have a bleed at 22 bar for mills direct drive turbines and at 6 bar for must
sterilization process and ethanol dehydration requirements. The boiler was modelled according to
previous studies [10] and [11].
Table 2 shows the mean parameters adopted in the simulation of conventional ethanol production
process.
reactor, in order to remove impurities. This operation would reduce the amount of reactants in
subsequent stages. As this study is an initial assessment, previous washing of bagasse and trash was
not considered.
Thus, stream of bagasse B2 is sent to the pre-treatment reactor PRE-TRAT. Some studies indicate
that it is not necessary the addition of a catalyst in the steam explosion reactor because it is an
autocatalytic process: there is a dissociation of water molecules in their ions H+ and OH-, which
hydrolyse hemicelluloses. The formation of acetic acid happens also, which catalyses the
subsequent reactions. On the other hand, some researchers report that the addition of acid catalyst,
such as SO2 or H2 SO4, is necessary in order to achieve higher yields of hemicellulose conversion. In
this study, it was adopted the addition of SO 2 catalyst in the pre-treatment reactor in a rate of 2%
w/w, according to [14].
In relation to the steam consumption by steam explosion pre-treatment, experimental data are
reported in the range of 0.55 to 0.65 kg of steam /kg moist bagasse for moisture bagasse contents in
the range of 38.6 to 65.5% [15].
In the pre-treatment reactor, the formation of xylose (C5 H10 O5 ), acetic acid (C2 H4 O2 ), furfural and
glucose (C6 H12 O6 ) was considered, according to [5]. Table 3 shows the reactions and conversion
yields considered in pre-treatment reactor, according to [14].
Table 3. Yields considered for the reactions in pre-treatment reactor
Equation
Product
Yield (%)
C5 H8O4 + H2O
C5 H10O5
Xylose
61.4
C5 H8O4 + H2O
2.5 C2 H4 O2
Acetic acid
9.2
C5 H10 O5
FURFURAL + 3 H2 O
Furfural
5.1
C6 H10 O5 + H2 O C6 H12O6
Glucose
4.1
From
Hemicellulose
Hemicellulose
Xylose
Cellulose
The decompression in pre-treatment tank was represented by the expansion valve VE2 and the
unitary block FLASH-3. Hence, there are the steam flashed V-FL-PR and the pre-treated material
B-PRE1.
In order to remove xylose and other components that could inhibit the posterior processes of
enzymatic hydrolysis and fermentation, pre-treated bagasse is washed in unit block SEPA-PE. After
that, two fractions are present: the liquid fraction L-PE (xylose liquor) and the solid fraction CELLIG. In this first study, the use of xylose liquor was not considered.
For the next stage, water is added to the process (stream H2O-HIDR) in tank T-MIST, in order to
achieve an appropriate concentration of water insoluble solids in the hydrolysis reactor. Three cases
were considered in this study: solid contents of 5%, 8% and 10% in the hydrolysis reactor. In the
next step, stream CEL-LIG4 goes to the hydrolysis reactor R-HIDROL, where enzymes (stream
ENZIMA) are added to catalyze the hydrolysis reactions. After hydrolysis stage, the hydrolysate
goes to a filter in order to separate the lignin cake (LI-CAKE) of the glucose hydrolysate (LI-GLI).
Table 4 shows the yields adopted in the hydrolysis reactor calculated from [14] data
Table 4. Yields considered for the reactions in the hydrolysis reactor for different solid contents in
the reactor (5%, 8% and 10%)
Reaction
From
Product
Yield (%)
5%
8%
10%
C6 H10 O5 + H2 O C6 H12O6
Cellulose
Glucose
69.2
60.6
55.8
C5 H8O4 + H2O
C5 H10O5
Hemicellulose
Xylose
46.9
44.4
40.6
163
164
Table 5. Parameters adopted for the simulation of ethanol production through enzymatic
hydrolysis
Parameter
Value
a
Pre-treatment reactor temperature, C
190
Pre-treatment reactor pressure, bar
12.5
Pre-treatment reactor steam consumption, kg of steam/kg of raw materialb
0.55
Pressure at unitary block FLASH-3, bar
1.01
Efficiency of solid in solution removal in unit block SEPA-PE, %a
90
a
Loss of soluble lignin in unit block SEPA-PE, %
6.3
Moisture content of solid fraction CEL- LIG, %c
60
c
Water for xylose washing, l/kg of dry material
15
d
Hydrolysis reactor temperature, C
50
Enzymatic load cellulose, FPU/g dry biomassa
53
a
Enzymatic load b glucosidase, IU/g dry biomass
83
Moisture content in solid fraction TORTA-LI0, %
70
Solid content in concentrate hydrolysate, %
19
e
Moisture content in sugarcane trash, %
10.05
e
Trash lower heating value, MJ/kg
13.9
Energy consumption in trash shredder, kWh/t of trashe
82.03
Energy consumption in cleaner station, kWh/t of trashf
13.6
Energy consumption in bagasse feeder, kWh/t of bagasse
0.459
Energy consumption in xylose separator SEPA-PE, kWh/t of material
2.3
3
Energy consumption in separator SEPA-L-T, kWh/m
0.4
Energy consumption in dewatering press of lignin cake, kWh/kg of dry matter
56.09
a
Carrasco et al. (2010) [14]; b Kling et al. (1987) [15]; c Palacios-Bereche, (2011) [4];
Hassuani et al. (2005) [20]; f Cella (2010) [26]
Due to the possible presence of soluble lignin in glucose hydrolysate, as well as phenolic groups
(they were not considered in this simulation), the detoxification of glucose hydrolysate is
recommended before its mixture with must of sugarcane juice [18, 19]. In this simulation the
removal of the components acetic acid, furfural and sulphurous acid was considered, before the
mixture with must.
166
From the iterative procedure explained before, the amount of bagasse for hydrolysis can be
determined, moreover, thermal integration procedure permits to determine the targets of minimal
energy consumption, maximizing, in this way, the production of ethanol from bagasse hydrolysis.
Table 6 shows the streams considered for thermal integration, as well as the exchange heat and the
initial and final temperature. Figure 6 shows the final Grand Composite Curve for the analyzed
case.
Table 6. Streams considered for thermal integration, 10% of solid content in the hydrolysis reactor
(Case V)
Heat streams
Sterilized juice
Fermented wine
Phlegmasse
Vinasse
Anhydrous ethanol
Vapour Condensates
Condenser column B
Condenser extractive
column
Condenser column D
Vapour steam
explosion
Ti
C
130.0
32.0
103.8
109.3
78.3
83.2
Tf
C
32
28
35
35
35
35
H
MW
50.8
13.2
3.7
45.9
10.2
16.8
81.6
81.6
30.5
78.3
78.3
8.7
85.1
35
34.6
100.7
100
17.9
Cold streams
Treatment juice
Pre-heating juice
Juice for sterilization
Final wine
Reboiler column A
Reboiler column B
Reboiler extractive
column
Reboiler recovery
column
Hydrolysis water
Pre-heating glucose
liquor
Imbibition water
Ti
Tf
C
C
34.2 105.0
98.1 115.0
89
130.0
31.2 90.0
109.3 109.3
103.4 103.8
H
MW
44.9
2.8
22.5
41.4
53
35.4
112.1
137.1
8.0
149.6
149.6
2.9
25
50
9.7
50
115
27.2
25
50
4.7
Fig. 4. Grand Composite Curve including the evaporation systems GCC (Case V).
conventional distillery without hydrolysis, while cases II, III and IV correspond to the hydrolysis
process coupled with the conventional distillery without thermal integration. Case V corresponds to
the hydrolysis process integrated with the conventional process according to the method presented
in item 3. A higher solids concentration in the hydrolysis process means lower steam consumption
in the glucose liquor concentration step. This fact permits to assign higher amounts of bagasse for
the hydrolysis process and consequently, to obtain higher increases in the ethanol production.
About electricity surplus, there is an increase of 47.9%, 29.3%, 21.3% in Cases II, III and IV
respectively in comparison with Case I, these increases happen owing to an increase in steam
consumption in the mentioned cases. On the other hand, in Case V there is a reduction of 27.9% in
comparison with Case I, fact that is explained because in Case V, a reduction of steam consumption
happens. In general, these results are due to the turbine configuration adopted in the cogeneration
system (backpressure steam turbine) and the reduction of steam consumption for higher solid
contents in the hydrolysis step.
From the thermal integration procedure, it can be observed a significant decrease in steam
consumption in Case V. Steam consumption in Case V is 11.1% lower than the consumption in
Case IV, while ethanol production in Case V is 4.6% higher than the production obtained in Case
IV.
Table 7. Results of the simulation: Anhydrous ethanol production l/t of cane, electricity surplus
kWh/t of cane and steam consumption kg/t of cane.
Parameter
Solid content in hydrolysis reactor
Anhydrous ethanol, (l/t of cane)
Electricity surplus, (kWh/t of cane)
Bagasse for hydrolysis, (kg/t of cane)
Steam consumption, (kg/t of cane)
Increase in ethanol production, (%)
Increase in ethanol production, (l/t of cane)
Case I
-79
42.3
0
489.6
Case II
5
86.7
62.5
110.3
794.2
9.7
7.7
Case V
10
92.8
30.5
239.7
661.8
17.4
13.7
In relation to vinasse production, there is an increase of 11.6%, 14.1%, 15.1% and 20.8% for Cases
II, III, IV and V in comparison with Case I. It is due to the higher amount of must processed in
distillation stage in comparison with Case I. In Brazilian current practice, the vinasse is used to
irrigate the fields, where the nutrients are recuperated.
Figure 5 shows the bagasse balance for each case. Bagasse for hydrolysis increases from Case II to
V, while bagasse for boiler decreases. The amount of trash for boiler is constant for all cases, 38.9
t/h considering a moisture content of 10%. The amount of lignin cake also increases from Case II to
V, being a percentage of the bagasse for hydrolysis. Thus, the amount of lignin cake at 50% of
moisture content was 0.43 and 0.46 kg/kg of bagasse for hydrolysis for Cases II and III respectively
and 0.48 kg/kg of bagasse for Cases IV and V. The lower heating value of lignin cake was
calculated through the simulator from their composition data. The lower heating values obtained for
lignin cake were: 8563 kJ/kg for Case II, 8441 kJ/kg for Case III and 8373 kJ/kg for Cases IV and
V. These differences occurred because of the conversion yields adopted in the hydrolysis reactor.
168
Fig. 5. Bagasse balance (kg/t of cane) for Cases I, II, III, IV and V.
Figure 6 shows the balance of electricity generated. A decrease in electricity surplus appear from
Cases II to V due to the reduction in electricity produced in the cogeneration system, as a
consequence of the decrease in the steam generated in boiler, and due to the consumption increase
in the hydrolysis process.
Fig. 6. Balance of electricity generated (kWh/t of cane) for Cases II, III, IV and V
The electricity consumed in the hydrolysis process was estimated from the data shown in Table 5. It
was identified a significant electricity consumption in the agitators of the hydrolysis reactors, due to
the large volume of the hydrolysis reactors. The reactor volume was calculated assuming reactors
CSRT (Continuous Stirred Tank Reactors) and residence time of 48h, in agreement with data of
Table 4. The total volume of reactors resulted in 20520 m3 , 17165 m3 , 15731 m3 and 21835 m3
respectively for the cases II, III, IV and V. The electricity consumption in the reactor agitators was
determined from a scale up procedure based on the data reported in [23].
Results of the present study can be compared with others of the literature. Walter and Ensinas [19]
indicated an increase in ethanol production of 25.6%, but in a future scenario and considering
glucose and xylose fermentation. Dias et al. [24] indicated ethanol production increases of 18.3%
and 20.3%, but considering an Organosolv hydrolysis process with diluted acid, while Leite [1]
indicated an ethanol production increase, through enzymatic hydrolysis process, of 12.2% and
25.6%, for scenarios corresponding to the years of 2015 and 2025 respectively.
169
Dias et al. [27] also accomplished a simulation study considering the ethanol production by
enzymatic hydrolysis and pre-treatment by steam explosion, and a reduction of 30.8% in steam
consumption was obtained from thermal integration. These authors indicate an ethanol production
in the integrated process in the range of 107.5 to 120.6 l/t of cane; these values represent an increase
of 15.9% and 30% in comparison to their base case.
5. Conclusions
An evaluation of the mass balance and energy consumption for the ethanol production process by
enzymatic hydrolysis was accomplished. Moreover, this study showed the potential ethanol
production increase due to the introduction of the bagasse enzymatic hydrolysis in the conventional
ethanol production process.
The results obtained for cases II, III an IV were modest because the study considers a conservative
conventional factory with average technology, thus, in order to increase the ethanol production the
optimization of the conventional process would be recommended, aiming an energy consumption
reduction. Hence, the ethanol production increase for cases II, III and IV was 9.7%, 11.5% and
12.3%, respectively. These results showed that a higher ethanol production is obtained for higher
solids concentrations in the hydrolysis process. They show that thermal energy consumption in the
hydrolysate concentration stage is very significant, thus, other separation technologies with lower
energy consumption should be studied. Reverse osmosis membranes in combination with
evaporation systems can be tested for this application.
Case V showed the highest ethanol production of 92.8 l/t of cane, which represents an increase of
17.4%; however, this is a prospective study and pilot plant/industrial data would be necessary in
order to adjust the modelling of the plant. The Pinch Analysis applied to this case showed to be a
useful tool to evaluate the thermal integration potential.
There is an intrinsic vinasse increase due to the higher amount of must being processed with the
introduction of the hydrolysis process; the highest vinasse production was 430.8 t/h for Case V
which represents an increase of 20.8% in comparison with Case I. Currently, vinasse is used as
fertirrigation in Brazilian sugarcane fields, however this practice is being regulated. Thus, the
introduction of technologies to reduce the vinasse production becomes necessary.
The energy balance in the cogeneration system shows that the energy supplied by lignin cake is
significant to operate the integrated process. For Case V, energy supplied by lignin cake represented
44.2% of total energy supplied to the cogeneration system. Hence, the study and characterization of
lignin cake is important in order to enable the ethanol production by enzymatic hydrolysis process.
Acknowledgments
The authors wish to thank to CNPq (Process 304820/2009-1), FAPESP (Process 11/05718-1) and
FINEP (Contract FINEP FUNCAMP Nr. 01/06/004700) for the financial support.
References
[1] Leite R.C.C., Fuel bioethanol an opportunity to Brazil. Brasilia: Centro de Gesto e Estudos
Estratgicos; 2009, (In Portuguese).
[2] Rein P., Cane Sugar Engineering. Berlin: Verlag Dr. Albert Bartens K. G.; 2007.
[3] Wooley R.J., Putsche V., Development of an ASPEN PLUS Physical Property Database for
Biofuels Components, National Renewable Energy Laboratory, 1996. Available at
<www.p2pays.org/ref/22/21210.pdf> [accessed 12.11. 2007].
170
[4] Palacios-Bereche R., Modeling and energy integration of the ethanol production process fro m
sugarcane biomass [PhD Thesis]. Campinas, Brazil: University of Campinas; 2011 (I n
Portuguese).
[5] Dias M.O.S., Simulation of ethanol production processes from sugar and sugarcane bagasse,
aiming process integration and maximization of energy and bagasse surplus. [dissertation].
Campinas, Brazil: University of Campinas; 2008 (In Portuguese).
[6] Eijsberg R., The design and economic analysis of a modern bioethanol factory located in Brazil.
[dissertation]. Delft, The Netherlands: University Delft; 2006.
[7] Rossell C.E.V., Sugarcane processing to ethanol for fuel purposes, In Chemistry and Processing
of Sugarbeet and Sugarcane, edited by Clarke M. A. and Godshall M.A., Elsevier Science
Publishers B. V., Amsterdam, 1988.
[8] Finguerut, et al., Fermentation, Hydrolysis and Distillation. In: Cortes et al. editors. Biomass for
Energy. Campinas, Brazil: Ed. Unicamp. 2008. p. 436-474.
[9] Dias M.O.S., Modesto M., Ensinas A.V., Nebra S.A., Maciel Filho R., Rossell C.E.V.,
Improving bioethanol production from sugarcane: evaluation of distillation, thermal integratio n
and cogeneration systems. Energy 2010; 36: 3691-3703.
[10] Palacios-Bereche R., Nebra S.A. Thermodynamic modeling of a cogeneration system for a
sugarcane mill using ASPEN PLUS, difficulties and challenges. COBEM 2009: Proceedings o f
the 20th International Congress of Mechanical Engineering, 2009 Nov 15-20, Gramado, Brazil.
[11] Magnusson H., Process simulation in Aspen Plus of an integrated ethanol and CHP plant.
[dissertation]. Sweden: Umea University. 2005.
[12] Sun Y., Cheng J., Hydrolysis of lignocellulosic materials for ethanol production: a review.
Bioresource Technology 2002; 83: 111.
[13] Efe C., Technical and economical feasibility of production of ethanol from sugar cane and
sugarcane bagasse. [dissertation], Delft, The Netherlands: TU-Delft; 2005.
[14] Carrasco C., Baudel H.M., Sendelius J., Modig T., Roslander C., Galbe M., Hahn-Hgerdal
B., Zacchi G., Lidn G. SO2 -catalyzed steam pretreatment and fermentation of enzymatically
hydrolyzed sugarcane bagasse. Enzyme and Microbial Technology 2010; 46: 6473.
[15] Kling S.H., Carvalho Neto C., Ferrara M.A., Torres J.C.R., Magalhaes D.B., Ryu D.D.Y.,
Enhancements of enzymatic hydrolysis of sugar cane bagasse by steam explosion pretreatment.
Biotechnology and Bioengineering 1987; 29: 1035 1039.
[16] Sanchez O.J., Cardona C.A., Trends in biotechnological production of fuel ethanol from
different feedstocks. Bioresource Technology 2008; 99: 52705295.
[17] Jorge L.M.M, Righetto A.R., Polli P.A., Santos O.A.A., Maciel Filho R., Simulation and
analysis of a sugarcane juice evaporation system. Journal of Food Engineering 2010; 99: 351
359.
[18] Cardona C.A., Sanchez O.J., Energy consumption analysis of integrated flowsheets for
production of fuel ethanol from lignocellulosic biomass. Energy 2006; 31 (13): 2447-2559.
[19] Walter A., Ensinas A.V., Combined production of second generation biofuels and
electricity from sugar-cane residues. Energy 2010; 35: 874-879.
[20] Hassuani S.J., Leal M.R.L.V., Macedo I.C., Biomass power generation: Sugarcane bagasse
and trash, Piracicaba, Brazil: Ed. PNUD and CTC, 2005.
[21] Michelazzo M.B., Sensitivity analysis of six systems for collection of sugarcane trash
(Saccharum spp.). [dissertation]. Campinas, Brazil: University of Campinas, 2005 (I n
Portuguese).
[22] Westphalen D.L., Wolf Maciel M.R., Pinch Analysis of evaporation systems, Brazilian
Journal of Chemical Engineering 2000; 17: 4-7.
171
[23] Pereira L.T.C., Teixeira R.S.S.T., Bom E.P.S., Freitas S.P. Sugarcane bagasse enzymatic
hydrolysis: rheological data as criteria for impeller selection. Journal of Industrial Microbiology
& Biotechnology 2010; DOI 10.1007/s10295-010-0857-8.
[24] Dias M.O.S., Ensinas A.V., Nebra S.A., Maciel Filho R., Rossell C.E.V., Production of
bioethanol and other bio-based materials from sugarcane bagasse: Integration to conventiona l
bioethanol production process. Chemical Engineering Research and Design 2009, 87: 12061216.
[25] Galbe, M., Zacchi, G. Ethanol production from lignocellulosic materials. In: CORTEZ,
L.A.B. (Coord.). Bioethanol from sugarcane: R&D for productivity and sustainability. Sao
Paulo: Blucher, 2010. Part. 4, Chap. 12, p. 697-716.
[26] Cella, N. Trash use as fuel in biomass boilers. Course Boilers, Environment and Renewable
Energy. Ribeirao Preto, 23-24 de june 2010.
[27] Dias, M.O.S.; Cunha, M.P., Maciel Filho, R., Bonomi, A., Jesus, C.D. F., Rossell, C.E.V.
Simulation of integrated first and second generation bioethanol production from sugarcane :
comparison between different biomass pre-treatment methods. Journal of Industria l
Microbiology & Biotechnology, 2011, 38: 955-966.
172
Dipartimento di Energetica Sergio Stecco - Universit degli Studi di Firenze (CA), Florence, Italy,
isabella.pecorini@unifi.it
b
Dipartimento di Energetica Sergio Stecco - Universit degli Studi di Firenze, Florence, Italy,
tommaso.olivieri@unifi.it
c
Dipartimento di Energetica Sergio Stecco - Universit degli Studi di Firenze, Florence, Italy,
donata.bacchi@unifi.it
d
Dipartimento di Energetica Sergio Stecco - Universit degli Studi di Firenze, Florence, Italy,
alessandro.paradisi@unifi.it
e
Dipartimento di Energetica Sergio Stecco - Universit degli Studi di Firenze, Florence, Italy,
lidia.lombardi@unifi.it
f
Dipartimento di Ingegneria dellInformazione - Universit degli Studi di Siena, Florence, Italy,
corti@dii.unisi.it
g
Dipartimento di Energetica Sergio Stecco - Universit degli Studi di Firenze, Florence, Italy,
ennio.carnevale@unifi.it
Abstract:
The Biochemical Methane Potential (BMP) of several components of the Organic Fraction of the Municipal
Solid Waste (OFMSW) were tested in order to assess the possibility to obtain a good estimate of the biogas
production of a real scale anaerobic digestion plant. In particular, five different fractions and a mixed food
waste sample were t ested with batch anaerobic digesters at 37C and both the BMP after 21 days (BMP21)
and final BMP (BMP f) were measured. Regarding the mixed food waste substrate it was found an average
BMP21 of about 405 NL/kgVS and a B MPf of 484 NL/kgVS with an average methane content of 57%. From
the experimental results, some industrial potential biogas production were defined to compare them with
data from real anaerobic digestion plants. In particular two different plants were considered: one located in a
rural area that treats the source selected OFMSW from a public collection point, another locat ed in a city
area with a curbside collection system. Furthermore, studying the BMP of the pre-treatment reject of these
plants, it was possible to study the pre-treatment efficiency and the difference performance of the two real
plants.
Keywords:
Anaerobic Digestion, Biochemical Methane Potential, Biogas, Organic Fraction Municipal Solid Waste.
1. Introduction
In 2009, more than 10 million tons of waste, corresponding at the 33,6% of the whole amount of
municipal solid waste (MSW) produced in Italy, were collected as source separated fractions. About
the 35% of these fractions were organic fraction from kitchen and yard and garden waste and, since
2005, a constant increasing of 11% every year have been recorded. Moreover ISPRA [1] shows that
most of them were treated in composting plants (about 281 facilities were registered in the 2009)
and about 540'000 t where instead stabilized in anaerobic digestion plants, in particular 18 plants of
which 15 working.
173
New strategies in MSW management, i.e., source-separate collection of the OFMSW and the need
to reduce the biodegradable-MSW allocated in landfill, have favoured the development of
composting and anaerobic digestion as useful biotechnologies for transforming organic waste into
suitable agricultural products [2]. Moreover, given that the amount of OFMSW is still increasing
and the attention to the environmental impacts is becoming all the time more important, the
possibility to recover not only compost form waste but also energy could enhance the anaerobic
digestion of OFMSW as way to provide a clean fuel from renewable energy [3]. In this way the
quality of the OFMSW in terms of potential methane production becomes important in order to
assess the biogas production expectation from the anaerobic treatment.
In the last years, several researches have been carried out for the analysis of biomethane potential of
several waste substrates. In particular, most of these utilize BMP analysis as a possible way to
characterize the biodegradability of the organic matter in order to assess its stability and the waste
treatments efficiencies such as composting or anaerobic digestion. Others are instead interested in
determining waste BMP as relevant in the context of treatment by anaerobic digestion and useful to
determine the amount of organic carbon that can be anaerobic converted to methane. This research
focus on this last purpose with the aim to understand how, by measuring the BMP of the main
component of the OFMSW is possible to estimate the potential biogas production of real digestion
plant. Moreover, to obtain more realistic results, this research focuses on how the pre-treatments
and the operating environment could affect the biogas production of real plants.
TS, %FM
33,10,24
94,50,04
14,00,05
6,50,61
40,30,02
VS, %TS
89,90,01
97,50,01
98,10,06
81,81,48
90,40,04
The source selected OFMSW from curbside collection was used for the substrates tested in the
batch assay. For each fractions water content and VS were measured in triplicate, in Table 1 only
the mean value are shown.
Furthermore, a mixing of those fractions (Mixed Food Waste MFW) was employed for the tests
too; its composition is shown in Table 2. TS and VS contents, as the sum of each fractions, were
about 42,40,38% (w/w) and 93,30,01% on TS basis respectively.
Table 2. Mixed food waste composition
Fraction
Fruit and vegetable
Proteins
Carbohydrates
Leafs
Cellulose
Weight (g)
8
8
8
3
3
Percentage %(w/w)
26,7
26,7
26,7
10
10
Pressure was measured using a membrane pressure gauge (Model HD2304.0, Delta Ohm S.r.L.,
Italy). The values of pressure measured were converted into biogas volume as:
Pmeasured TNTP
Vbiogas
Vr
PNTP Tr
where:
Vbiogas, volume of daily biogas production, expressed in Normal litre (NL);
Pmeasured, headspace pressure before the gas sampling (atm);
Tr and Vr, temperature (K) and volume (L) of the reactor;
TNTP and PNTP , Normal temperature and pressure, 273,15K and 1 atm respectively.
The headspace volume, calculated as the difference between the total volume of the batch and the
volume occupied by the sample considering a sample density of 1g/mL, was about 600ml for each
bottle.
The gas produced was routinely analyzed using an IR gas analyzer (ECOPROBE 5 - RS Dynamics).
After every measurement the bottles were shaken to guarantee homogeneous conditions in the assay
vessels [5].
Experimental ID
MFW (a)
MFW (b)
Proteins
Carbohydrates
Fruit and Veget
Yard and Garden waste
Dirty paper
Inoculum
NL/kgTS
413
396
386
91
250
115
315
47
NL/kgVS
542
520
528
109
353
175
422
74
Error, %
14
14
14
67
21
42
17
-
Preliminary experiments on similar waste showed that 90 days of incubation at 35C, after the lag
period, was sufficient to insure the total gas production expression [8], therefore the values of BMP
measured after 93 days has been considered as the final biogas produced by each substrate. The
results obtained, deducted the inoculum yield, seem to be comparable to that obtained in other
similar studies [9] and [10].
Table 4. BMPf
Experimental ID
MFW (a)
MFW (b)
Proteins
Carbohydrates
Fruit and Veget
Yard and Garden waste
Dirty paper
Inoculum
NL/kgTS
501
466
452
101
327
205
377
130
NL/kgVS
657
611
617
120
462
312
504
203
Error, %
31
33
33
168
44
65
40
-
The highest BMP value was obtained for the MFW, while, observing the calculated errors, it is
clear that for charbohydrates some problem of acidification occurred.
This can be also noticed in the temporal plot of the average cumulative biochemical methane yields
(Fig. 3): for the soon after few days, no significant methane production was detected.
With reference to Fig. 4, the correlation between the BMP21 and the BMPf has been studied [4]. As
results from the chart, the biogas obtained at 21 days corresponds to the 89% of ultimate potential
methane. In fact, as supported by the comparison between the cumulative and the daily MFW
average BMP, the main part of the biogas totally produced was released during the first 20 days
(Fig. 5).
Focusing on the MFW results, the peak value of daily biogas production was about 51 NL/d*kg VS,
while the cumulative biogas were 542 and 520 NL/kgVS for the MFW (a) and (b) sample.
Also for the other samples, the biogas composition was analysed in order to understand if the
methanogenesis phase was correctly taking place during the digestion process.
From the biogas analysis it has also been possible to esteem the quantity of methane totally
produced, simply as the product between the biogas released and the methane percentage of its
composition (Fig. 7). Considering an average methane percentage of about the 57%, it was
esteemed a methane yield of 309 and 296 NLCH4 /kgVS for the sample MFW (a) and MFW (b)
respectively.
This is probably due to an accumulation of volatile fatty acid and acid conditions. In fact, in the
preparation of the carbohydrates batches, too high concentrations of substrate were probably used,
as well as the substrate to inoculums ratio was higher than 1,2, value suggested by the authors to
avoid acidification[7] given that this ratio is recognised as one of the major parameter affecting the
results of anaerobic assay[11].
180
As Fig. 5, the graphic in Fig. 9 shows that, comparing with the background methane production
from the inoculum determined in blank assays, the biogas production after 40th day of digestion can
be neglect and the most of the biogas was produced during the first 20th day of digestion.
Furthermore, looking at the daily biogas pruduction,it can be noticed that, according with Zhu B.,
Gikas P. and Zhang R. et al. [7], the biogas production duration of food waste was prolunged, with
initial daily biogas yields lower compare to others.
BMPf, Nm3 /t
213
198
150
95
46
13
152
5
Adding up the BMP21 of each waste component and considering a typicall source selected
OFMSW composition shown in Fig. 11, three potential industrial biogas production are define
(Table 6):
181
a maximum biogas production (GPmax), calculated assuming that the pre-treatment before the
anaerobic digestion is able to remove only the undesired waste fractions with an efficiency of the
100%, i.e. wood packaging, plastic film and plastic packaging, other plastic, rubber, leather,
ferrous and non ferrous metals, inert and hazardous waste;
a potential production, in which is consider the 80% of pre-treatment efficiency and the 20% of
biodegradable fraction removed wrongly (GP80%);
another potential production defined as above but assuming an 75% removal efficiency and the
25% of organic fractions separate erroneously (GP75%).
These estimates could be compared with the biogas production of real anaerobic digestion plants
(GPreal). In particular, two different anaerobic digestion plants are considered: one located in a
rural area treating source selected OFMSW from a public collection point with a GP of 98 Nm3 /t,
another located in a city area with a curbside collection system characterized by a higher GP (Table
7). It is possible to notice that the estimated biogas production (GP80% and GP75%) and the GPreal
have similar values and that the rural area digester is characterized by the lowest biogas production.
Table 7: Biogas production, real industrial anaerobic digestion plant
Anaerobic digestion
Collection system
GPreal, Nm 3 /t
plant
Rural Area
Public collection point
98
City Area
Curbside collection
108
Data Source
Management data
Bozano Gandolfi P., [12]
As said before, during the experimental assay, it has been possible to measure also the BMP of the
pre-treatment rejects. In particular the light fraction (LF) and the small heavy fraction (SHF) from
the pre-treatment of the two anaerobic digesters were tested (Table 8).
182
BMP21, NL/kgVS
240
261
508
214
BMP21, Nm3 /t
57
8
99
33
As before, also the BMP21 of both LF and SHF form the rural area digester were lower than the
ones from the city area. This is probably due to an high presence of yard and garden waste in the
rural waste composition, given that the yard and garden waste has a low BMP21 (with reference to
Table 5, 7 Nm3 /t). On the other side, the rejects from the city area digestion plat are characterized
by a high BMP21, probably because a considerable part of biodegradable fraction of input waste,
with the highest BMP21, is wrongly removed by the pre-treatment.
5. Conclusions
The results of this study show that the biochemical methane potential assay could provide useful
data to study the pre-treatment efficiency and the performance of real anaerobic digestion plant.
Focusing on the experimental work, the MFW had the highest biochemical methane potential, with
an average BMP21 of 405 NL/kgTS and a BMPf of about 484 NL/kgTS and an average methane
content of 57%. The laboratory equipments developed prove to be suitable to this kind of
experimentation, given that no airtight problem occurred. However, the results obtained for the
carbohydrates sample and the errors behaviour show that the measurement protocol and the sample
preparation have to be improved in some parts. In particular more attention and specific evaluation
have to be done for the substrate to inoculum ratio as well as for the inoculum characteristics.
Actually, the estimates obtained with the experimental assay were comparable with the biogas
production of real anaerobic digestion plant and it has been possible to assess the pre-treatment
efficiency and the performance of some real cases. From the comparison results that: the GP of the
rural plant is affected by the presence of yard and garden waste, as it is also supported by the BMP
of the pre-treatment rejects; the pre-treatment reject of the city area plant, that treats OFMSW
collected by a curbside system, has a high BMP probably because a considerable biodegradable
fraction is removed by the pre-treatment.
Nomenclature
BMP Biochemical Methane Potential, NL/kgTS or NL/kgVS
BMP Biochemical Methane Potential after 21 days, NL/kgTS or NL/kgVS
BMPf Final Biochemical Methane Potential, NL/kgTS or NL/kgVS
GP80%
Industrial biogas production, considering 80% of pre-treatment efficiency, Nm3 /t
GP75%
Industrial biogas production, considering 75% of pre-treatment efficiency, Nm3 /t
GPmax
Industrial biogas production, considering maximum pre-treatment efficiency, Nm3 /t
GPreal Industrial biogas production, considering existing real plants, Nm3 /t
LF
Light Fraction, anaerobic digestion plant rejects
MFW Mixed Food Waste
MSW Municipal Solid Waste
NL
Normal litre
NTP Normal Temperature and Pressure conditions, 273,15K and 1 atm respectively
183
OFMSW
Organic Fraction from Municipal Solid Waste
SHF Small Heavy Fraction, anaerobic digestion plant rejects
TS
Total Solid, % v/v on wet weight basis
VS
Volatile Total solid, %v/v on TS basis
References
[1] Istituto Superiore per la Protezione e la Ricerca Ambientale. Rapporto Rifiuti Urbani Edizione
2009
Available
at:
<
http://www.apat.gov.it/site/itit/apat/pubblicazioni/rapporti/documento/rapporto_108_2010_rifiuti.html>.
[accessed
17.1.2012]
[2] Pognani M., BarrenaR., Font X., Scaglia B:, Adani F., and Snchez A., Monitoring the organic
matter properties in a combined anaerobic/aerobic full-scale municipal source-separated waste
treatment plant. Bioresource Technol, 2010; 101:6873-6877
[3] Schievano A., Pognani M., DImporzano G., and Adani F., Predicting anaerobic
biogassification potential of ingestates and digestates of a full- scale biogas plant using chemica l
and biological parameters. Bioresource Technol, 2008; 99:8112-8117
[4] Ponsa S., Gea T., Alerm L., Cerezo J., Sanchez A., Comparison of aerobic and anaerobic
stability indices through a MSW biological treatment process. Waste Manage, 2008; 28:27352742
[5] Angelidaki I., Alves M., Bolzonella D., Borzacconi L., Campos J.L., Guwy A.J., Kalyuzhnyi
S., Jenicek P and van Lier J.B., Defining the biomethane potential (BMP) of solid organic
wastes and energy crops: a proposed protocol for batch assays. Water Sci Technol, 2009;
59:927-934
[6] Hansen L.T., Schmidt E.J., Angelidaki I., Marca E., Jansen J.L.C., Mosb k H. and Christense n
T.H., Method for determination of methane potentials of solid organic waste. Waste Manage,
2004; 24:393-400
[7] Zhu B., Gikas P., Zhang R., Lord J., Jenkins B. and Li X., Characteristics and productio n
potential of municipal solid waste pretreated with a rotary drum reactor. Bioresource Technol,
2009; 100:1122-1129
[8] Bayard R., de Arajo Morais J., Ducom G., Achour F., Rouez M. and Gourdon R., Assessment
of the effectiveness of an industrial unit of mechanical-biological treatment of municipal solid
waste, 2010; 175:23-32
[9] Lesteur M., Bellon-Maurel V., Gonzalez C., Latrille E., Roger J.M., Junqua G. and Steyer J.P.,
Alternative methods for determining anaerobic biodegradabilit: A review. Process Biochem,
2010; 45: 431-440
[10] Cho J.K. and Park S.C., Biochemical methane potential and solid state anaerobic digestion
of Korean food waste. Bioresource Technol, 1995; 52: 245-253
[11] Neves L., Oliveira R. and Alves M. M., Influence of inoculum activity on the biomethanization of a kitchen waste under different waste/inoculum ratios. Process Biochem,
2004, 39:2019-2024
[12] Bozano Gandolfi P., Pratical experiences in the production of biogas and energy from
wastes Available at: < http://bta-international.de/downloads.html?&lang=3>. [accessed
24.1.2012]
184
Abstract:
This paper employs exergy analysis to derive a general equation for the exergy efficiency of flat plate
collectors and thus optimize its design and operation. Exergy analysis of a flat-plate solar collector is a more
effective method of finding the optimum relationship between flow rate and the collector area. The fixed
parameters in this optimization are the collector inlet temperature, available solar radiation, the collector
transmittance-absorptance product and the ambient temperature. The collector heat loss coefficient is
estimated according to the collector plate temperature, wind convection loss, number of glass covers,
collector inclination, ambient temperature and emittance of collector plate and glass cover. In order to find
the optimum value of this multivariable problem genetic algorithms are used which are based on the
principles of genetics and survival of the fittest. The optimization parameter is exergy efficiency and the
objective is to maximize this parameter. Genetic algorithms proved suitable and very quick in obtaining the
required results. These results prove that the exergy efficiency of a flat-plate solar collector is maximized for
small distances between the riser tubes and for very small diameter of these tubes. By using a more
practical distance of 10 centimetres between the tubes, and excluding this parameter from the optimization
procedure, very small differences are observed in maximum exergy efficiency and if the cost of the
materials is accounted this is a more cost-effective solution. Other findings are that the exergy efficiency
increases considerably at higher solar radiation and that the transmittance absorptance product affects to a
great extent the exergy efficiency.
Keywords:
Exergy analysis, Genetic algorithms, Flat-plate collectors, Optimisation, Efficiency.
1. Introduction
Solar energy collectors are special kind of heat exchangers that transform solar radiation energy to
internal energy of a transport fluid. Flat plate collectors are the most popular type of solar devises
for low temperature applications. The main use of these collectors is in solar water heating systems
operating at maximum temperatures of 80-90C. M ost of these systems are operating
thermosiphonically at very small flow rate created by the small density differences between the hot
and cold water. A number of researchers have used exergy analysis to design flat plate collectors.
Badescu [1] optimized the width and thickness of the fins of a flat-plate collector by minimizing the
cost per unit useful heat flux. The proposed procedure allows computation of the necessary
collection surface area. A rather involved, but still simple, flat plate solar collector model is used in
the calculations. M odel implementation requires a specific geographical location with a detailed
meteorological database available. Fins of both uniform and variable thickness were considered.
The optimum fin cross section is very close to an isosceles triangle. The fin width is shorter and the
seasonal influence is weaker at lower operation temperatures. Fin width and thickness at the base
depend on the season. The optimum distance between the tubes is increased by increasing the inlet
fluid temperature, and it is larger in the cold season than in the warm season.
Badescu [2] also considered the best operation strategies for open loop flat-plate solar collector
systems. A direct optimal control method (the TOM P algorithm) is implemented. A detailed
collector model and realistic meteorological data from both cold and warm seasons are used. The
maximum exergetic efficiency is low (usually less than 3%), in good agreement with experimental
185
measurements reported in literature. The optimum mass-flow rate increases near sunrise and sunset
and by increasing the fluid inlet temperature. The optimum mass-flow rate is well correlated with
global solar irradiance during the warm season. Also, operation at a properly defined constant massflow rate may be close to the optimal operation.
Torres-Reyes et al. [3] presented a procedure to establish the optimal performance parameters for
the minimum entropy generation during the collection of solar energy. The Entropy Generation
Number, Ns, and the criterion for the optimal thermodynamic operation of a collector under nonisothermally, finite-time conditions, are reviewed. The M ass Flow Number, M , corresponding to the
optimum flow of working fluid as a function of the solar collection area, is also considered. A
general method for the preliminary solar collector design, based on Ns, M and the SunAir or
stagnation temperature, is developed. This last concept is defined as the maximum temperature that
the collector reaches at non-flow conditions for a given geographic location, geometry and
construction materials. The thermodynamic optimization procedure was used to determine the
optimal performance parameters of an experimental solar collector.
Torres-Reyes et al. [4] also presented the thermodynamic optimization of flat plate collectors based
on the first and the second law, developed to determine the optimal performance parameters and to
design a solar to thermal energy conversion system. An exergy analysis is presented to determine
the optimum outlet temperature of the working fluid and the optimum path flow length of solar
collectors with various configurations. The collectors used to heat the air flow during solar-tothermal energy conversion are internally arranged in different ways with respect to the absorber
plates and heat transfer elements. The exergy balance and the dimensionless exergy relationships
are derived by taking into account the irreversibilities produced by the pressure drop in the flow of
the working fluid through the collector. Design formulas for different air duct and absorber plate
arrangements are obtained.
The use of genetic algorithms (GAs) for the optimal design of solar collectors is well known [5, 6].
Genetic algorithms have been used as a design support tool by Loomans and Visser [7] for the
optimization of large hot water systems. The tool calculates the yield and the costs of solar hot
water systems based on technical and financial data of the system components. The genetic
algorithm allows for the optimization of separate variables as the collector type, the number of
collectors, the heat storage capacity and the collector heat exchanger area.
Kalogirou [8] used also genetic algorithms together with a neural network for the optimization of
the design of solar energy systems. The method is presented by means of an example referring to an
industrial process heat system. The genetic algorithm is used to determine the optimum values of
collector area and the storage tank size of the system which minimize the solar energy price.
According to the author the solution reached is more accurate than the traditional trial and error
method and the design time is reduced substantially.
Krause et al. [9] presented a study in which two solar domestic hot water systems in Germany have
been optimized by employing validated TRNSYS models in combination with genetic algorithms.
Three different optimization procedures are presented. The first concerns the planning phase. The
second one concerns the operation of the systems and should be carried out after about one year of
data is collected. The third one examines the daily performance by considering predictions of
weather and hot water consumption and actual temperature level in the storage tank.
The objective of the present work is to maximize the exergy efficiency of flat-plate collectors. The
fixed parameters in this optimization are the collector inlet temperature, available solar radiation,
the collector transmittance-absorptance product and the ambient temperature. The collector heat
loss coefficient is estimated according to the collector plate temperature, wind convection loss,
number of glass covers, collector inclination, ambient temperature and emittance of collector plate
and glass cover. For this purpose an evolution strategy based on genetic algorithms is used to
determine the optimum solution.
2. Energy Analysis
186
In this section various relations that are required in order to determine the useful energy collected
and the interaction of the various constructional parameters on the performance of a collector are
presented.
The useful energy collected from a collector can be obtained from the following formula [10]:
Qu
Ac FR Gt (
) UL Tf,i
Ta
mc p Tf,o Tf,i
(1)
mc p
U F ' Ac
1 Exp L
A c UL
mc p
(2)
In (2) F is the collector efficiency factor which is calculated by considering the temperature
distribution between two pipes of the collector absorber and by assuming that the temperature
gradient in the flow direction is negligible [10]. This analysis can be performed by considering the
sheet-tube configuration shown in Fig. 1, where the distance between the tubes is W, the tube
diameter is D, and the sheet thickness is .
1
UL
F'
1
W
U L D (W
D) F
1
Cb
(3)
1
Di hfi
In (3), Cb is the bond conductance which can be estimated from knowledge of the bond thermal
conductivity, the average bond thickness, and the bond width. The bond conductance can be very
important in accurately describing the collector performance and generally it is necessary to have
good metal-to-metal contact so that the bond conductance is greater that 30 W/mK and preferably
the tube should be welded to the fin.
Factor F in (3) is the standard fin efficiency for straight fins with rectangular profile, obtained from:
tanh n(W D) / 2
n(W D) / 2
(4)
UL
k
(5)
The collector efficiency factor is essentially a constant factor for any collector design and fluid flow
rate. The ratio of U L to Cb, the ratio of UL to hfi, and the fin efficiency F are the only variables
appearing in (3) that may be functions of temperature. For most collector designs F is the most
important of these variables in determining F . The factor F is a function of U L and hfi, each of
which has some temperature dependence, but it is not a strong function of temperature.
Additionally, the collector efficiency factor decreases with increased tube center-to-center distances
and increases with increases in both material thicknesses and thermal conductivity. Increasing the
overall loss coefficient decreases F while increasing the fluid-tube heat transfer coefficient
increases F .
Therefore it is obvious from the above analysis that by increasing F more energy can be intercepted
by the collector. By keeping all other factors constant increase of F can be obtained by decreasing
W. However, decrease in W means increased number of tubes and therefore extra cost would be
required for the construction of the collector.
The collector efficiency is found by dividing Q u given in (1), by the incident radiation A c Gt. By
doing so the following Equation is obtained:
FR (
) FR U L
Tf ,i Ta
Gt
mc p Tf ,o Tf,i
(6)
AcG t
By plotting against
/Gt a straight line is obtained with the slope equal to F RUL, called the loss
coefficient and the intercept on the y-axis equal to F R ), called optical efficiency.
Generally the efficiency of a solar thermal system increases by increasing the flow rate. This
increase is asymptotic, i.e., the increase is rapid at small flow rates and becomes almost asymptotic
to a certain maximum value at higher values of flow rate. In a similar way the collector outlet
temperature increases with the collector area and decreases with increasing flow rate and vice versa.
This creates difficulties in selecting appropriate values of flow rate and collector area.
In a real system the variation of the fluid inlet temperature depends to a great extent on the storage
tank configuration, thermal load demand and the consequent make-up water to the storage tank
which affects the fluid inlet temperature to the collector.
3. Exergy Analysis
According to Kalogirou [10] the temperature at any position y at a fluid inlet temperature T f,i is
given by:
Tf
Ta
S
UL
Tf ,i
Ta
S
UL
Exp
U L F' NWy
mc p
(7)
For y=L the collector area is given by Ac = NWL. If T f,i = T a the fluid temperature variation at the
exit from the solar collector is:
Tf ,o Ta
S
UL
1 Exp
U L F' A c
cp m
(8)
The flow rate of exergy transferred from the sun to the fluid that is heated while crossing the riser
pipe is:
Ef
mef
m h f ,o
h f ,i
Ta s f ,o sf ,i
(9)
188
Where hf,o - hf,i = cp(T f,o - T f,i) is the variation of specific enthalpy. The variation of the specific
entropy is:
sf ,o s f ,i
c p ln
Tf ,o
(10)
Tf ,i
Therefore:
Ef
mcp
T Ta ln
Tf ,o
(11)
Tf ,i
And the exergy efficiency is given by dividing E f by the available solar radiation Qs,in which is
equal to Ac Gt:
ex
Ef
Q S,in
Replacing
mc p
ex
A cG t
T Ta ln
mc p
Tf ,o
Tf ,i
(12)
AcGt
given by (8) in (12), the following relation for the exergy efficiency can be obtained:
S
1 Exp
UL
U L F ' Ac
mcp
Ta ln 1
S
1 Exp
Ta U L
U L F ' Ac
mc p
(13)
As can be seen from the exergy analysis, a general equation for the exergy efficiency of flat plate
collectors is derived. This efficiency depends on the values of flow rate, collector area and collector
efficiency factor. The latter depends on the distance between consequent riser tubes, the diameter of
the riser tubes, the collector fin efficiency, the internal riser tube diameter (depends on outside tube
diameter) and the convection coefficient inside the tube - which depends on the mass flow rate and
tube inside diameter, which affect the Reynolds number and thus the Nusselt number.
4. Method Description
The objective of this work is to find the parameters that maximize the exergy efficiency. In order to
do this a numbers of parameters need to be considered as constants. These are shown in Table 1.
Table 1. Constant parameters used in exergy optimization
Parameter
Fluid specific heat, cp
Available solar radiation, G t
Transmittance-absorptance product, ( )
Ambient temperature, Ta
Bond conductance, Cb
Absorbing plate thickness,
Thermal conductivity of fluid, k
Value
4185 J/kg-K
500-1000 W/m2
0.60-0.90
25C
100 W/m2-K
0.5 mm
410 W/m-K
From these constant parameters a number of other parameters are evaluated. These are:
1. Heat loss coefficient, U L, estimated from [10]:
189
Tp
1
Ng
UL
C Tp Ta
Tp N g f
0.33
1
hw
1
0.05N g 1
Ta
p
Tp Ta
2N g f 1
(14)
Ng
Where:
f
1 0.04h w
0.0005h w2 1 0.091N g
C 365.9 1 0.00883
hw
0.0001298
(15)
(16)
8.6V 0.6
L0.4
(17)
This estimation ignores the bottom and edge losses and considers the wind velocity to be 1 m/s,
collector slope, , equal to 45, one glass cover and emittance values of 0.8. The estimation is also
done at a plate temperature T p to be 20C above the mean collector fluid temperature.
2. Factor n estimated from (5)
3. Fin efficiency F, estimated from (4)
4. The convection heat transfer coefficient inside the pipe, hf,i, estimated from the principles of
heat transfer and according to the type of flow, turbulent or laminar according to the mass flow
rate and the riser tube diameter.
5. The collector efficiency factor F , estimated from (3)
For the above parameters the inputs required are the distance between successive riser tubes, W, the
riser tube diameter, D, and the collector mass flow rate. These are varied during the optimization
process until the exergy efficiency is maximized using a genetic algorithm described in the
following section.
It should be noted that a proper exergy optimization should include optimization of both the
process parameters and the size of equipment by considering two key parameters, i.e., exergy
efficiency and cost, which include the capital investment (proportional to the collector area) and
operating cost due to the pressure drop. However, in this work the pressure drop is not included as
the collector is assumed to operate at very small flow rate (similar to the thermosyphonic one),
therefore, no exergy consumption for the fluid pumping is included in the analysis and this should
shift the optimum to the smaller diameters of tubes and distances between the tubes. This
assumption will be proved by the obtained results. Additionally, no other cost-connected parameter
is included in the optimization, such as exergy cost of equipment, so the optimization related to the
collector area only affects the amount of radiation collected by the system.
5. Genetic Algorithm
The genetic algorithm (GA) is a model of machine learning, which derives its behavior from a
representation of the processes of evolution in nature. This is done by the creation within a
machine/computer of a population of individuals represented by chromosomes. Essentially these are
a set of character strings that are analogous to the chromosomes that we see in the DNA of human
beings. The individuals in the population then go through a process of evolution.
It should be noted that evolution as occurring in nature or elsewhere is not a purposive or directed
process, i.e., there is no evidence to support the assertion that the goal of evolution is to produce
M ankind. Indeed, the processes of nature seem to end to different individuals competing for
resources in the environment. Some are better than others are, those that are better are more likely
to survive and propagate their genetic material.
190
In nature, the encoding for the genetic information is done in a way that admits asexual
reproduction, which typically results in offspring that are genetically identical to the parent. Sexual
reproduction allows the creation of genetically radically different offspring that are still of the same
general species.
In an over simplified consideration, at the molecular level what happens is that a pair of
chromosomes bump into one another, exchange chunks of genetic information and drift apart. This
is the recombination operation, which in GAs is generally referred to as crossover because of the
way that genetic material crosses over from one chromosome to another.
The crossover operation happens in an environment where the selection of who gets to mate is a
function of the fitness of the individual, i.e., how good the individual is at competing in its
environment. Some GAs use a simple function of the fitness measure to select individuals
(probabilistically) to undergo genetic operations such as crossover or asexual reproduction, i.e., the
propagation of genetic material remains unaltered. This is fitness - proportionate selection. Other
implementations use a model in which certain randomly selected individuals in a subgroup compete
and the fittest is selected. This is called tournament selection. The two processes that most
contribute to evolution are crossover and fitness based selection/reproduction. M utation also plays a
role in this process.
GAs are used for a number of different application areas. An example of this would be
multidimensional optimization problems in which the character string of the chromosome can be
used to encode the values for the different parameters being optimized.
In practice, this genetic model of computation can be implemented by having arrays of bits or
characters to represent the chromosomes. Simple bit manipulation operations allow the
implementation of crossover, mutation and other operations.
When the GA is executed, it is usually done in a manner that involves the following cycle [11]:
Evaluate the fitness of all of the individuals in the population.
Create a new population by performing operations such as crossover, fitness-proportionate
reproduction and mutation on the individuals whose fitness has just been measured.
Discard the old population and iterate using the new population.
One iteration of this loop is referred to as a generation. M ore details on genetic algorithms can be
found in Goldberg [12].
The first generation of this process operates on a population of randomly generated individuals.
From there on, the genetic operations, in concert with the fitness measure, operate to improve the
population. Genetic algorithms (GA) are suitable for finding the optimum solution in problems
where a fitness function is present. Genetic algorithms use a fitness measure to determine which
of the individuals in the population survive and reproduce. Thus, survival of the fittest causes good
solutions to progress. A genetic algorithm works by selective breeding of a population of
individuals, each of which could be a potential solution to the problem. The genetic algorithm is
seeking to breed an individual, which either maximizes, minimizes or it is focused on a particular
solution of a problem. In this case, the genetic algorithm is seeking to breed an individual that
maximizes the exergy efficiency of the solar collector.
The larger the breeding pool size, the greater the potential of it producing a better individual.
However, the fitness value produced by every individual must be compared with all other fitness
values of all the other individuals on every reproductive cycle, so larger breeding pools take longer
time. After testing all of the individuals in the pool, a new generation of individuals is produced
for testing.
A genetic algorithm is not gradient based, and uses an implicitly parallel sampling of the solutions
space. The population approach and multiple sampling means that it is less subject to becoming
trapped to local minima than traditional direct approaches, and can navigate a large solution space
with a highly efficient number of samples. Although not guaranteed to provide a globally optimum
191
solution, GAs have been shown to be highly efficient at reaching a very near optimum solution in a
computationally efficient manner.
During the setting up of the GA the user has to specify the adjustable chromosomes, i.e. the
parameters that would be modified during evolution to obtain the maximum or minimum values of
the fitness functions. In this work, the fitness function is exergy efficiency given by (13).
Additionally the user has to specify the range of the input parameters called constraints.
The genetic algorithm parameters used in the present work are:
Population size=50
Population size is the size of the genetic breeding pool, i.e., the number of individuals contained in
the pool. If this parameter is set to a low value, there would not be enough different kinds of
individuals to solve the problem satisfactorily. On the other hand, if there are too many in the
population, a good solution will take longer to be found because the fitness function must be
calculated for every individual in every generation.
Crossover rate=90%
Crossover rate determines the probability that the crossover operator will be applied to a particular
chromosome during a generation.
M utation rate=1%
M utation rate determines the probability that the mutation operator will be applied to a particular
chromosome during a generation.
Generation gap=96%
Generation gap determines the fraction of those individuals that do not go into the next generation.
It is sometimes desirable that individuals in the population be allowed to go into next generation.
This is especially important if individuals selected are the most fit ones in the population.
Chromosome type=continuous
Populations are composed of individuals, and individuals are composed of chromosomes, which are
equivalent to variables. Chromosomes are composed of smaller units called genes. There are two
types of chromosomes, continuous and enumerated. Continuous are implemented in the computer as
binary bits. The two distinct values of a gene, 0 and 1, are called alleles. M ultiple chromosomes
make up the individual. Each partition is one chromosome, each binary bit is a gene, and the value
of each bit (1, 0, 0, 1, 1, 0) is an allele. Enumerated chromosomes consist of genes, which can have
more allele values than just 0 and 1.
The genetic algorithm is usually stopped after best fitness remained unchanged for a number of
generations or when the optimum solution is reached. In this work the genetic algorithm was
stopped after best fitness remained unchanged for 75 generations.
6. Results
The input parameters (adjustable chromosomes) were used in a genetic algorithm program to find
the values that maximize collector exergy efficiency. The whole model was set up in a
spreadsheet program in which the various parameters are entered into different cells. The optimum
values of the various parameters were used in (13) to estimate the collector exergy efficiency which
is the fitness function that needs to be maximized. The various input parameters are constrained to
vary between certain values. The ones used in this work are shown in Table 2.
192
Symbol
m
Ac
W
D
Range
0.001-0.08 kg/s
1-10 m2
0.03-0.15 m
0.004-0.022 m
The optimum results can be presented graphically or in tables. Here the table presentation is
preferred so as to show the exact values obtained. The results are shown in Tables 3-5. It should be
noted that for each run of the program the optimum solution was reached in less than 5 seconds on a
Pentium 3.2 GHz machine, which is very fast.
Table 3 presents the results of the optimization process with respect to the effect of inlet
temperature and radiation.
Table 3. Results of the optimization process-effect of inlet temperature and radiation
Gt (W/m2)
( ) T in (C)
W (m) D (m)
T out (C)
Ac (m2)
ex (%)
25
0.03
0.004
3.07
66.02
7.49
27
0.03
0.004
4.85
59.86
9.98
500
0.8
30
0.03
0.004
7.90
52.24
9.61
35
0.03
0.013
13.77
43.27
8.76
m (kg/s)
0.0106
0.0196
0.0304
0.0797
According to the results shown in Table 3 by increasing the collector inlet temperature the collector
outlet temperature reduces, the optimum flow rate increases and the exergy efficiency increases. As
can be seen the distance between the riser tubes remains to the minimum value, same as the riser
tube diameter except the last value which differs from the minimum value possible, determined by
the constrains. It is apparent that the optimum collector area reaches a maximum value at the inlet
temperature of 27C and then drops for bigger values. The effect of solar radiation is shown in
Table 4.
Table 4. Results of the optimization process-effect of solar radiation
Gt (W/m2)
( ) T in (C)
W (m) D (m)
T out (C)
ex (%)
500
25
0.03
0.004
3.07
66.02
500
27
0.03
0.004
4.85
59.86
0.8
1000
25
0.03
0.004
5.19
95.88
1000
27
0.03
0.004
6.21
92.22
Ac (m2)
7.49
9.98
8.24
9.93
m (kg/s)
0.0106
0.0196
0.0140
0.0190
As can be seen from Table 4 by increasing the solar radiation and keeping the other parameters
constant, the collector outlet temperature increases, as expected, and the same applies to the
optimum collector area and the optimum mass flow rate. Both the riser tube spacing, W, and
diameter, D, remain at their minimum values. The effect of solar radiation and the value of the
transmittance-absorptance product are presented in Table 5.
Table 5. Results of the optimization process-effect of radiation and transmittance absorptance
product
Gt (W/m2)
( ) T in (C)
W (m) D (m)
T out (C)
Ac (m2)
m (kg/s)
ex (%)
500
0.6
0.03
0.004
3.57
49.56
10.0
0.0224
500
0.9
0.03
0.004
5.60
64.56
9.96
0.0190
27
1000
0.6
0.03
0.004
4.10
77.32
9.68
0.0180
1000
0.9
0.03
0.004
7.39
99.03
9.26
0.0180
The results presented in Table 5 reveal that the increase of the transmittance-absorptance product
causes increase in the collector outlet temperature, and a slight decrease in the optimum collector
193
area and the optimum mass flow rate. The exergy efficiency however almost doubles. It is thus
required to achieve a high value of ( ) as possible in flat-plate collectors. Again both, the riser tube
spacing, W, and diameter, D, remain at their minimum values.
As seen from the above tables the distance between the riser tubes, W, is very small. In fact it could
be zero but was stuck to the minimum value specified as a constrain. For this reason a more
practical distance of 10 centimetres between the tubes is used and this parameter is excluded from
the optimization procedure. By doing so the results shown in Table 6 are obtained.
Table 6. Optimization results for a fixed distance between riser tubes of 10cm
Gt (W/m2)
)
T in (C)
D (m)
T out (C)
Ac (m2)
ex (%)
500
0.8
27
0.004
4.72
59.6
9.98
1000
0.8
27
0.004
6.02
92.1
8.38
m (kg/s)
0.0193
0.0156
As can be seen for this more practical case, and by comparing the values shown with the values
presented in previous tables, the exergy efficiency is not affected too much. This case however is
much more cost effective because the minimum the distance between the riser tubes the maximum
would be the number of tubes and the collector will cost more.
Another practical case that needs to be investigated is the effect of riser pipe diameter on the exergy
efficiency. For this exercise the distance between the riser pipes is kept to 10 cm and the other
parameters, like optimum collector area and mass flow rate, as the ones presented in Table 6. The
results of this exercise are shown in Table 7.
Table 7. Effect of pipe diameter for a fixed distance between riser tubes of 10 cm
Gt (W/m2)
)
T in (C)
Ac (m2)
m (kg/s)
D (m)
500
0.8
27
9.98
0.0193
0.004
0.008
0.012
0.015
1000
0.8
27
8.38
0.0156
0.004
0.008
0.012
0.015
ex (%)
4.72
4.71
4.70
4.69
6.02
6.01
5.99
5.99
As can be seen the pipe diameter can be increased from the optimum small diameter without any
problem as the exergy efficiency is marginally affected for the bigger sizes. So this needs to be
decided solely on cost, which increases for bigger diameter pipes and the friction factor imposed by
the smaller diameter pipe, which needs to be kept as small as possible in order not to block the
thermosiphonic flow.
It should be noted that in all results presented in Tables 3-7 a very small flow rate is given as the
optimum and the actual value approaches the thermosiphonically created one which is the most
frequently used operation mode for flat-plate collectors. So the assumption made earlier about the
small flow rate and the exclusion of fluid pumping from the analysis is proved.
7. Conclusions
It is proved in this paper that exergy analysis and genetic algorithms are very suitable for obtaining
the required results quickly. These results prove that the exergy efficiency of a flat-plate solar
collector is maximized for small distances between the riser tubes and for very small diameter of
these tubes. By using a more practical distance of 10 centimeters between the tubes, and excluding
this parameter from the optimization procedure, very small differences are observed in maximum
exergy efficiency and if the cost of the materials is accounted this is a more cost-effective solution.
The exergy efficiency it is also insensitive to the size of the riser pipe diameter. Other findings
194
prove that exergy efficiency increases considerably at higher solar radiation and that the
transmittance absorptance product affects to a great extent the exergy efficiency.
Nomenclature
Ac
B
C
Cb
cp
D
Di
collector area, m2
bond width , m
factor given by (16)
bond conductance, W/(mK)
specific heat capacity, J/(kgK)
riser tube outside diameter, m
riser tube inside diameter, m
Ef
F
f
FR
Gt
h
hfi
hw
I
k
kb
L
m
n
N
Ng
Qs,in
Qu
s
S
Ta
Tf
T f,i
T f,o
Tp
UL
V
W
p
g
References
[1] Badescu V., Optimum fin geometry in flat plate solar collector systems, Energy Conversion and
M anagement 2006; 47(15-16): 2397-413.
[2] Badescu V., Optimal control of flow in solar collectors for maximum exergy extraction,
International Journal of Heat and M ass Transfer 2007; 50: 4311-22.
[3] Torres-Reyes E., Cervantes-de Gortari J.G., Ibarra-Salazar B.A., Picon-Nuez M ., A design
method of flat-plate solar collectors based on minimum entropy generation, Exergy the
International Journal 2001; 1(1): 46-52.
[4] Torres-Reyes E., Navarrete-Gonzalez J.J., Zaleta-Aguilar A., Cervantes-de Gortari J.G.,
Optimal process of solar to thermal energy conversion and design of irreversible flat-plate solar
collectors, Energy 2003; 28(2): 99-113.
[5] Kalogirou S.A., Use of genetic algorithms for the optimal design of flat plate solar collectors,
Proceedings of the ISES 2003 Solar World Congress on Solar energy for a sustainable future,
2003 June 14-19; Goteborg, Sweden, on CD ROM .
[6] Kalogirou S.A., Artificial neural networks and genetic algorithms for the optimisation of solar
thermal systems, In: Sayigh, A.S., Editor, WREC IX: Proceedings of the IX World Renewable
Energy Congress on CD-ROM , 2006; Florence, Italy.
[7] Loomans M ., Visser H., Application of the genetic algorithm for the optimisation of large solar
hot water systems, Solar Energy 2002; 72: 427-39.
[8] Kalogirou, S.A., Use of artificial intelligence for the optimisation of solar systems, International
Journal of Renewable Energy Engineering 2002; 4: 499-505.
[9] Krause M ., Valen K., Wiese F., Ackermann H., Investigations on optimizing large solar thermal
systems, Solar Energy 2002; 73: 217-25.
[10] Kalogirou S.A., Solar Energy Engineering: Processes and Systems, New York: Academic
Press, Elsevier Science; 2009.
[11] Zalzala A., Fleming P., Genetic Algorithms in Engineering Systems, London, UK: The
Institution of Electrical Engineers; 1997.
[12] Goldberg D.E., Genetic Algorithms in Search Optimisation and M achine Learning, Reading,
MA: Addison-Wesley; 1989.
196
Abstract:
Biomass gasification is not a new technology, but there is a renewed interest in its further development,
mainly to produce power and heat as part of locally based combined heat and power plants. The produced
gas mainly consists of H2, CO, CO2, CH4, H2O and some trace impurities such as H2S, COS, NH3, HCN, HCl,
alkali, tar and particulate matter. These im purities are the responsible for clogging, corrosing, poisoning and
carbon deposition on different elements of the power systems, like alternative internal combustion engines,
gas turbines, also in fuel cells and auxiliary equipments, being necessary their removal or the adjusting of its
concentration level, depending on the final application.
This work presents the experimental evaluation of the tar and particle content in t he produced gas from fixed
bed downdraft gasifier with two stages of air supply. A very widely considered technology for t he biomass
gasification is the downdraft fixed-bed reactor, because had shown to produce a gas with lower tar level,
compared with other gasifiers. The ex periments were carried out varying the amount of air supplied to the
reactor and the air ratio between the two stages (AR). The temperatures in different points along of gasifier
and the gas compositions were also measured. The results show that there is an operational point where
there is a coincidence of the highest conversion of the gasifier and the better quality of the gas, (higher
calorific and lower tar content) and also that the use of a s econd stage can reduce the gas tar content up to
87%.
Keywords:
Biomass gasification, Double stage downdraft gasifier, Produced gas, Tar content, particle content.
1. Introduction
Biomass has been considered as a promising source of energy for the partial substitution of fossil
fuels, not only because its global potential, but as a neutral source of CO 2 ; its thermal conversion
generates low emissions of SO2 and NOx, it is cheap and relatively fast-growing, can be grown on
marginal land, without affecting food crops and helps the retention of water and fertilizers in the
soil. The possibility to produce fuels from biomass on a large scale reduces the greenhouse effect,
the environmental pollution and also increases the security and energetic independence of the
countries, but its challenging complexity lies on the need to implement all the stages of the
technological biomass production chain, for the conversion into biofuels and chemicals or in
electrical generation. There are several processes to transform chemical energy from biomass into
thermal energy (combustion, gasification and pyrolysis), of which, gasification is the one with
greatest perspectives, because it offers advantages as: a higher efficiency of conversion, compared
with combustion and pyrolysis. In the practice, the gasification can convert from 60 to 90% of the
197
biomass energy conversion in gas energy [1,2]. The produced gas by biomass gasification is
composed mainly of H2 , CO, CO2 , H2O, CH4 , light hydrocarbons (C x Hy ) and some impurities, as
heavy hydrocarbons (tar), nitrogen compounds (NH3 , HCN), sulphur compounds (H2 S, COS) and
solid particles. This gas can be used for energy production in internal combustion engines, gas
turbines and fuel cells or for the production of synthetic natural gas (SNG) and liquid fuels or
chemicals through the Fischer Tropsch synthetic path [3-7]. Each application needs a specific
quality of the gas used, especially in terms of impurities concentration, which depend on the
characteristics of biomass, the type of the reactor used and the operational conditions [8-12].
Tar formation are one of the biggest problems that occurs during the gasification of biomass, it can
cause blockade, fouling of pipes and valves and operation problems in the equipments where the
gas is used, increasing the maintenance cost of equipments. Thus, it is necessary a gas cleaning
system for the removal of tar, whose complexity depends on the concentration limits for these
pollutants according to the final application [13,14].
Double stage gasification is being used as an economical primary method to reduce the tar content
in the produced gas from biomass [15-16]. The basic concept of a two-stage gasification process is
the separation of the two zones the pyrolysis zone and the combustion zone, with the air injection in
each stage, which may occur in a same reactor or in separate reactors. Recently, experimental
evaluations of the tar content in two and three stage downdraft gasifiers had been reported. These
gasifiers are distinguished by their constructive characteristics, where different biomasses had been
used (composition and shape) and different experimental conditions had been evaluated
(temperature, ratio of the air supplied between the stages), getting different results in terms of tar
and particle content of the gas [17-20], as it is shown in Table 1.
Table 1. Tar content in gas from downdraft gasifiers.
Operation
Tar
Temperature
(mg/
Biomass
(o C)
Nm3 )
Eucalyptus Pelets
(2 cm)
800
Air injection
43.2
950
10
Primary and
secondary air
preheated up to 210
0
C
Coconut husk
700-900
28
780-850
100
Reactor
size
25 cm
diameter,
100 cm
height
25cm
diameter,
190 cm
height,
20 cm
diameter,
Ref.
[17]
[18]
[19]
[20]
Jaojaruek et al [17] made an experimental study and evaluated the relation between the tar content
of the gas and the gasifier operational parameters, conducted in three conditions: a) the reactor
working with a single stage, b) with two stages and c) two stages, with an air-gas mixture fided at
the second stage. The results showed that the second stage reduces the gas tar content in the gas.
Additionally, using an air-gas mixture for the second stage an increase in the calorific value of the
gas was achieved. Bhattacharya and Dutta [18], tested a two stage gasifier and with a layer of coal
for the start up of the reactor. The influence of the level of the coal layer, type of biomass and
moisture content over the tar content and gas composition were checked. It was possible to find out
198
that during the start-up and warming of the gasifier, due to low temperatures, there was a large tar
production; to reduce this period its recommended to use a layer of charcoal, with a height of the
layer slightly above the primary air inlet and adjusting combustion conditions at the second air
stage. They also concluded that, for a constant airflow at the primary stage, increasing the airflow
through the secondary stage leads to a decrease of the tar content and increase the CO and H2
content. Regardless of the results achieved, there is not always a charcoal availability for the
operation of the gasifiers and warming of the air up to 210 o C requires an additional heat source.
Bhattacharya et al [19] tested a three stage gasifier. During the start up of the reactor a layer of
charcoal was also used and was possible observe that the airflow distribution of 40% at the first
stage, 28 % at the second, and 30% at 3rd stage, getting a significant decrease of tar content.
As a conclusion, the decreasing of tar formation is achieved when: a) preheating the air is made, b)
use dual stages of air injection, c) air-gas mixture in the second stage, d) optimal parameter of
operation as ER of 0.3-0.4 and moisture content of 6-20%. The tar content in these gasifiers is in the
range of 10-100 mg/Nm3 , values lower than those obtained in other types of gasifiers.
The objective of this work is to evaluate the tar and particulate content from the produced gas in a
downdraft gasifier with two-stage air injection, analyzing the influence of the total air inlet and the
ratio between stage one and stage two in these pollutants. The tested biomass was eucalyptus wood,
charcoal was not used during the start up of the gasifier and the air supplied was not heated.
measures the temperature of the gas at the gasifier exhaust. The air is supplied by a blower (1900
mmH2 O). The gases generated in the reactor go out by the lower section, after crossing the
gasification zone and the grid and pass finally by a cyclone where the larger solid particles are
removed.
The biomass gasification starts by burning the wood layer by and external heat source, by this way
the reactor is heated up to 150 o C (temperature at the combustion zone in the inner wall of the
reactor). Once this temperature is attained, air is fed into the gasifier. Quasi- stoichiometric
combustion conditions are adjusted reaching a bed temperature (in the combustion zone) of around
600 o C. Air flow values are measured through an orifice plate, according to the methodology given
in ISO 5167-2 [21]. After reaching the steady state (no significant variations in temperature and gas
concentrations are observed) the operating conditions are fixed, the sampling of tar and particles are
made.
Previous experimental tests [22] showed that the condition of the higher gasifier cold efficiency ( )
is 68%. That corresponds to 20 Nm3 /h of total airflow, ratio air/biomass (ER) of 0.4, with air flow
ratio between the first and second stage (AR) of 80%. Under these conditions the concentrations of
CO and H2 was 19.04% and 16.83% respectively, and low heating value of the gas was 4.53
MJ/Nm3 . These conditions were take for the experimental planning. The details of the experimental
installation and termocouples position in the gasifier are shown in Figure 1.
Va
1 est
Va
2 est
.100%
(1)
Where:
(Va) 1est : Air flow through stage 1 (Nm3 /h)
200
Total air
18
20
22
18
20
22
18
20
22
The module one is the pre-conditioning stage of gas and particle collection, consists of an isokinetic
tube connected to the gas line, heated with an electrical resistance to prevent condensation of the tar
inside of the tube and the vessel that contains the filter for the particles collection. Thimble quartz
filter Advantec No. 86R was used. The tube and the port filter are heated up to 250 o C. The
module two, six impingers bottles for tar dilution. The first impinger that is empty, acts as moisture
collector; the others four impinger filled with Isopropanol, in which water and tar are condensed
from the produced gas and the last impinger with silica to dry the gas. Salt/ice/water mixture to
keep the impingers at low temperature was used. The module three contains the vacuum pump to
extract the gas, flow meter and temperature indicator.
The sample is taken during an hour after established operational conditions. After the sampling, a
gravimetric analysis to determine the tar and particles content is made. The filter is weight before of
sampling. After the sampling, the filter that not only retains particle but also a part of the tar is
placed in a soxhlet extraction system, where the residual tar from the particles is separated using
isopropanol. Them, the filter is dried and the particles mass is obtained for weight difference. The
bottles content and the liquid produced in the soxhlet extraction are evaporated by rotavaporation to
remove the solvent and the tar mass is obtained. An analytical balance model BL 210S to determine
the mass of particles and tar in the sample was used. Finally, the tar and particles content per
volume of gas sampled is reported (mg/Nm3 ).
202
Figure 3. Temperature profile of the gasifier for AR of 80% and total airflow of 20 Nm 3 /h
In Figures 2 and 3 its shown that 20 minutes after ignition begins a sudden rise in gasifier
temperatures, is possible to observe how the use of a second stage of air supply increases the
pyrolysis zone temperature, bringing it closer to the temperature of the combustion zone in the
reactor. According of Figure 2 and 3, the average temperature in the pyrolysis zone was 539 o C and
661 o C, and the difference with respect to average temperature in the combustion zone was 174 o C
and 40 o C for one stage and two stages respectively. This behavior reduces the amount of tar
because a high temperature inside the reactor induces the tar destruction. Figure 4 shows the
average temperature of different zone in the gasifier depending of the total air flow.
Figure 4. Average temperature inside of the gasifier for AR of 80% and different airflow.
As its shown in the Figure 4, its possible to confirm that in a two stage downdraft gasifier, the
temperature in the pyrolysis zone is closer to the temperature in the combustion zone, being highly
differentiated at high air flows.
203
LHV
10790. % H 2
12630. % CO
35800. %CH4
(2)
Figure 5 shows the CO, CO 2 , CH4 , H2 concentration of the produced gas for AR of 80% for runs 7
to 9. It also shows LHV, as a function of the total airflow fed to the gasifier. At 20 Nm3 /h1 of total
air flow (run 8) the highest H2 concentration was 17.14 %v and the LHV of 4738 kJ/Nm3 in the
produced gas is reached. At 22 Nm3 /h (run 9), the H2 concentration and the LHV showed a slight
decrease reaching 16.56 %v and 4714 kJ/Nm3 respectively. From this total airflow, the process
begins to be favored by combustion given the temperature increase along the gasifier. The higher
concentrations at these conditions are also attributed to the good stability and performance of the
combustion and pyrolysis zones.
Units
204
According to Figure 6, at the point of higher efficiency of the gasifier when AR is 80% and total air
flow is 20 Nm3 /h, the tar content of 54.09 mg/Nm3 is within the range and even lower than values
reported in Table 1 for other two stage downdraft gasifiers, and without the use of an air- gas
mixture through the second stage. As it is observed in Figure 7, at same conditions, the particles
content presents its lower value of 22.69 mg/Nm3 . The increase of the total airflow and of the AR
values decrease the tar and particles contents in the gas as its observed in Table 4 and confirmed in
Figures 6 and 7. Specifically, this suggests that when an 80% of the total air is entering into the first
stage, the biomass volatization is favored, allowing the formation of lighter compounds that are
more easily cracked when the gas stream pass through the combustion zone, and in this way
reducing the tar content of the produced gas.
205
4. Conclusions
For the gasifier used in this study with operational conditions of 20 Nm3 /h of total air flow, ER of
0.4 and AR of 80%, a fuel gas with levels of CO, CH4 , CO2 , H2 contents of 19.2, 1.34, 14.21 and
17.14 %v was obtained, at these conditions the low heating value of the gas was 4738.4 kJ/Nm3 and
the tar content decrease to 54.09 mg/Nm3 , compared with CO, CH4 , CO2 , H2 contents and LHV of
17.90, 1.72, 13.90, 15.71 %v and 5103 kJ/Nm3 respectively, and 418.95 mg/Nm3 tar content
obtained at AR of 0% and 20 Nm3 /h. These results showed a decreasing in the tar content of 87%.
The results suggest a relationship between the CH4 concentration with the tar content in the
produced gas. In this way, increasing the air supply and use of a second stage decrease the tar
content and the CH4 concentration. This suggest advantages for the biomass devolatilization in the
pyrolysis zone, by production of lighter compounds that are easily cracked when the gas passes
through the combustion zone.
The use of a second stage of air supply increases the temperature at the pyrolysis zone. This allows
to reducing from four to three temperature zones inside the reactor: drying, the first zone, pyrolysiscombustion, the second one and gasification, the third one. This behavior leads to a decrease of the
tar content in the produced gas, so this method is considered an efficient and economic primary
method for tar conversion without air/gas mixture o preheated air as was reported in literature.
Acknowledgments
The authors would like thank the Energy Company of Minas Gerais (CEMIG) and the Secretary of
Science and Technology (SECTES) for the financial support received for the development of this
project. Also to thank the Committee on Coordination of Improvements in Higher Education
(CAPES), the National Research Council of Brazil (CNPq) and the Foundation for Research
Support of Minas Gerais State (FAPEMIG) for the graduate and research productivity grants
allocations.
References
[1] Marsh R, Hewlett S, Grif ths T, Williams K. Advanced thermal treatment for solid waste a
waste managers guide. Proceeding of the 22nd international conference on solid waste
management and technology. Philadelphia (USA); 2007
[2] Kristina Difs, Elisabeth Wetterlund, Louise Trygg, Mats Sderstrm. Biomass gasi catio n
opportunities in a district heating system. Biomass and Bioenergy. 2010: 34: 637651.
[3] Kurkela E, Simell P, mckeough P, Kurkela M. Production of synthesis gas and clean fuel gas
[synteesikaasun ja puhtaan polttokaasun valmistus]. In: VTT Publications, vol. 682. Technica l
Research Centre of Finland, VTT; 2008.
[4] Goyal HB, Seal D, Saxena RC. Bio- fuels from thermochemical conversion of renewable
resources: a review. Renew Sust Energy Rev 2008: 12: 50417.
206
[5] K.D. Panopoulos, L. Fryda, J. Karl, S. Poulou, E. Kakaras. High temperature solid oxide fue l
cell integrated with novel allothermal biomass gasi cation Part II: Exergy analysis. Journal o f
Power Sources 2006: 159: 586594
[6] H. Moritaa, F. Yoshibaa, N. Woudstrab, K. Hemmesc, H. Spliethoff. Feasibility study of wood
biomass gasi cation/molten carbonate fuel cell power systemcomparative characterization o f
fuel cell and gas turbine systems. Journal of Power Sources 2004: 138: 3140.
[7] Anastasia A. Zabaniotou, Vasiliki K. Skoulou, Dimitris P. Mertzis, George S. Koufodimos, and
Zissis C. Samaras. Mobile Gasi cation Units for Sustainable Electricity Production in Rura l
Areas: The smart-CHP Project. Ind. Eng. Chem. Res. 2011: 50: 602608.
[8] Jos Corella, Jos Manuel Toledo and Gregorio Molina. Biomass gasification with pure stea m
in fluidised bed: 12 variables that affect the effectiveness of the biomass gasifier. Int. J. Oil, Gas
and Coal Technology 2008: 1: 194-207.
[9] Li, C.-Z. Some recent advances in the understanding of the pyrolysis and gasi cation behavior
of Victorian brown coal. Fuel 2007: 86: 16641683.
[10] Hong Cui, Scott Q. Turn, Vheissu Keffer, Donald Evans, Thai Tran, and Michael Foley.
Contaminant Estimates and Removal in Product Gas from Biomass Steam Gasification. Energy
Fuels 2010: 24: 12221233.
[11] S. Monteiro Nunes, N. Paterson, D. R. Dugwell, and R. Kandiyoti. Tar Formation and
Destruction in a Simulated Downdraft, Fixed-Bed Gasifier: Reactor Design and Initial Results.
Energy & Fuels 2007: 21: 3028-3035.
[12] Takashi Yamazaki, Hirokazu Kozu, Sadamu Yamagata, Naoto Murao, Sachio Ohta, Satoru
Shiya and Tatsuo Ohba. Effect of Superficial Velocity on Tar from Downdraft Gasification o f
Biomass. Energy & Fuels 2005: 19: 1186-1191
[13] Yan Cao, Yang Wang b, John T. Riley, Wei-Ping Pan. A novel biomass air gasification
process for producing tar- free higher heating value fuel gas. Fuel Processing Technology 2006 :
87: 343 353.
[14] Devi L, Ptasinski KJ, Janssen FJJG. A review of the primary measures for tar elimination in
biomass gasi cation processes. Biomass Bioenergy 2003: 24: 125-1240.
[15] Abu El-Rub Z, Bramer EA, Brem G. Review of catalysts for tar elimination in biomass
gasi cation processes. Ind Eng Chem Res 2004: 43: 6911-6919
[16] Han J, Kim H. The reduction and control technology of tar during biomass
gasi cation/pyrolysis: an overview. Renew Sust Energy Rev 2008: 12: 397-416.
[17] Kitipong Jaojaruek, Sompop Jarungthammachote, Maria Kathrina B. Gratuito , Hataitep
Wongsuwan, Suwan Homhual. Experimental study of wood downdraft gasi cation for a n
improved produced gas quality through an innovative two-stage air and premixed air/gas supply
approach. Bioresource Technology 2011: 102: 48344840.
[18] S. C. Bhattacharya and Animesh Dutta. Two-stage gasification of wood with preheated air
supply: a promising technique for producing gas of low tar content. Energy Program, Asia n
Institute of technology, P O Box 4, Klong Luang, Pathumthani 12120, Thailand.
[19] S.C. Bhattacharya, San Shwe Hla, Hoang-Luang phamm. A study on a multi-stage hybrid
gasifer-engine system. Biomass and Bioenergy 2001: 21: 445-460
[20] Susanto H, Beenackers AACM. A moving-bed gasifier with internal recycle of pyrolysis
gas. Fuel 1996: 75: 1339-47.
[21] ISO 5167-2. Measurement of fluid flow by means of pressure differential devices inserted in
circular cross-section conduits running full. Part 2: Orifice plates. European standard. 2003.
[22] Martnez J. D, Silva L. E. E, Viera A. R, Lesme J. R. Experimental study on biomass
gasification in a double air stage downdraft reactor. Biomass and Bioenergy. 2011: 35: 34653480.
207
[23] Neeft J, Knoef H, Zielke U, Sjstrm K, Hasler, P, Simell P. Guideline for sampling and
anaysis of Tar and particles in Biomass produced gases (Tar protocol). Energy Project ERKGCT1999-2002
208
Abstract
In the present paper we have analysed the possibility to realize an anaerobic digester in a bio-Energy Park
located in Citt della Pieve, a small town in Central Italy. The use of anaerobic digesters is quite common in
Europe for reducing the environmental impact of manure in a co-digestion procedure with vegetables
materials. In addition, for several areas of Central Italy there is the need to find alternative productions to
improve farmers incomes, as traditional cropping systems are loosing convenience. An interesting
alternative seems to be cultivation of energy crops because of the favourable conditions of the electric
energy market. We are suggesting a low input cropping system to be implemented in areas where low input
food/feed crops are no more profitable.
In particular our case-study is an example based on the use of a forage legume, alfalfa (Medicago sativa L.),
together with other crops, like sorghum, to realize small-size bio-digesters plants.
Alfalfa: is a highly sustainable crop as it is able to fix nitrogen and therefore it does not require this
fertilization with the consequence of avoiding underground water pollution. Moreover alfalfa residual products
are nitrogen rich thus improving soil structure and fertility more than popular graminaceous energy crops
such as corn. Beside, alfalfa mostly does not need irrigation in the typical Central Italy environment, all these
traits make it one of the species with the lowest energy needs for growing.
The aims of this feasibility study are: i) optimization of plant materials feeding the bio-digester, ii) typology of
bio-digester, iii) size of bio-digester in relation with land availability for growing energetic cultures, iv) the
utilization of bio-gas produced by bio-digester plant to produce electric and thermal energy using
cogeneration engines, vi) disposal of waste-water produced according to regional and national laws.
The final aim of this study is to verify the possibility to develop an alternative economical use of marginal
soils in relatively dry areas of Central Italy that would be replicable in other European areas with a similar
climatic situation.
Keywords:
Alfalfa, Anaerobic digestion, Digestate, Silage, Sorghum.
1. Introduction
Anaerobic digestion is an appropriate technique for converting biomass such as ensiled energy
crops into renewable energy. In addition, since the digested residue can be used as a fertilizer, a
cropping system based on energy crops has favourable traits of sustainability.
The interest in using ensiled crops for anaerobic digestion is increasing. In Europe the development
of anaerobic digestion began in the sector of civil sewage treatment plants for the stabilization of
sludge and currently it is estimated that there are more than 1600 operational digesters.
At the moment this technique is considered to be one of the best for the treatment of the wastewater
from agro-industrial complexes with high organic content. As early as 1994 there were about 400
business and consortium biogas units while now there are more than 3500 anaerobic digesters
operating on livestock effluent in all countries of the European Union. The highest number is in
Germany followed by United Kingdom and Italy. There are currently about 450 active plants for the
recovery of biogas from MSW landfills with a high concentration in Great Britain. This type of
209
treatment is being increasingly supplemented in recent years by the treatment of the organic fraction
deriving from the differentiated collection of municipal waste (bio waste), digested with other
organic industrial waste and livestock slurries. In Denmark alone there are now 21 centralized codigestion plants of this type, treating about 1,750,000 tons of livestock slurry and 450,000 tons of
organic industrial waste and bio waste.
In 2010, primary energy production from biogas had an impressive increase of 31,3%. Biogas
produced more than 10.9 M toe in 2010, which is an additional 2.6 M toe in just twelve months, and
energy was primarily channeled into electricity production. The power output from this source
should be as much as 30.3 TWh in 2010, which is 20.9% more than 2009 [1].
The country that developed anaerobic digestion at the highest degree in the last ten years is
Germany, particularly in the livestock sector: the German biogas association reports that the country
had 7,100 methanation plants in 2010 with 2,780 M W of electrical capacity. [1]
In 2010 Italy should become Europes number three biogas producer, with primary energy
production estimated at 478.5 ktoe. This is the result of the policy of incentives adopted by the
national government which, in addition to providing a contribution for the investment, pays a price
for electrical energy from biogas which may reach 0.28 /kWh over a period of 15 years.
Central Italy is characterized by a great percentage of farmland localized in marginal areas. These
kinds of areas are suffering to a greater extent of the general crisis of the primary sector due to
lowering of incomes, abandoning of farms and the scarce appeal to the new generations. Eurostat
reports a reduction of farmers income of 3.3% in Italy. Energy from biomasses has become an
interesting alternative to food/fodder crop production in the last years. The majority of power plants
settled in Italy are based on biogas production in medium-large scale farms with animal husbandry.
Because of the general Italian condition for farming this occurs mostly in Northern Italy. Anyway
due to constant loss of income for traditional crops, farmers from marginal areas in central Italy are
seeking an alternative to improve the profitability of their land. The present study analyse the
possibility to suggest a model for biogas production in typical medium-small size farm in marginal
areas where animal husbandry is not as common as it used to be. We are suggesting the use of two
low/modest input crops with a special emphasis on alfalfa. Calculations are based on literature data
but experimental analysis is in progress to test crop yield and biogas profitability in our conditions.
The present work takes into account the Regolamento regionale 4 maggio 2011, n. 4 of Regione
Umbria, concerning the management of facilities for the treatment of livestock manure and biomass
for biogas production and utilization [2]. The new regional framework, in twenty articles,
establishes: 1) the requirements for operation and management of anaerobic digestion plants and
company, 2) inter-treating livestock manure and/or biomass to produce electricity and heat from
biogas power up to 1 M W, 3) the terms of the agronomic use of the resulting digestate from
anaerobic digestion.
In particular, the regulation sets that the materials to be treated must not come from more than thirty
kilometres from the plant and that the same distance should be respected for the transport of
digestate from the plant to the land of the company (art. 9). With articles 10 and 11 regulates the
management and agronomic use of digestate and possible further treatment.
over 100C and at a variety of moisture contents from around 50% to more than 99%.. Bacteria
living optimally at temperatures between 35 40 C are called mesophiles, those surviving warmer
and more hostile conditions at 5560 C are called thermophiles.
In the absence of oxygen, anaerobic bacteria ferment biodegradable matter into methane and carbon
dioxide, a mixture called biogas. Biogas contains 6070% methane and 3040% carbon dioxide
depending on the feedstock type [3]. Trace amounts of hydrogen sulfide, ammonia, hydrogen,
nitrogen, carbon monoxide, oxygen and siloxanes are occasionally present in the biogas. Usually,
the mixed gas is saturated with water vapor.
Anaerobic digestion takes place in basically three stages. In the first stage, complex organic
macromolecules are hydrolyzed into simpler soluble molecules. In the second stage these molecules
are converted by acid forming bacteria to simple organic acids, carbon dioxide and hydrogen; the
principal acids produced are acetic acid, propionic acid, butyric acid and ethanol. In the third stage,
methanogen bacteria form methane, either by breaking down the acids to methane and carbon
dioxide, or by reducing carbon dioxide with hydrogen [4].
The biogas produced in an anaerobic digestion energy plant consists of 5580% CH 4, 2045% CO2,
0.01.0% H2S, and 0.00.05% NH3, and it is saturated with water [5].
211
212
alfalfa, /ha
90.00 100.00
50.00 60.00
50.00 60.00
60.00 70.00
250.00 300.00
500.00 600.00
As equipment for harvesting of sorghum, a combined forage harvester (mowing, chopping, loading)
was accounted, for alfalfa swathing, raking, chopping of windrows and loading operations were
considered.
We have used a combination of data from different literature sources to proceed with the
calculations reported in the following parts of the paper.
4. Feasibility study
The chain for producing methane through anaerobic digestion from energy crops is presented in
Fig. 1, from the production and harvest of crop biomass, to storage and pre-treatment of the
biomass, production and utilization of biogas, storage, post-methanation and post-treatment of the
digestate, and finally returning the digestate back to the crop production areas as fertilizer and soilimprovement medium [12].
The most important parameter in choosing crops for methane production is the net energy yield per
hectare, which is defined mainly by biomass yield and convertibility of the biomass to methane, as
well as cultivation inputs. Energy crops should be easy to cultivate, harvest and store, tolerant to
weeds, pests, diseases, drought and frost, have good winter hardiness and be able to grow on soil of
poor quality with low nutrient input.
Due to the problems related to the animal manure management, the present feasibility study
concerns a co-digestion of vegetable matrices, in particular alfalfa and sorghum.
The co-digestion of two vegetable matrices may create problems in regard to the activation of the
anaerobic digestion process. Because of this it may be necessary to make an inoculum of slurry to
help the activation process.
Table 2. Methane and gross energy potentials of energy crops and crops residues
Substrate
Methane potential
Gross energy
potential
Forage beet
Alfalfa
Potato
Maize
Wheat
Barley
Rape
Grass
Clover
Marrow
kale
Jerusalem
artichoke
Sugar beet
tops
Straw
Ref.
(m3CH4 kg-1TSadded )
0.46
0.36
0.41
0.32
0.28
0.41
0.39
0.36
0.34
0.41
0.27
0.27-0.35
0.35
0.14-0.21
0.26
0.32
0.27
n.r.
0.32 c
n.r.
0.28 c
n.r.
n.r.
n.r.
n.r.
n.r.
n.r.
0.24 c
0.25-0.32
n.r.
0.12-0.19
n.r.
0.28 c
0.24 c
n.r.
55 c
n.r.
56 c
n.r.
n.r.
n.r.
n.r.
n.r.
n.r.
46 c
64-83
n.r.
24-36
n.r.
42 c
49 c
5800 a
3240 b
3965 a
2304 b
2280 a
5780 a
2960 a
2030 a
1190 a
4060 a
1908 b
n.r.
2530 a
n.r.
1680 a
2304 b
2862 b
56 ac
34 b
38 ac
24 b
22 ac
56 ac
28 ac
20 ac
12 ac
39 ac
20 b
n.r.
25 ac
n.r.
16 ac
24 b
30 b
1
2
1
2
1
1
1
1
1
1
2
3
1
3
1
2
2
0.23
0.36-0.38
0.25-0.26
0.30 c
0.19 c
0.29-0.31 c
0.23-0.24
0.25 c
n.r.
36-38 c
139-145
n.r.
n.r.
n.r.
n.r.
n.r.
n.r.
n.r.
n.r.
n.r.
4
5
3
6
in Germany, b in Sweden, c Values calculated from the data report ed, a = year, n.r. = not reported. 1: Weiland 2003, 2:
Brolin et al. 1988, 3: Kaparaju et al. 2002, 4: Gunaseelan 2004, 5: Zubr 1986, 6: Badger et al. 1979.
This study assesses the sizing of the anaerobic digestion dimension plant, the quantification of the
necessary agricultural area for the crops cultivation and for the agronomic utilization of the digested
produced by the anaerobic digestion process.
214
We have considered a size of the CHP engine of 250 kW, which is a power that fits the needs of a
typical medium-small size farm. In addition the following technical parameters are assumed:
M esophilic temperature range;
Hydraulic retention time (HRT) equal to 28 days;
Co-digestion of alfalfa and sorghum.
Following these assumptions, we have determined the amount of alfalfa and sorghum necessary to
feed the 250 kW CHP engine. Considering a lower heating value of the biogas equal to 6.8 kWh/m3,
an efficiency of the engine equal to 36% and 7,500 operating hours for year, we obtain an amount
of biogas equal to 766,544 m3 per year.
In order to evaluate the amount of alfalfa and sorghum necessary for the production of biogas
amount stated above, we have considered the parameters reported in table 3 (regarding alfalfa) and
table 4 (regarding sorghum).
Table 3. Alfalfa silage parameters
Alfalfa silage
Dry matter, DM
40%
Organic dry matter, ODM
85%
Biogas yield
0,55 m3 biogas/kg ODM
fifty centimeters. The design must include all the necessary measures to minimize odorous
emissions [2].
In fact, the application of digestate at times of the year when there is little plant uptake, for instance
autumn and winter, can result in nutrient leaching and runoff into ground and surface waters (e.g. of
N and P). Digestate must therefore be stored until the correct time of application.
Digestate applications should be matched with crop nutrient requirements; this will minimize any
unintended negative impact to the environment and also maximize farmers profits. Table 5 reports
application rates (especially for nitrogen), length of storage periods, and timings for applications
that must also comply with national limits.
Table 5. National limits regulating nitrogen loading on farmland, required storage capacity and its
spreading season
Maximum nutrient load
Required storage
Compulsory season for
capacity
spreading
Austria
170 kg N/ha/year
6 months
28 feb 5 oct
Denmark
170 kg N/ha/year (cattle)
9 months
1 feb harvest
140 kg N/ha/year (pig)
Italy
170 - 340 kg N/ha/year
150 days
depends on the weather
conditions
Sweden
170 kg N/ha/year
6 10 months
1 feb 1 dec
Northern Ireland
170 kg N/ha/year
4 months
1 feb 14 oct
Germany
170 kg N/ha/year
6 months
1 feb 31 oct arable land
1 feb 14 nov grassland
5. Conclusions
This study has assessed the opportunity to realize an anaerobic digester in a bio-Energy Park
located in Citt della Pieve, a small town in Central Italy.
We have chose to evaluate the anaerobic co-digestion of vegetable matrices such as alfalfa and
sorghum silage and we have evaluated the principal process parameters.
In particular, assuming a 250 kW CHP engine, we have determined the biogas necessary, the
volume of the digester, the amount of alfalfa silage and sorghum, and also the areas required to their
production.
The area requested seems to be rather relevant, depending on the lower productivity of the crops
chosen as alternative to the largely used energy crops for biogas production, such as maize. The
economic feasibility is being currently investigated also taking into account smaller digester size
and that marginal lands should be used. Further analysis including other low-input energy crops as
suggested by latest literature in the field will be carried out.
Finally we have reported some important consideration about the digestate management, that could
be a great opportunity but also a critical point of the system.
Nomenclature
CSTRs
HRT
ODM
TS
TVFA
VFA
VS
ww
Volatile Solids
wet weight
References
[1] Eurobserver, The state of renewable energies in Europe. 2011 edition.
[2] Regolamento regionale 4 maggio 2011, n. 4 Norme di attuazione dell'articolo 4, comma 1,
lettera e) della legge regionale 10 dicembre 2009, n. 25 concernente la gestione degli impianti
per il trattamento degli effluenti di allevamento e delle biomasse per la produzione di biogas e
l'utilizzazione agronomica delle frazioni palabili e non palabili Bollettino Ufficiale della
Regione Umbria n. 21 del 11 M aggio 2011.
[3] M . Plchl, M . Heiermann, B. Linke, H. Schelle. Biogas Crops Part II: Balance of Greenhouse
Gas Emissions and Energy from Using Field Crops for Anaerobic Digestion. A gricultural
Engineering International: the CIGR Ejournal 2009. M anuscript number 1086. Vol. XI. M onth.
[4] R. M . Jingura, R. M atengaifa, Optimization of biogas production by anaerobic digestion for
sustainable energy development in Zimbabwe, Renewable and Sustainable Energy Reviews 13
(2009) 11161120.
[5] L. Xiaohui, N. Sugiura, F. Chuanping, M. Takaaki, Pretreatment of anaerobic digestion effluent
with ammonia stripping and biogas purification, Bioresource Technology 100 (2009) 5478
5484.
[6] Plch M ., Heiermann M . Biogas Farming in Central and Northern Europe: A Strategy for
Developing Countries?. A gricultural Engineering International: the CIGR Ejournal. Invited
Overview No. 8. Vol. VIII. March, 2006.
[7] Kttner M ., Dry fermentation a new method for biological treatment in ecological sanitation
systems (ECOSAN) for biogas and fertilizer production from stackable biomass suitable for
semiarid climates. In Proc. 3rd International Conference on Environmental M anagement, pp 16.
Johannesburg, South Africa 2002.
[8] Chen F., Dixon RA Lignin modification improves fermentable sugar yields for biofuel
production Nat Biotechnol 2007. 25(7):759-61.
[9] Borreani G, Tabacco E. Insilamento dellerba medica per valorizzare le proteine. Terra e vita
2008, 14: 80-82.
[10] M . Heiermann, M . Plchl, B. Linke, H. Schelle, C. Herrmann. Biogas Crops Part I:
Specifications and Suitability of Field Crops for Anaerobic Digestion. Agricultural
Engineering International: the CIGR Ejournal. M anuscript 1087 2009. Vol. XI. June.
[11] Candolo G. Energia dalle biomasse vegetali: le opportunit per le aziende agricole
Agronomica 2006, 4:26-35.
[12] A. Nordberg, A. Jarvis, B. Stenberg, B. M athisen, B. H. Svensson. Anaerobic digestion of
alfalfa silage with recirculation of process liquid. 2006.
217
Abstract:
For both environment al and economic reasons, the use of biogas for heat and power generation (CHP),
especially on a small and decentralized scale, is predicted to increase dramatically in the years to come.
However, these unconventional fuel gases present new challenges to manufacturers of combustion systems
as their properties differ from natural gas. In general, their calorific values (LCV) are much lower than those
of natural gas, as they contain large amounts of inert species such as CO 2 or N2. Also, their chemical
compositions may vary significantly over time and they may contain species such as HCN or NH3, leading to
increased NOx formation during combustion due to fuel-bound nitrogen. While NOx formation due to fuelbound nitrogen is common in coal combustion, NOx reduction measures for the combustion of gaseous fuels
are usually aimed towards the reduction of thermal NOx formation and are thus not able to prevent the
conversion of chemically bound nitrogen in biogas into nit rogen oxides.
In the course of several res earch projects, Gaswrme-Institut e.V. Essen (GWI) investigated on how to best
make use of these renewable fuels in future combustion systems. Using both numerical and experimental
techniques, several burner systems were developed which can achieve a stable combustion of different
types of biogases with a minimum of NOx formation. Using CFD simulations, burners based on the COS TAIR
and flameless oxidation (FLOX ) principles were modified to operate with low calorific value fuel gases. The
performance of these burners was then further investigated by experiment al investigations in GWIs semiindustrial test rigs where a satisfactory agreement between numerical and experimental data was observed.
In a further step, the COSTA IR burner was then mounted into a commercially available 100 kW micro gas
turbine (MGT) and tested under real operating conditions. It was shown that the combustion system was
able t o operate in a stable manner while producing only a minimum of NOx -emissions, making the
combination of a MGT and a burner system optimized for low calorific value gases an ideal choice for small
scale decentralized combined heat and power applications.
Keywords:
Low Calorific Gases, Alternative Fuels, Simulation, Fuel-Bound Nitrogen, COSTA IR, Flameless Oxidation,
Product Gases formed from Biomass, NOx -Emissions
1. Introduction
Product gases formed in biomass gasification plants usually have a fluctuating lower heating value
and contain several different chemical species, depending on the type of process and biomass being
used. The gasified biomass produces a low calorific value gas that can be used in a similar manner
as natural gas. Generally some of the more common chemical species of low calorific value product
gas are e.g. methane (CH4 ), carbon monoxide (CO), hydrogen (H2 ), carbon dioxide (CO2 ), nitrogen
(N2 ) and traces of benzol, toluol and tars. Low calorific product gases are defined as gases with
lower heating values (LCV) less then < 3 kWh/mN3 . The large percentage of N 2 and CO2 and the
low CH4 contents of many (LCV) product gases are the reason why the lower heating value
decreases significantly compared to natural gas. However the term low calorific value gas is not
solely restricted to product gases formed in gasification processes, but in a broader sense sums up
various types of product gases produced in steel mills, chemical processing facilities or landfills. An
218
overview of typical chemical compositions and properties of some common low calorific gases is
listed in table 1.
Table 1. Compositions and Properties of various Low Calorific Value Gas
In a recent research project conducted by the Gaswrme-Institut e. V. Essen, two burners based on
the different concepts of FLameless OXidation (FLOXT M) and COntinuously STaged Air with
Internal Recirculation (COSTAIR) were subject to extensive numerical and experimental
investigations. The main goal of the project was to develop a method to reduce the NO x -emission
levels during combustion of biogases containing fuel-bound nitrogen. In past research projects, both
burner concepts have proven to be low NO x -emission technologies capable of handling fluctuating
fuel gas qualities and low calorific values [2]. The prototype burners were developed in three steps.
As a first step, the unmodified burners were simulated with CFD FLUENT with regards to their
predicted NO x-emission levels while combusting low calorific product gas containing fuel-bound
nitrogen. In the next step of the project several geometric modifications and improvements to the
burners were simulated, again applying the same gases and simulation parameters. Finally, both
burners were built and experimentally tested at GWI. The aim of this paper is to give readers a
greater insight into the numerical and experimental results as well as the approaches taken to
develop a low NO x combustion system for low calorific gases containing fuel-bound nitrogen.
2. Numerical Development
The COSTAIR burner concept applies continuously staged air with internal recirculation of flue
gases to achieve a stable combustion with low NO X- and CO- emission levels. Air is injected
perpendicularly to the main flow by means over a large number of small nozzles on the surface of
the air distributor, thus continuously varying the local air ratio. The internal recirculation of the flue
gas helps to produce a stable, uniform combustion that minimizes peak flame temperatures and
pockets of oxygen in the combustion reaction zone [3]. The fuel gas nozzles are equally spaced
around the central air distributor to ensure that the combustion products mix appropriately with the
continuously staged air. In Fig. 2 a simplified sketch of the COSTAIR burner is given.
Throughout the course of this project, CFD simulations played an important role in the development
and optimization of COSTAIR burner s. The main idea behind reducing the high NO x -emission
levels of product gases with fuel-bound nitrogen was to stage the air in such a manner that the
combustion reaction zone is globally split into a sub-stochiometric ( < 1) reduction zone followed
by over-stochiometric ( < 1) burnout region. The gases simulated in the numerical models
consisted of the chemical species CO, H2 , CH4 , CO2 and N2 . By injecting small amounts of
ammonia gas (NH3 ) into the product gas,the presence of fuel-bound nitrogen was emulated. The
simulation results show that adding small amounts of NH3 to the fuel gas, caused the NO x emission
levels to increase drastically during combustion since the fuel NOx formation pathway yields much
220
more nitric oxides than the thermal NOx production pathway which is usually dominant in the
combustion of gaseous fuels.. During the development of the COSTAIR burner a strong emphasis
was focused on the design of the air distributor. The initial CFD results showed that the positioning
and size of the air jets on the air distributor had a tremendous impact on the stability of the
combustion zone and and the increase of NO x emissions levels. The results also confirm that
sectioning the air jets on the air distributor causes the reaction zone to shift within the combustion
chamber. If for instance the reaction zone is shifted behind the air distributor a considerable rise in
NOx -emission levels is calculated. After determining the basic design of the modified COSTAIR
burner, further optimization of the burner and air distributor geometries was required in order to
effectively burn low calorific value gas containing fuel-bound nitrogen. A promising approach
taken was to extend the length of the air distributor in order to divide the reaction zone more clearly
into a sub-stochiometric and over-stochiometric reaction zone. Another important aspect during the
development of the prototype burners was the size and positioning of the fuel gas nozzles.
Therefore the optimum distance and angle of the fuel jets to the air distributor were numerically
determined as well. In Fig. 3 the simulated velocity and temperature distributions of both
COSTAIR burner variants are shown. The temperature and velocity distributions confirm that a
stable combustion is possible with both COSTAIR variants. By extending the length of the air
distributor the combustion reaction zone is clearly divided into two regions with different air ratios.
Most importantly, the CFD results confirm that a reduction of NO x emissions from fuel-bound
nitrogen can be achieved with the optimized air distributor.
experimental tests performed at GWIs testing facilities, the same CFD models and boundary
conditions used to investigate the COSTAIR burner were applied to adapt a FLOX burner system
to low calorific product gas. The cross sections of the fuel and air nozzles were increased in order to
generate the required gas velocities for flameless oxidation.
Further evaluation of the results showed that placing the secondary air inlets behind the combustion
reaction zone leads to a considerable decrease in NO x -emission levels. The temperature and velocity
distributions shown in Fig. 5., illustrates that the injection of secondary air does not influence the
recirculation of flue gases in the combustion reaction zone. The maximum peak flame temperature
of the FLOX burner is 250 C lower compared to the COSTAIR burner. The largest reduction of
NOx -emissions was achieved when the primary and secondary air inlets were spaced 1000 mm
apart. Injecting the secondary air behind the main combustion reaction zone causes the maximum
flame temperatures to decrease while simultaneously the primary reaction zone is divided into two
different reaction regions with different air ratios, thus leading to a decline of NO x - and CO
emission during the combustion of low calorific product gas containing gaseous ammonia.
222
3. Experimental Investigations
Fig. 6. Layout of the Test Rig (bottom left COSTAIR Burner, bottom right FLOX Burner)
A schematic drawing of the test rig along with images of the FLOX and COSTAIR burners are
presented in Fig. 6. The experimental investigations of the COSTAIR and FLOX prototype
burners were conducted in two testing stages. During the first phase the original COSTAIR and
FLOX burners were tested without modifying the original burner geometries and secondary air
staging inlets. During the second stage of testing the air staging concepts developed for the both
burner prototypes were investigated and evaluated. The testing conditions for both burners were
kept constant throughout the course of the experimental investigations in order to directly compare
the testing results. The low calorific gases used in the experimental tests at GWI were synthetically
produced with a gas mixing station, while a gas pre-heater warmed up the gas mixtures to 350C in
order to simulate the actual operating temperatures of a variety of product gases formed in landfills
or gasification processes. One of the fuel gases used was seeded with small amounts of ammonia
gas (NH3 ) in order to investigate the impact of fuel-bound nitrogen on NO x-emissions. The
ammonia gas was added to the fuel gas in increments of 1000, 3000 and 5000 ppm. The
composition of the flue gas was measured with an exhaust probe to determine concentrations of CO,
CO2 , O2 , NO and NO2 . Moreover, the pressures of the fuel gas and air along with the temperature of
the flue gas were recorded.
223
COSTAIR
COSTAIR
[ppm]
NOx [ppm]
NOx [ppm]
@ 3 Vol.-% O2
@ 3 Vol.-% O2
NOx
Reduction
[%]
1.2
9.77
7.51
23.13
1.2
1000
384.45
37.45
90.26
1.2
3000
735.44
78.93
89.27
1.2
5000
1275.33
151.42
88.13
224
The data collected during the first set of burner tests served as points of reference in order to
determine the total reduction of NOx -emissions and the effectiveness of the optimized air distributor
on NOx -emissions during the second stage of testing. A tremendous decline in NO x -emissions was
demonstrated by imposing a strong gradient of the local air ratio along the distributor length of the
COSTAIR burner while burning product gases containing fuel-bound nitrogen.
0
1000
3000
5000
6.91
56.85
118.66
193.74
6.38
36.18
64.62
96.88
NOx
Reduction
[%]
7.67
36.36
45.54
50.01
The measurements of the basic FLOX burner variant reveal that less NOx-emissions are formed
when compared to the basic COSTAIR burner. However the use of optimized air distributor had
225
greater impact on reducing NO x -emissions compared to the secondary air used with the FLOX
system. A direct comparison of the optimized burners generally shows that a reduction of NO x emissions is achieved equally well with both combustion concepts. The results confirm that nearly a
50% reduction of NO x -emissions from product gases containing fuel-bound nitrogen is possible
with the FLOX burner when the combustion air is staged.
Fig. 9. Close Up View of the MGT Fig. 10. Simulation of the Exit Velocities of the
COSTAIR Burner
Pilot Burner Nozzle
Further evaluation of the numerical results concluded that a small area circa 30 mm above the gas
jet forms a low velocity field containing an ignitable gas/air mixture. Corresponding to the
numerical results a new burner flange plate was constructed and installed in the MGT T100 for
226
validation testing. In Fig. 12 the optimized COSTAIR burner is shown (please note the position of
the pilot gas nozzle and spark ignition). In the adjacent images the entire COSTAIR burner
including the head pipe of the MGT T100 is displayed.
Fig. 11. Radial Velocities 30 mm above the Fig. 12. Adjusted Burner Geometry(left) and Head
Gas Nozzle
Pipe Mounted to the Burner (right)
5. Conclusion
Throughout the course of this project, the suitability and optimization of the COSTAIR and FLOX
combustion systems using low calorific product gases containing fuel-bound nitrogen were
investigated. One of the first steps taken in this project was to develop two burner systems for low
calorific product gases. The influence of different fuel gas compositions and properties was
simulated with CFD FLUENT under realistic burner operating conditions. During the second stage
of development, both burner designs were optimized to efficiently burn low calorific value gases
and reduce NO x -emissions of product gas containing fuel bound nitrogen. A variety of NO x reduction methods were considered and simulated in numerous simulation approaches. Finally, both
prototype burners were installed and tested in a GWI test rig. The results show that secondary air
staging is an effective method to reduce NO x- emissions due to fuel-bound nitrogen. Also both
burners responded stably to fluctuations of the fuel gas composition and maintained low emissions
over a wide range of operating conditions.
The outcome of these projects shows that applying secondary air staging to FLOX and COSTAIR
burner systems leads to a considerable decline in NO x -emission levels. Yet a further conclusion is
that small traces of fuel bound nitrogen in product gases causes the NO x -emission levels to increase
immensely. The developments and combustion concepts investigated in this project, may soon pave
the way for new and innovative CHP applications such as micro gas turbines, fuel cells, gas
powered engines to utilize low calorific product gases containing fuel-bound nitrogen in an
economically feasible manner. The developed burner is able to combust low calorific fuel gases
with LHVs as low as 1.25 kWh/Nm3 (corresponding to a CH4 concentration of about 12.5 vol- %)
without releasing high NO x -emissions. Furthermore, the necessity to treat product gases containing
fuel-bound nitrogen can be reduced considerably using both combustion concepts as primary
measures to reduce NO X formation in the combustion space. The use of alternative fuels combined
with innovative burner systems may give industrial plant operators a technical advantage to react in
a more competitive and flexible manner to fuel price fluctuations. On behalf of all the
corresponding authors, we thank the Department of Energy Plant Technology, Ruhr-University of
Bochum, Germany.
227
References
[1] Brink, A., Huppa, M., Kurkela, E., Suomalainen, M.: Nitric oxide yield from combustion of a
low calorific gasification product gas: numerical and experimental results, Progress in
Computational Fluid Dynamics, Vol. 6, No. 4/5, 2006
[2] AiF-Projekt MGT: Neue Brennersysteme zur dezentralen Nutzung von schwachkalorige n
Gasen in Mikro-Gasturbinen. AiF-Vorhaben Nr.: 13216 N, Projektdauer: 01.03.2002 bis
31.08.2004
[3] Al-Halbouni, A.: Entwicklung NO x-emissionsminimierter Heizkesselfeuerungen. Habilitation,
Otto-von-Guericke-Universitt Magdeburg, Shaker Verlag 2001
[4] Maclean, S., Tali, E., Giese, A., Leicher, J., Grner, K.: Investigations on the Reduction of NO x
Emissions of Product Gases containing Fuel Bound Nitrogen in Thermal Processing Plants, 9th
European Conference on Industrial Furnaces and Boilers, Porto, Portugal, 2011
[5] Leicher, J., Giese, A., Grner, K., Scherer, V., Schulzke, T.: Development of a Burner Syste m
for Use of Low Calorific Fuel Gases in Micro Gas Turbines, Proceedings of the Europea n
Combustion Meeting, Cardiff, England 2011 .
228
Abstract:
SEAKERS (SEA Kinetic Energy Recovery System) is a research project, funded wit hin t he 7th EU
Framework Programme and officially started on January 1st, 2011, whose goal is to develop an innovative
devic e consisting in a kinetic energy recovery system for sailing yachts based on the conversion of boat
oscillations (heave, pitch and roll) caus ed by the sea into electric energy by means of a linear generator.
Therefore, SEAKERS addresses a well known unsatisfied requirement of yacht owners, since energy is a
resource of primary importance in a boat, especially in a sailing one: it is well known that during a one day
cruise, electricity consumption has to be carefully managed (for instance the refrigerator is switched off), so
as not to be short of energy at night. It often happens that, after one day of sail cruise, it is necessary to
recharge the batteries through the onboard generator, which means keeping it on for hours, producing very
annoying noise, smoke and pollution.
The device that is going to be developed aims at recovering as much kinetic energy as possible from the
natural movements of a sailing yacht on the sea, therefore taking the view of a boat as a moving wave
energy converter with energy harvesting capacity. The boats motions can be vertical oscillations due to the
buoyancy in the presence of sea waves, both when the boat is still or sailing, and rolling and pitching
motions originated both by sailing in wavy waters and by the normal boat dynamics due t o the sails
propulsion. Linear generators will convert kinetic energy into electrical energy to be used as green
electricity for any possible application on board.
Preliminary calculations show that a properly configured system could be able to recover 100-400 W under
most sea conditions, which can be an extremely attractive result since an electric energy availability of 1-2
kWh on a sailing yacht is of signific ant interest.
Keywords:
Wave Energy Recovery, Linear Generator, Sail Yacht.
1. Introduction
This paper presents some preliminary results obtained in the SEAKERS project, whose aim is to
design and test a kinetic energy recovery system to be used on board of sail yachts in order to
recover energy from the wave-induced boats vertical motion.
Such a system is able to recover actual free energy, as opposed to other devices, already
commercially available, that subtracts energy from the propulsion offered by the winds lift on the
sails, as in the case of micro-wind turbines installed on the boat, which are set into motion by the
apparent wind originating from the yachts motion.
In practical terms, the SEAKERS device is intended to be a linear oscillator, with a mass oscillating
vertically inside a prismatic guide and gaining kinetic energy; if the mass is the moving element of
a linear generator, the resulting mechanical energy can be extracted and converted into electricity.
The oscillating mass incorporates permanent magnets which, moving in proximity of stator
windings, generate electric power due to electromagnetic induction.
The SEAKERS device addresses a well known unsatisfied requirement of yacht owners, since
energy is a resource of primary importance in a boat, especially in a sailing one: it is well known
that during a one day cruise, electricity consumption has to be carefully managed (for instance the
refrigerator is switched off), so as not to be short of energy at night. It often happens that, after one
229
day of sail cruise, it is necessary to recharge the batteries through the onboard generator, which
means keeping it on for hours, producing very annoying noise, smoke and pollution.
The idea of a linear generator originates from work carried out at the University of Uppsala [1-4],
where such devices have been designed and tested in order to recover wave energy from a buoy,
oscillating on the sea surface, connected to a rope that makes a piston move inside a generator
placed on the seafloor. In the SEAKERS project, the oscillating mass is set into motion not directly
by the sea waves but by its inertia as the yacht is subject to heave, pitch and roll motions.
In order to design the test-bed for testing the generator, it is necessary to set up a reliable model of
different sea conditions that could be of practical interest for a normal cruise on a sail yacht (thus
there is no need to consider extreme, stormy waves) and of the ship motion due to such sea states.
Furthermore, it is interesting to find out, by means of a very simple mechanical model of the linear
generator, how much power could be extracted under these simplifying assumptions, in order to
decide whether the projects outcome could in principle be commercially viable, and quickly to
provide data against which results from more detailed analytical models and experimental tests
could later be compared.
This paper presents the results obtained in this first stage of the project, detailing first the model of
sea conditions (section 2), then the outcome of simulations on the yachts motion carried out by
means a commercial software (section 3), and finally the results of the analysis carried out on a
linear mechanical system located on the boat (section 4).
2. Wave excitation
2.1.
Wave spectra
The main characteristic of sea waves is randomness. Indeed, by checking even a short a time series,
two characteristics arise: height and period of a wave are different from height and period of
another wave. For this reason, the free surface elevation of sea waves is modelled as a stochastic
process and is assumed to be a random, Gaussian, ergodic process in the time domain [5-8].
Mathematically, sea elevation can be reconstructed in one dimension as a Fourier series as follows:
n
x, t
Z j cos
t kjx
(1)
j 1
In this equation, Zj is the wave amplitude for the j-th wave form of circular frequency j, k j its wave
number (dependent on j through the dispersion relation) and j its phase shift. The dispersion
relation defines the relationship between wave frequency and wave number; in deep water it is
expressed as [5-8]:
2
kg
(2)
where g is the acceleration of gravity. (It may be useful to recall that wave number and wave length
are mutually dependent: k 2 / ).
Given a sea elevation time pattern for a given spatial coordinate x, the amplitudes Zj of its Fourier
series may be evaluated as Fourier transforms of :
1
T
Zj
T
2
T
2
t exp
i2
(3)
t dt
The most meaningful representation from a statistical point of view of a particular sea state is given
in the frequency domain by means of the wave spectrum S
, which is defined as:
S
Z j Z *j
Zj
(4)
230
Z *j being the complex conjugate of Zj. Therefore, the spectrum S j is proportional to the energy
content of the j-th wave form of circular frequency j, while the area under the spectrum s
proportional to the overall energy content of the sea state described by sea elevation t :
n
Zj
(5)
j 1
(6)
By virtue of equation (5), the energy spectrum is correlated to the overall energy content of the sea
state, because the energy content of a single sinusoidal wave is proportional to the square of its
height. Furthermore, statistical data that can be gleaned from the energy spectrum correspond to
important parameters for the description of a sea state. Of particular importance is the 0-th spectrum
moment, which is equivalent to the area under the wave spectrum curve:
m0
(7)
For a narrow band spectrum, it can be demonstrated that the root mean square (RMS) wave
amplitude is given by m0 , and the RMS value of wave height (crest to trough) is therefore:
H RMS
2 m0
(8)
One of the most useful parameter to represent the sea state is the significant wave height, which is
the mean of highest third wave heights, and for narrow band spectrum it is given by [5-8]:
Hs
4 m0
(9)
2.2.
2 m0
(10)
Simulation assumptions
The simulations that will be presented in the following sections were carried out taking into account
statistical wave data for the Mediterranean Sea, with particular reference to the measurements taken
at Capo Linaro (Civitavecchia, Italy) 1 .
In the case of random waves, it is possible to find a particular set of parameters that make the
JONSWAP spectrum suitable to represent sea conditions in the location of interest (the above
mentioned Capo Linaro near Civitavecchia).
The JONSWAP spectrum was developed from extensive field measurements in the context of the
Joint North Sea Wave Project [5-8]. This formulation is suitable for wind-generated waves in fetch
limited locations. The inputs are the wind speed and the fetch length. The mathematical formulation
is given by equation:
Personal co mmunications with Prof. Felice Arena, University of Reggio Calab ria, 2011.
231
S ( )
5
p
5
4
exp
p
2
)2
2
p
(11)
by
0.0076( gx / U ) 0. 22 , where x is the fetch length and U the mean wind speed, and is the
peak-shape parameter. For practical applications, can be assumed equal to 0.08 in the whole
frequency domain.
In deep water, wave period and length are correlated by the dispersion relation (2), which may be
rewritten as follows:
gT 2 2
(12)
(13)
Furthermore, wave period is also related to the significant wave height through parameters
Tp
Hs g
and :
(14)
Thus, higher waves are longer (12), propagate faster (13) and are less frequent (14). The JONSWAP
spectrum is completely defined when the significant wave height Hs and parameters and are
specified.
In order to represent correctly sea conditions at Capo Linaro, values of , , f and Tp are chosen
according to the following table. The corresponding wave spectra are illustrated in fig. 1.
Table 1. Parameters used to represent random sea waves at Capo Linaro near Civitavecchia, Italy.
Hs [m]
Tp [s]
f ,
0.5
0.016
1.0
13.2
2.98
1.0
0.008
2.0
14.9
4.75
1.5
0.010
0.5
15.5
6.06
2.0
0.008
0.5
16.4
7.40
Fig. 1. Wave spectra representing sea conditions at Capo Linaro near Civitavecchia, Italy.
232
3. Yachts response
3.1.
Encounter frequency
Due to its forward speed V, the wave spectrum for the ship is different than for a fixed observer.
When studying the ships response it is therefore necessary to take into account the frequency at
which it actually encounters the waves (encounter frequency). The encounter frequency depends on
wave velocity and ship speed and relative direction with respect to waves . Angle is defined
between the forward directions of wave and ship: thus for bow waves = , for transverse waves
= /2, and for aft waves = 0.
c V cos
(15)
Vrel
(16)
c V cos
(17)
For seakeeping purposes, the assumption of deep water may be applied; in this case, taking into
account the dispersion relation (2), the encounter frequency can be finally derived as:
2
V
cos
g
(18)
The wave energy spectrum must also be modified according to the encounter frequency (it is
practically a Doppler shift of the spectrum). Since the energy content of a spectrum must be the
same for any observer, fixed or moving with the ship, the 0-th momentum must be the same:
m0
Se
(19)
Therefore, the relation between wave spectrum and encounter spectrum is the following:
Se e d e S d
(20)
2
In this paper, the term frequency will be used indifferently to identify both frequency f, measured in Hz, or angular
(circular) frequency , measured in rad/s.
233
1 2 V cos
Se
(21)
V
1 2
cos
g
3.2.
gd :
The ship response is usually described in terms of transfer functions (RAO, Response Amplitude
Operator), which give the normalised amplitude of the resulting ships motion for a sinusoidal
excitation of frequency e, the normalization factor being the wave amplitude 0 for linear motions,
2
the wave slope k 0 2 0 /
for angular motions and the wave acceleration
e 0 for
accelerations:
z0
RAO z e
(22)
0
RAO
(23)
a0
RAO a
(24)
2
e
z 0 exp i
exp i
exp i
(25)
(26)
(27)
yt
zt
L t
B t
(29)
Therefore, being the sum of harmonic motions (phasors), the vertical oscillation y t
represented by a harmonic oscillation:
yt
y 0 exp i e t
(30)
y
is also
and it is possible to define a RAO for the particular point of interest on the ship:
y0
RAO y e
(31)
0
In case of a random wave excitation, with the assumption that the response is a linear function of
wave amplitude and applying the superposition principle, vertical motion can be reconstructed as:
n
yt
y 0, j cos
e, j
(32)
y, j
j 1
where each oscillation amplitude y0,j is a function of frequency and amplitude of the j-th harmonic,
according to (31).
Furthermore, it is possible to demonstrate that the ships response energy spectrum is given by the
product of the square of the RAO and the wave energy spectrum. Thus, heave motions energy
spectrum is:
Sz
RAO 2z
(33)
and analogous equations hold for the other motions, while for any point on the ship the energy
spectrum associated to its wave- induced motion is:
Sy
3.3.
RAO 2y
(34)
Simulation results
The foundation for the commercial software package Seakeeper 3 , which was used to carry out the
computation of the yachts motions under different wave conditions, is the linear strip theory based
on the work of Salvesen [10], which is used to calculate the coupled heave and pitch response of the
vessel; the roll response is calculated using linear roll damping theory [11].
The main purpose of the kinematic model presented is to provide reasonable data about the
response of a generic yacht to different sea conditions, in order to have reliable information on the
motion which the SEAKERS device is subjected to. Since the project does not address a particular
yacht model, nor even a specific size of boat, there was no point in developing a focused in-house
software: hence the choice of adopting a commercial software that has a proven record of reliability,
using it to simulate the response of a yacht of adequate length included in the extensive library
provided.
The yachts model used in the numerical simulations is one of the library models that can be found
in Seakeepers library, since it has geometric and mass properties comparable to those of
commercial sail yachts of interest for the SEAKERS project.
The most relevant hydrostatic properties of this yacht are given in table 2. The generator considered
in the simulations presented is placed at bow on the longitudinal axis (B = 0) at a distance L =
5.17 m from the center of gravity.
Seakeeper is a software by Formation Design Systems Pty Ltd (trad ing as FormSys); website:
http://www.formsys.com/maxsurf/ msproducts/seakeeper.
235
Displacement
Volume (displaced)
Overall length
Draft amidships
Immersed depth
Waterline length
Max beam on waterline
Max section area
Waterplane area
Prismatic coefficient (Cp)
Block coefficient (Cb)
UoM
t
t
m
m
m
m
m
m2
m2
Fig. 4. Yachts mapped sections used to evaluate hydrodynamic coefficients in the equations of
motion by the Seakeeper software.
As illustrated in section 3.2, the ships response is defined by means of Response Amplitude
Operators (RAO) and phase shifts, with reference to a sinusoidal wave excitation. Figures 5 and 7
show values of RAO for each motion (heave, pitch and roll) for two different speeds (V = 5 knt and
V = 8 knt respectively), while figs. 6 and 8 show the phase shifts, as obtained by means of the
Seakeeper software. For the roll motion, the default value of non-dimensional damping factor
proposed by the software has been taken into account.
The response to random waves is illustrated in figs. 9 and 10 in terms of energy spectra of the
vertical oscillations (34). The significant oscillation amplitudes are obtained from these spectra in
the same way as the significant wave amplitude (10) is calculated from the wave spectrum:
y 0s
2 m0 y
Sy
(35)
Values of significant vertical oscillation amplitudes, corresponding to the energy spectra of figs. 9
and 10, are given in table 3.
236
237
238
239
240
Fig. 9. Energy spectrum of vertical oscillations at the generators location at speed V = 5 knt.
241
Fig. 10. Energy spectrum of vertical oscillations at the generators location at speed V = 8 knt.
Table 3. Yachts significant vertical oscillation amplitudes at the bow generators location.
y0s [m]
V = 5 knt
V = 8 knt
Hs [m]
= 90 deg
= 135 deg
= 180 deg
= 90 deg
= 135 deg
= 180 deg
0.5
0.433
0.383
0.322
0.405
0.366
0.318
1.0
0.721
0.684
0.688
0.680
0.697
0.732
1.5
0.980
0.943
0.967
0.937
0.964
1.024
2.0
1.147
1.127
1.175
1.120
1.152
1.233
242
4. Linear oscillator
4.1.
General remarks
The linear generator that will be used in the SEAKERS device to recover energy from the waveinduced motions of the yacht is analysed and approximated in this paper as a simple linear
mechanical oscillator, where the damping element represents a linear approximation of the effect of
the electromagnetic force exerted by the generator as it provides a voltage difference proportional to
the square of its relative velocity with respect to its basement, and the spring represents the stiffness
of the generators support. It is further assumed that the damping coefficient can be dynamically
varied depending on sea conditions: this could be achieved in the final system by means of a
variation of some electrical parameters in the associated circuit. The equation of motion is thus:
mx c x
K x
mg
Fs
(36)
where m is the generators mass, x its position in an inertial frame of reference, y is the basements
position, c is the damping coefficient, K the springs stiffness, g the acceleration of gravity, Fs a
static force provided by the support in order to balance the weight mg such that Fs = mg.
It is assumed here that the support can exert such a static force in order to balance the mass weight;
it can be seen that mechanical springs alone cannot play such a role, because the resulting stiffness
would be too high for the typical forcing frequencies. Indeed, if the spring were to counterbalance
the weight with a limited elongation at rest l = 0.05 m, the resulting natural frequency would be n
= g/l 2.2 Hz, which is much larger than the forcing frequency of sea waves: as the following
section explains, this would make the system too stiff, i.e. the mass would move rigidly with the
basement, with no relative motion between the two and, thus, no power extracted.
Equation (36) can thus be rewritten eliminating all static forces and introducing the relative position
s = x-y of the mass in a frame of reference moving with the basement:
m s cs Ks
my
(37)
which becomes the well-known second order ordinary differential equation for an oscillating body:
s
2
n
(38)
K m
(39)
4.2.
c 2m
(40)
The steady-state response of the linear mechanical system to a harmonic forcing of the type
yt
y 0 exp i e t is itself harmonic:
st
exp i
y0
n2
(41)
0
given by:
n2
1 i2 n
(42)
The magnitude of 0 gives the amplitude s0 of the harmonic motion of the generator (fig. 11, top),
and its ratio with the ships oscillation amplitude is the response amplitude operator for the
generators relative motion:
s0
y0
RAO s
n,
n2
e
1 n
2 2
(44)
2 n
gives the phase of the generators motion with respect to the forcing oscillation
(45)
Fig. 11. Frequency response of the harmonic oscillator: amplitude (top) and phase (bottom).
244
In this model, damping the oscillations in the linear mechanical system provides the mean to extract
energy from the wave excitation; thus, it is interesting to identify optimal values for the damping
coefficient c in order to extract the maximum power. The power absorbed is given by:
cs 2
Pt
(47)
and its average value over one cycle (which will be indicated as ) is:
1 T
1 2 2
P t dt
c e s0
(48)
T 0
2
the above expression, taking into account (40) and (44), becomes:
n3
3 2
e 0
1 n2
2 n
(49)
Since the oscillations are constrained by the size of the generator, two different scenarios must be
considered. In the first one, let us imagine that the undamped oscillations do not reach the
maximum range allowed smax : in this case, increasing the damping coefficient from 0 initially yields
higher values of even though s0 decreases according to (44), until a maximum for is reached,
beyond which it decreases. The optimum value of can thus be found when d /d = 0:
opt
2
11 n
2 n
(50)
with corresponding optimum damping coefficient, maximum average power and oscillation
amplitude given by:
1
(51)
c opt m 1
n2
1
m
4
max
s0
y0
3
e
1
y 1
n2
2
0
1
n2
(52)
(53)
In the second case, the undamped oscillations would be larger than the maximum allowed range
smax : then it is possible to extract more power by increasing the damping coefficient while the
oscillation amplitude y0 remains at its maximum permissible level smax . It is possible to show that
the maximum power is obtained when the damping coefficient is such that the oscillation given by
(44) is exactly equal to the stroke (s0 = smax ):
opt
c opt
n
2
y0
s max
y0
s max
1
n2
(54)
1
n2
(55)
1
m
4
3 2
e max
y0
s max
1
1
n2
(56)
245
It is possible to find out which wave excitations make the system reach its maximum stroke smax by
setting the oscillation amplitude given by (53) equal to smax , yielding:
y0
s max
21
1
n2
(58)
Thus, for wave amplitudes originating boat oscillations lower than the limit set by the above
equation, the system oscillates freely and equations (51)-(53) apply, while for higher waves more
damping, and thus more power, is available, in order to constrain the system within the maximum
stroke allowed, and (55)-(56) apply.
It is worth to point out that in both cases the optimum value for the damping coefficient is directly
proportional to the oscillators mass and to the forcing frequency: the average power absorbed is
therefore proportional to mass m and to the third power of forcing frequency ( e3 ), as shown by
(52) and (56).
In particular, the linear dependence on the oscillating mass m is, on the one hand, almost obvious
because energy recovery depends on inertia and kinetic energy, but on the other hand it is an
important property to be taken into account because it allows to design, test and prototype modular
systems of relatively low mass, with the overall power extracted given by the sum of power
available from different modules. For this reason, the results discussed in section 4.4 will be given
with reference to a unit mass m = 1 kg.
4.3.
The frequency response of the harmonic oscillator can also be used when the external forcing is not
harmonic (as in the case of real wave excitation): if Sy( e) is the energy spectrum associated to the
vertical oscillations of a particular point of interest, which can be evaluated from the wave energy
spectrum by means of (34), then the energy spectrum associated to the relative motion s of a linear
system such as the one described in the previous section is given by:
Ss
RAO s
2
e
Sy
(59)
In the following considerations the dependence of relative motion and its spectrum on natural
frequency n will be implicitly assumed, so that Ss ( , e) Ss ( , n , e).
The spectrum of relative motion allows the evaluation of the significant oscillation amplitude as
follows:
s0 s
2 m0 s
Ss
(60)
Since power generation depends on the square of the generators velocity (47), the energy spectrum
related to relative velocity must be introduced. This is simply given by:
Ss
2
e
Ss
2
e
RAO s
2
e
Sy
(61)
The 0-th moment of the velocity spectrum gives velocitys root mean square (RMS), which is
related to average power generation:
s RMS
m0s
Ss
(62)
246
c
s RMS
m
RMS
(63)
n RMS
As in the case of sinusoidal waves, for a given natural frequency the generators motion depends on
the choice of the damping coefficient , for which an optimum value is found by maximizing power
output (63) with the constraint that the significant oscillation amplitude is lower than the maximum
allowed range smax (for this non- linear optimization procedure the MATLAB function fmincon
has been used):
d
RMS
0 with s 0s
opt
(64)
s max
opt
4.4.
Simulation results
In this section, results obtained with the mechanical model of the linear generator subject to random
wave excitations for two different speeds (V = 5 knt and V = 8 knt) are discussed. As detailed in
section 3.3, the yacht's oscillation is described by energy spectra represented in figs. 9 and 10.
The maximum stroke taken into consideration in the simulations is smax = 0.5 m, since this value is
about the highest possible on sail yachts of length from 10 to 14 m, which are the main target for
the SEAKERS project. The natural frequency n is taken as 0.25 Hz, in order to make the
mechanical system almost resonant with most sea conditions that may be encountered.
Figures 12 and 13 show the energy spectra associated to the generators relative motion and RMS
power generation for different speeds and directions. Power generation values are also given in
table 4, while table 5 gives calculated values of significant oscillation amplitudes.
From table 5 it is possible to see that for wave heights higher than 0.5 m, the generators oscillation
is always limited to the maximum range smax : in order to analyze the results, it is thus useful to
consider the simplified equation for average power generation (57), which shows that, if the
mechanical system is tuned to the forcing wave condition (n 1), average power is proportional
to the third power of the encounter frequency, and to the product of boats vertical oscillation y0 and
maximum range smax .
Therefore, an increase in significant wave height gives rise to two opposite effects on power
generation: on the one hand it increases due to its dependence on y0 , but on the other hand the wave
energy spectrum shifts towards lower frequencies (fig. 1), resulting in a decrease in power
generation. Clearly, this gives rise to a maximum power generation for a particular sea state, that
under the assumptions taken into account in this paper correspond to a significant wave height of
1.5 m, as figs. 12 and 13, along with table 4 show.
In other words, even with high values of significant wave height, which correspond to rather low
values of peak frequencies (table 1), if the full spectrum is taken into account significant
contributions to the excitation can be found also at frequencies higher than the peak one, and these
contributions increase average velocities and, consequently, power generation. Nonetheless, it is
still possible to find that increasing wave heights beyond a certain threshold decreases the power
output, because in this case significant contributions can indeed be found only at low frequencies.
Tables 6 and 7 report values of optimum damping coefficients and damping ratios as defined by the
optimization procedure (64).
247
[W/kg]
V = 8 knt
= 135 deg
0.352
0.644
0.852
0.757
= 180 deg
0.230
0.630
0.848
0.789
= 180 deg
0.263
0.500
0.500
0.500
Hs [m]
0.5
1.0
1.5
2.0
= 90 deg
0.552
0.617
0.753
0.664
V = 5 knt
= 135 deg
0.401
0.620
0.807
0.715
= 180 deg
0.247
0.546
0.721
0.656
= 90 deg
0.500
0.572
0.704
0.619
= 90 deg
1.522
2.410
3.096
3.107
V = 5 knt
= 135 deg
2.709
2.729
3.665
3.732
= 180 deg
2.724
2.816
3.851
4.017
= 90 deg
1.469
2.232
2.894
2.934
V = 8 knt
= 135 deg
3.142
3.036
4.139
4.223
= 180 deg
3.228
3.413
4.725
4.922
= 90 deg
0.468
0.710
0.921
0.934
V = 8 knt
= 135 deg
1.000
0.967
1.317
1.344
= 180 deg
1.027
1.086
1.504
1.567
Hs [m]
0.5
1.0
1.5
2.0
= 90 deg
0.484
0.767
0.986
0.989
V = 5 knt
= 135 deg
0.862
0.869
1.167
1.188
= 180 deg
0.867
0.896
1.226
1.279
248
Fig. 12. Energy spectra associated to the linear mechanical systems motion at speed V = 5 knt.
249
Fig. 13. Energy spectra associated to the linear mechanical systems motion at speed V = 8 knt.
250
5. Conclusions
This paper has provided an overview of the results obtained with the kinematic and mechanical
models of the yachts response to different wave excitations, and of the linear generator taken as a
simple mechanical linear system which extracts power from the wave- induced motion by means of
an ideal linear damping.
These preliminary evaluations, even though based on a rather simplified model of the generator,
have produced some important insights on system dynamics and on the range of values to be
assigned to several significant parameters, such as mechanical stiffness and damping ratio.
In particular, given the particular range of forcing frequencies, the mechanical stiffness must be
chosen so as to obtain a natural frequency within the range of most forcing frequencies: a value of
0.25 Hz has been considered in this paper, with a resulting stiffness around 2.5 N/(m kg).
The choice of damping ratios is based on the maximisation of power output for a given natural
frequency for different wave excitation conditions. Values of damping coefficients in the range 1.55.0 N s/(m kg) have been found, and this result will be useful in the definition of the electric
circuits physical parameters associated to the linear generator.
Power generation of up to 0.85 W/kg have been obtained in the most favourable sea conditions, and
anyway values higher than 0.5 W/kg are available in most cases, which represent an interesting
result for this particular application. Indeed, a total weight for the SEAKERS device of up to 200 kg
can be considered acceptable on sail yachts with length in the range 12-14 m, especially in the case
of a first equipment (as opposed to retrofitting an already existing yacht, because in this case many
more design constraints should be addressed). With an optimized design, it is conceivable from
preliminary evaluations that up to 50% of this weight (100 kg) could be allocated to the oscillating
masses; in this case, an average power generation of almost 100 W is feasible, which could make
possible to recover at least 1 kWh at the end of a day- long cruise. This amount of energy generation
could indeed be interesting for this particular application.
Obviously, the issue related to the influence of this moving mass on sailing performance should be
addressed: but specific calculations, which have not been reported here for brevitys sake, show that
the inertial forces generated by the mass motion are at least two orders of magnitude lower than the
forces exerted by the sea on the boat. After all, this must be the case because the energy absorbed
by the linear system is but a small fraction of the total energy of the incoming waves. Therefore, in
all probability the impact of the added mass due to the generator on sailing performance can be
safely deemed negligible.
It should be observed that the concept of power availability (or availability factor) routinely
used to appreciate the performance of a system based on renewable energy, is much less useful for
this particular application, because the final goal is not the generation of electricity per se on a
continuous basis but, rather, only on the particular occasion when the yacht is used for a cruise. An
availability factor should therefore be considered only with reference to a single cruise, and it would
take into account the time frequency of encountered wave height during a typical cruise of a sailing
boat, because this is the main parameter influencing average power output. However, it is rather
difficult to make any prediction about this probability, other than saying that sail cruises are most
common when wind (and, consequently, wave) conditions are not extreme (i.e. with moderate
winds and waves, while it is safe to assume that calm or stormy seas are avoided); furthermore, a
leisure cruise usually requires stable weather conditions. In the end, it is reasonable to infer that in
most cases sea conditions encountered on a leisure cruise are reasonably constant and marked by a
significant wave height within the range 0.5-1.5 m. Under these assumptions, average power
generation is expected to be almost constant during the whole cruise.
The next activities in the SEAKERS project will be focused on the implementation of a non- linear
electro-mechanical model of the generator, in order to define suitable ranges for the most important
251
electrical parameters in the system, and to narrow down the set of possible runs of the more detailed
2D and 3D models that have been set up.
Acknowledgments
The research leading to these results has received funding from the European Union Seventh
Framework Programme (FP7/2007-2011) under grant agreement n 262591. The authors would
also like to acknowledge the invaluable effort provided by all the staff of Labor srl, the projects
coordinator, as well as the contribution of the SMEs and RTD performers involved in the project.
References
[1] Ivanova I.A., Bernhoff H., gren O., Leijon M., Simulated generator for wave energy
extracction in deep water. Ocean Engineering 2005;32(14-15):1664-1678.
[2] Thorburn K., Leijon M., Farm size comparison with analytical model of linear generator wave
energy converters. Ocean Engineering 2007;34(5-6):908-916.
[3] Eriksson M., Isberg J., Leijon M., Hydrodynamic modelling of a direct drive wave energy
converter. International Journal of Engineering Science 2005;43(17-18):1377-1387.
[4] Leijon M., Danielsson O., Eriksson M., Thorburn K., Bernhoff H., Isberg J., Sundberg J.,
Ivanova I., Sjstedt E., gren O., Karlsson K.E., Wolfbrandt A., An electrical approach to
wave energy conversion. Renewable Energy 2006;31(9):1309-1319.
[5] Bertram V., Practical ship hydrodynamics, Oxford, UK: Butterworth-Heinemann; 1999.
[6] Biran A., Ship hydrostatics and stability, Oxford, UK: Butterworth-Heinemann; 2003.
[7] Jensen J.J., Load and global response of ships, Oxford, UK: Elsevier; 2001.
[8] Rawson K.J., Tupper E.C., Basic ship theory, Oxford, UK: Butterworth-Heinemann; 2001.
[9] Perez N.A., Sanguinetti C.F.O., Scale model tests of a fishing vessel in roll motion parametric
resonance. Sntesis Tecnolgica 2006;3(1):33-37. ISSN 0718-025X. Available at
<http://mingaonline.uach.cl/scielo.php?pid=S0718-025X2006000100004&script=sci_arttext>.
[10] Salvesen N., Tuck O.E., Faltinsen O., Ship motions and sea loads. Transactions, Society of
Naval Architects and Marine Engineers 1970;78:250-287.
[11] Formation Design Systems Ltd. Seakeeper Manual Available at:
<www.formsys.com/extras/FDS/webhelp/seakeeper/skmanual.htm> [accessed 3.2.2012].
252
Abstract:
This paper critically reviews the state of the art of an approach to supply energy to earth from space mirrors
that would be placed in orbit with angle control to reflect solar radiation to specific sites on earth for
illumination. These mirrors would be of t he order of a square kilometer or more, planned to be made of thin
plastic reflective films, which are launched to some optimal orbit around the Earth. One could, for example,
thereby provide night or emergency illumination to cities and other locations, or illuminate agricultural
production areas to lengthen the growing season, or to illuminate photovoltaic or thermal collectors on earth
for producing electricity or heat. Proposals were also made for using such mirrors for weather modification,
and we added here the possibility of using t he spac e mirrors for shading the earth to reduce global warming.
Experiments with spac e mirrors were conducted in the past by the former Soviet Union. In addition, thin film
aluminized Kapton mirrors were manufactured and optically an mechanically tested to examine their property
changes when exposed to a cryogenic temperature, an economic analysis related to several applications
was performed, and leading issues that must be taken int o account in the sustainability analysis of the
concept were described. Our experiments with thin film mirror have shown that the reflectance at 77K is
always higher (about 1.7 fold) than that at 300K, and the ultimate tensile strength and modulus increased at
the cryogenic temperature. Without (yet) consideration of environmental and social impact externalities, our
economic analysis agrees with past studies that if transportation costs to mirror orbit are reduced to a few
hundred $/kg, as planned, the use of orbiting space mirrors for providing energy to earth is an investment
with a good rate of return and a cost effective alternative to other power sources. This energy concept is very
appealing relative to other options for addressing the severe energy and global warming problems that we
face, and des erves much and urgent R&D attention.
Keywords:
Space mirrors, Space energy, Solar mirrors, Mirrors, Cryogenic mirrors, Illumination
typically of the order of a square kilometer or more each, highly reflective, planned to be made of
thin plastic reflective films to minimize weight and cost, mounted in an appropriate light frame.
They would be launched to one or more orbits around the earth. One could, for example, thereby
provide night illumination to cities and other locations and for emergency lighting, or provide
sunshine for agricultural production in some areas to enable or lengthen the growing season, or for
applications such as crop drying ad water desalination, or to illuminate photovoltaic (PV) and
thermal collectors (including salt-gradient solar ponds) on earth for producing electricity or for
heating. This approach was originally briefly proposed by the space science pioneer H. Oberth in
1928 [15]. This concept seems to have laid dormant and was then advanced most notably by
Buckingham and Watson (in 1968, 60 years later [16]), NASA (Billman, Gilbreath and Bowen)
[17-19], Ehricke [20-22], and others [23-25]. The review portion in this paper relies strongly on
[16-21] in recognition of the pioneering work of these authors.
When considering space power generation, the major advantages of the space mirrors approach are:
(1) instead of PV collectors on the energy source spacecraft there are only mirrors (optical
reflectors), planned to be made from very thin film coated polymers (microns thick); (2) the energy
is transmitted directly to earth in the form of solar light, without need for conversion of the
collected solar energy to microwave or laser beams and their transmission through the atmosphere
to earth; (3) sunlight is less threatening environmentally than the transmission of microwave or laser
radiation; (4) no requirement for power management and distribution or thermal management
systems on the spacecraft; (5) constructed of light thin ( m order) films, mirrors are easier to bring
to orbit and deploy than the equivalent PV cells; (6) if used for power generation, it would probably
need smaller collector and energy conversion fields on earth because of the safety-dictated need to
make microwave beams diffuse when PV satellites are used; (7) no need for technical energy
conversion systems on earth when the reflected sunlight is used for lighting, agriculture or bioenrichment. There are, however, also a number of significant technical challenges: (1) the reflected
sunlight arriving at the earth surface is more subject to the effects of weather, such as overcast, haze
and atmospheric refraction, than microwave or laser beams; (2) amounts of sunshine reflected to
earth that are sufficient to help supply significant fractions of needed global energy, and to be
commercially viable, would require very large (order of 1 km2 or more) mirrors, that must be
optically flat (to a fraction of a wavelength of light) over these huge areas, and durable both
mechanically and optically; (3) environmental effects, such as associated glitter and other light
pollution. These challenges have contributed to the fact that significantly less has been done so far,
or planned to be done, on space mirrors, when compared with PV solar satellites.
A significant albeit brief step in the development of space mirrors was the Russian Space Mirror
Project Znamya (banner) developed by the Space Regatta Consortium (SRC) [24] established
in 1990 by the Russian space agency and the corporation Energia [26] (which specializes in space
and launch vehicles and rocket boosters). The purpose of SRC in project Znamya, according to their
official website, was the development of thin sheet technologies for solar reflection and solar sails
and then for illuminating high latitude earth regions during winter months.
Detailed information about the Znamya experiments (Fig. 1) is somewhat sparse [24,26-28] , and
the following is available. The first SRC to be tested in space was Znamya-2, on February 4,
1993. The mirror was a 20 m diameter circular 5 m-thick aluminized PETF (Mylar) film, with an
estimated aerial density of 22 g/cm2 that was composed of 8 sections with radial gaps between
them. It was installed together with the unfolding mechanism inside the docking compartment of
the cargo space vehicle Progress M-15 which disengaged from the MIR space station. The
deployment test was successful. According to the SRC website, the spot of light produced by the
mirror was about 5 km in diameter and moved across the earths surface (starting in France and
through Eastern Europe and Asia) at a speed of around 8 km/s. The brightness of the mirror as seen
from the earth was reported to have been similar to that of a single full moon (<1 lux).
Znamya 2.5 was the second attempt to launch a space mirror, as a continuation of SRC's space
254
reflector experiments that hopefully will lead to the deployment of 200- m-diameter reflectors. The
reflector was 25- m-diameter and was constructed of materials and design similar to Znamya 2.
Deployment of Znamya 2.5 was attempted on February 4, 1999. As the sail unfurled it collided with
and wrapped around the docking antenna, and the whole apparatus crashed into the ocean. Since
then, there have been no attempts to launch a solar mirror. The Znamya experiments received much
attention from the media, including criticisms about light pollution that such space mirrors may
create [29-31].
a.
b.
Fig. 2 The illuminated ground area as a function of the space mirror orbit altitude h [17].
For mirrors of diameters considered in the studies so far, the illumination Ie at the earth's surface
by a space reflector in terms of solar illuminance Is. reflectance , cloudiness factor C (C = 1 for
cloudless sky), reflecting area Ar (of single or cluster of reflectors), angle of incidence between the
255
reflected beam and the illuminated earth spot, and angle at the reflector between the incident and
reflected light beam, is given by
Ie
A
(1)
C r f ( ) cos cos
Is
Ae
2
where f ( ) is a function representing the intensity extinction, due to haze and zenith distance (that
increases the beams path length through the atmosphere) and is the elevation angle of the
reflector above the horizon; for = 00 , f ( ) =1.
Equation (1) shows that he earth spot illumination intensity (Ie) increases in proportion to the
reflector area (Ar).
If the solar reflector is above the atmosphere, as typically planned for such space mirrors, the solar
radiation intensity at the reflector is at the atmospheres edge, Isc the Luminous Solar Constant is
133,334 lx or 134,108 lx and the Solar Constant = 1,366.1 kW/m. If it is within the atmosphere,
the illuminance Is at the reflector is diminished by effects of air molecules, dust and water vapor
along the beam path, with this diminution expressed by Co , the overall coefficient of absorption ad
reflection in a cloudless atmosphere, so
Is
Co
(2)
I sc
The image area on the earth of an orbiting mirror of area Ar positioned at a height h above that
image is expressed by
Ae
Ae
Ar
(3)
2
where is the angular subtense of the Sun, = 1.39xl06 km/l.5xl08 km = 9.27 mrad, h = r - rearth , r
is the radial distance between the orbiting mirror and the center of the earth, and rearth is the earth
radius. The area illuminated on the ground is an ellipse with major axis (Dm + h cos ) and minor
axis (Dm + h). Equation (3) shows that the illuminated earth area becomes gradually independent
of the reflector area as the orbit altitude increases.
Very significantly, the overall reflectance of the mirror depends not only the surface specular
reflectance but also on its flatness to within a small fraction of the sun's angular diameter a. This
means that all parts of the mirror's surface must point in the same direction to within 1 or 2 mrad.
The reflector diameter influences the sharpness of the image. For a reflector of diameter Dr,
the earth spot image will have a penumbra region of shadow of the same diameter, which thus does
not practically affect the spot size.
For a synchronous orbit of h = 22,400 miles (36,049 km), the diameter of the illuminated spot on
earth is then 208 miles (~335 km). Obviously, if a smaller illumination area is needed the satellite
can be placed in lower orbits but then, as shown by Eq. (4), the illumination will take place for
shorter periods of time. This can be remedied by using a number of mirrors in the same or similar
orbits.
The period of a satellite (T) and the mean distance from the earth (h) are related by the
equation: T
2
h3
3600 K
1/2
(4)
where K GM earth , G is the Universal Gravitational Constant, G = 6.673 x 10-11 N m2 /kg2 , and
Mearth = 5.9742 x 1024 kg, so K = 398,659 km2 /s2 .
256
0o
REFLECTOR SATELLITES
45o
SUN'S
RAYS
NORTH
90o
POLE
SYNCHRONOUS
ORBIT
180o
Fig. 3 Buckingham and Watsons basic concept of a reflector satellite and structure for supporting
the thin-film mirror [16].
A very comprehensive conceptual, technical and socio economic study and exposition of space
mirrors as conducted by the space visionary Krafft Ehricke [20-22]. He proposed and analyzed in
some detail a number of generic applications for providing lunar-type night illumination service
(Lunetta). solar type light energy services (Soletta), insolation for bio-production enhancement
(Biosoletta) to produce food and biomass, insolation for agricultural weather stabilization,
precipitation management, crop drying and desalination (Agrisoletta), and insolation for
generating electricity on earth (Powersoletta). He made an economic feasibility study and
predicted that very competitive electricity generation costs can be obtained; for example he
predicted that Powersoletta with a PV energy conversion on earth can produce electricity at
4.8c/kWh.
He added a number of new concepts [21] beyond past considerations:
use of a variety of sub-geosynchronous orbits, particularly, sun-synchronous ones,
splitting" of large single reflectors into a number of smaller reflectors operating in clusters and,
to reduce the size of the individual reflector, lower cost, increase system robustness. The
illumination pattern in this configuration is determined by the number of co-orbits, the time
position of their maximum latitude passage, and the lighting power (number of reflector units)
assigned to each co-orbit.
257
splitting of one orbit into several co-orbits (Fig. 4) which is particularly advantageous for urban
night illumination where multi-directional illumination creates a more diffuse and uniform
lighting effect;
More possible applications;
the concept of retro-reflection (called by some others relay mirrors), i.e. reflecting light from a
mirror that does not have direct optical sight line to an area on earth that needs to be illuminated,
to one in orbit that does (Fig. 5; thus enabling day and night operation raises the system's
utilization factor.
Fig. 4 The NASA SOLARES multiple orbiting mirror concept. Note 2 co-orbits and that several
mirrors exposed to the sun at the same time are reflecting to the same earth spot [19].
Fig. 5. Retro-reflection technique for daytime use of space light (adapted from [21]). Primary
reflectors (marked 1) are those that face the Sun and reflect the light to secondary (2) reflectors
that beam to the ground service area. Reflectors numbered with a prime beam solar radiation
directly to the ground service area without retro-reflection.
We add, without further analysis, the possibility of using the space mirrors for shading the earth, a
possibly useful application that may be locally useful for a number of obvious reasons, or as a geoengineering way to reduce global warming, a concern that did not exist when the early researchers
did their studies of solar mirrors in the 1970-s. In the extreme case, the mirrors, or some of them,
could be turned towards the sun and thus prevent the solar radiation reaching the atmosphere and
earth area that are that is now in the mirrors shadow.
258
NASA performed some detailed preliminary feasibility studies of the design, deployment and use of
space mirrors, published in the late 1970-s [17-19] 1 , and concluded that The use of orbiting mirrors
for providing energy to ground conversion stations to produce electrical power is shown to be a
viable, cost effective and environmentally sound alternative to satellite solar power stations and
conventional power sources.
Their proposal, which they called SOLARES, was to use a cluster of free-flying very lightweight
(10 g/m2 ) metal-coated polymeric film mirrors, optimally 1 km2 each which, after deployment at
altitude of 800 km, are placed in operational orbit and controlled by solar radiation pressure, to
almost continuously illuminate a chosen surface on earth an intensity of Ie ~ 1kW/m2 (at a fairly
constant level, which, however, must take into account atmospheric variability with time). This
would increase the available insolation at the earth energy collection and conversion station about
4-fold, and, if the insolation is uniform enough over time would also eliminate or reduce the need
for energy storage.
They developed equations showing the influence of a number of parameters - mirror altitude, orbit
inclination, period, mirror size and number, and atmospheric effects - on the reflected insolation
that may be received by a round spot as a function of location. In their economic analysis they
found that generated electricity costs range may be as low as about 1.6 c/kWh (in 1977/8 US cents).
At the same time, as discussed in more detail in Section 7 below, they used extremely low costs for
transportation into orbit, which make the costs of electricity and heat generated they determined
much too low when considering current technology. As the environmental issues of principal
concern they identified the perpetual twilight that neighboring communities might experience and
the land area required, and felt that atmospheric effects are minimal and to their opinion perhaps
beneficial. More details about their economic and environmental study can be found in Sections 7
and 8. They expressed the opinion that SOLARES could supply the entire global energy
requirement.
Other authors have proposed mirror deployment at geostationary orbits (GEO), but Eq. (3) shows
that at this altitude of h = 35,800 km the area illuminated on earth would have the huge diameter of
about 3,329 km. At the chosen ground intensity of 1 kW/m2 the mirror area would then have to have
an area of about 150,000 km2 . The annual energy generated at one such location with 15% ground
conversion efficiency would be, if atmospheric solar radiation transmission effects are ignored, up
to about 41,200 EJ, 82 fold of the current world usage of 500 EJ. To achieve a practical ground area
size with realistic capital investment and energy output, to provide energy to more than a single
ground station, and to be able to employ the enhanced insolation for nonelectrical applications if
desired, they postulated the use of a large number of flat1-km diameter reflectors in lower orbits
(Fig. 4). Such configuration would allow each selected ground site could be insolated at all times
(excluding eclipse and inclement periods), and any given mirror could be used for other tasks,
including the insolation of other sites. The use of many and small reflectors clearly also allows the
desirable feature capability of incremental implementation and easier repair and replacement.
Smaller reflector areas also require much lower torque for their control, since their moment of
inertia scales as Ii ~ ARi2 t where is the average areal mass density, A is the mirror area and Ri is
the characteristic radial dimension along the ith rotational axis.
These NASA studies also calculated the daily and annual variation in the solar flux, both the natural
one and that supplied by the orbital mirror system, and it is shown in Fig. 6. The most impressive
feature is that although the direct solar input varies seasonally by more than a factor of two, the
mirror input is constant to about 10%, making the system suitable for baseload electricity
generation use.
It is noteworthy that while these NASA publications all had the same objective and focus, the assumptions and results
changed with the publication time, probably indicating an evolution of the concept.
259
Thorough techno-economic analysis is required to find the optimal system, so they only considered
an example of the mirrors at an altitude of 4,146 km in a 3-hr periodic orbit with inclination of 45
relative to the equatorial plane (or several inclined orbits separated by latitude). The mirrors would
be deployed or erected at an altitude of approximately 800 km. From this altitude, where the solar
radiation pressure is much larger than the drag force, it is possible to "solar sail the mirrors to their
operational orbit (i.e. 4,146 km), requiring about a 3- mo transit time.
Fig. 6 The annual variation of the solar flux at the area illuminated on earth by the sun alone, by
the proposed mirror system, and by their combination [19]
Using Eqs. (1) and (3), and as best it can be concluded from [17-19] assuming 23% losses and
geometric spreading due to the sun- mirror-site angle, eclipsing, non- zenith mean apparent reflector
location and atmospheric effects, 62,800 km2 of mirror area was stated to be required to deliver an
average 1.25 kW/m2 (0.25 kW/m2 from regular solar incidence + 1 kW/m2 from the mirrors) to
ground stations at 30 latitude. With their proposal to build individual mirrors of 1 km diameter
each, this translates to the need for 80,000 orbiting mirrors having a total mass of about 6.3 l08 kg.
This mirror system was estimated to be able to supply this flux to at least 5 (of a theoretical 13)
ground sites around the world. For each site, about 70% of the incident insolation falls within area
of diameter
(38.4 km, Ae = 1,167 km2 ) and 99% within 2
(76.8 km, Ae = 2,334 km2 ). We note
that a conclusion from [19] appears to be that the ratio of the mirror area and the ground area
steadily insolated with the added 1 kW/m2 from the mirrors is 62,800/(5x1,167) = 10.76. With 15%
conversion of just the
insolation to the five ground stations, up to 27.6 EJ of electricity can be
generated, amounting to about the electricity generated globally in 1977 (at the time of the NASA
studies) and 36% of the electricity generated in 2010. After deducting the energy converted in the
PV system, the remaining 85% of the energy at , as well as the energy in the annulus between
and 2 could additionally constitute a large usable energy resource.
260
Based on such a concept, the NASA [20] study proposed that the additional insolation from the
space mirrors can be significantly greater than average ambient sunlight, and they have chosen to
supply an earth surface and additional solar intensity of Ie ~ 1kW/m2 (total of ~1.25 kW/m2
including the natural, unreflected, insolation) for sizing the space system and ground stations and
for deriving costs. They point out that the maximal average U.S. value of normal sunlight is about
0.25 kW/m2 , and therefore this increase intensity should reduce the area-related terrestrial solar
converter system costs (for collectors, converters, land preparation, etc.) fivefold [(1 + 0.25)/0.25 5] from that of a solar power generation system of equal output that operates without the space
mirrors.
The NASA study [18-20] has shown that for a given orbital inclination the number of mirrors
needed to provide continuous insolation at a given ground site increases with decreasing altitude,
and thus the total mirror area for a fixed ground site intensity decreases. However, several factors
place a limit on the lowest usable altitude. First, atmospheric drag necessitates an altitude above
1,500 km for the 15 g/m2 structure they proposed to allow a system life of 30 years. A remedy is to
employ solar sailing for countering drag, thus perhaps providing the desire system life down to
altitudes of 1,000 km. Second, the angular acceleration needed for the mirror to insolate a given
spot during its transit varies approximately inversely with the third power of the altitude, thus
creating significantly tougher demands on structural characteristics and control at lower altitudes.
Third, lower orbits increase the fraction of time the mirror is eclipsed by the Earth. They thus
conclude that the lower bound for an operational reflector system is probably not less than 1000 km.
particles originating from the sun and propelled by the Earths magnetic field. The composition of
solar wind includes approximately 2x108 H+ /cm2 s protons (96%), of about 6x106 alpha
particles/cm2 s (3-4%) and a few 105 ions/cm2 s of higher mass with average velocities of about 400
km/h (corresponding to energies of 0.85, 3.4, and 10 keV, respectively) as well as 3-30 keV
electrons. There are also some occasionally emitted particles as the result of solar flares or storms.
Studies in [35] have shown a rapid worsening in the optical properties, including loss of reflectivity
and defocusing due to blistering, due to the effects of solar wind, affecting the very beginning of a
solar mirrors operation. They suggest the mirrors film thickness should be at least 0.1 m, which
is the maximum penetration depth of the solar particles. As discussed below, this is also the same
minimal thickness required for the metallic thin film to remain opaque to visible light and to other
low frequencies of electromagnetic radiation. Another type of radiation is the ultraviolet (UV) part
of sunlight with wavelengths between 4 and 400nm. Two UV bands are particularly relevant to
materials degradation, the near UV range (200 400nm) and vacuum UV range (100 200nm)
[36]. This type of radiation causes the greatest material degradation to polymers and most of this
damage is sustained by the first 0.3 microns from the surface.
A commonly used mirror substrate is polyethyleneglycolterepthalat (PETP) with the net
composition (C10 H8O4 )n , known as Mylar, Hostaphan, etc., and are produced by companies like
Bayer, and Du Pont, and Kapton, which is poly(4,4'-oxydiphenylene-pyromellitimide) made by
DuPont. They are coated with a thin film metallic surface by chemical vapor deposition (CVD) to
provide high reflection. The most common metal is aluminum due to its good reflectance and low
cost, but gold and silver were also used for small mirrors due to their stability or higher reflectance.
It is generally agreed that the film mirrors should be maintained periodically, say once in 10 years,
by applying a thin fresh Al (or other reflective material) layer in situ, which could be accomplished
by flying a furnace with evaporating Al along the foil surface at a certain distance [34]. This was
expected to restore the initial reflectivity (smoothing of the blistered flaked areas), and sintering
together the eventual brittle foil surface areas (flakes) by the freshly evaporated material. to
prolong the mirror reflectivity to the order of a 100 years. Such maintenance would also be needed
to repair possible holes due to meteorite impact, or for replacement of part or all of the reflector
film. It was estimated, however, that meteoroid damage would be very small, 3% for 30 years in
orbit [17].
The solar mirror must have mechanical properties that can withstand its temperature in space, that
goes down to 3K (with the associated embrittlement) but also rises to much higher values especially
on its non-reflecting parts which are intermittently exposed to the sun. The mirrored surface
substrate has the leading role in the mirrors overall structural integrity.
NASA. The size of the available vacuum evaporator dictated the diameter of the mirror film to be
134.6 mm. The pure aluminum coating was done by vapor deposition, and the manufactured film
mirror supported by a ring structure is shown In Fig. 7.
300K
77K
Our film
mirror
Reference
film mirror
Fig. 8. 500X microscope surface photos of our thin film mirror and of the reference one, at 300K
and 77K. Note: the bean-shaped dark area in the center of each photograph is a smudge on the
microscope lens, not a defect.
Figure 8 shows that the number of defects/imperfections per unit area (i.e. surface defect density) is
higher for our film mirror than for the reference one. These defects include mostly pitting for the
samples and spits in the standard, typical of using a vapor deposition. This may therefore be the
reason for the measured lower reflectance of our mirror. Since the vapor deposition process we
used can be improved significantly, there is no doubt that better reflectance can be obtained, as
shown by many.
263
An interesting observation is that for both films the imperfection surface density is lower for the
cryogenic temperature, perhaps thus justifying the higher reflectance at this temperature. Another
interesting observation is that nevertheless, the reference film experience significantly higher
degradation when exposed to the low temperature, as shown by lines the along the vertical direction
of the image that appear to be micro-cracks that formed after exposure to liquid nitrogen (~77 K).
In fact, when the reference film was placed into the liquid nitrogen it made a loud crackling noise
that lasted for a few seconds, while no sound was heard for our film. In contrast, the sample did not
exhibit any sounds at all when placed inside the container. No micro-cracks developed in our film.
The likely explanation is that the thermal expansion coefficients of silver and Mylar (the reference
film) differ by 6-fold, while those between Aluminum and Kapton (our film) differ by only ~15%.
Table 2 below shows the experimental values for ultimate tensile strength and Youngs modulus
from our mechanical tests.
Table 2: Measured Mechanical Properties for our film mirrors exposed to 300K and 77K
Film temperature, K Ultimate Tensile Strength, MPa
Young's Modulus, MPa
300
9.9
129
77
13.3
448
The observed trends are consistent with the behavior of thermoset polymers, like polyamides, which
are characterized by a high tensile strength and modulus with a small total elongation, prior to
failure, brittle behavior, and an increase of the ultimate tensile strength and tensile modulus at lower
temperatures.
264
Fig. 9 Proposed structural configuration of a space mirror and its flywheel control system [17,19].
The design draws on interest in constructing and using solar sailing for transporting payloads in
deep space exploration, in one case of which an ultralight truss mast that can be deployed to
kilometer lengths was designed, and a means to fabricate a large quadrant of sail material and stow
it without material creasing or trapped air were proposed [28,38,39]. Sailcraft areal densities
ranging from 8.9g/m2 down to 4.6g/m2 were proposed by NASA as a goal, and since the sail
material used in NASA experiments had an areal density of 3g/m2 , this leaves 1.6g/m2 for the
supporting structure, bus and payload.
Parallel applications of more immediate interest are high-precision space reflectors for
communications, earth observation, or radio-astronomy. At a few tens of meters in diameter, the
ones considered are orders of magnitude smaller than space mirrors and they can afford an order of
magnitude higher areal mass of typically 0.5 kg/m2 , but some of the goals and technology are
synergistic. Here we mention the work by Datashvilli, Baier and co-workers [40,41], with a small
model of their design shown folded and then deployed in Fig. 10.
Fig. 10. Deployment of membrane reflector model (0.6m diameter), a) stowed, b) deploying, c)
(scaled) deployed [40].
7. System Economics
While it is obviously extremely difficult to predict the costs of generated energy by space mirrors,
because of the novelty of the concept, uncertainties about developments that will arise during the
265
decades needed for its materialization, as well as about all of the externalities that will accompany
it, several economics analyses were conducted to at least provide an initial prediction.
A study by Ehricke in 1979 [21] estimated that the investment for Biosoletta is of the order of
$1,200 billion, that is about $80 billion/year for the 15 yr construction time he predicted.
The NASA economic analysis of their SOLARES space mirror concept [19] see also Section 3) to
supply approximately the global electricity demand at that time (32 EJ, which in 2010 is 77 EJ [2])
based on a desired 15% capital return, 30-year system life, and a load factor which takes into
account eclipse and inclement periods, found that generated electricity costs range from about 2.5
c/kWh to less than 16 c/kWh (in 1977 US cents), and that the ground station for converting the solar
radiation received from the mirrors to electricity by using PV is the major component of the total
system investment, since the cost of reflectors in space is much lower. We note that this was based
on PV system costs of $5/Wpeak that have since then dropped in some favorable cases up to less than
$2/Wpeak [42], but that is likely to have also been true for other components of the space mirror
system. The same study states that if the solar radiation incident on earth from SOLARES was used
just as heat, the cost would be about 1 c/100MJ thermal.
In that analysis, performed on the assumption of making 80,000 space mirrors of 1 km diameter
each to be placed in a 4,146 km altitude orbit, they assumed the use of a thin film mirror made of a
0.1 m Al reflector on 2.5 m polymer substrate weighing 4 gr/m2 , and the structure made from
HM graphite-epoxy. The total mirror weight, including also controls, instrumentation and growth
allowance, was estimated to be 10.01 gr/m2 , i.e. 7,860 kg/mirror, and the total cost of the mirror
including prorated R&D costs excluding transportation costs into orbit was $1,654,000 per mirror,
i.e. $2.11/m2 . Including the transportation they came up with a total price of $2 million per mirror,
i.e. $2.55/m2 . They estimated the cost of the total system to supply approximately the entire world
1977 electricity demand of about 32 EJ to be about $500 billion or an average of nearly $40
billion/year with their assumption of a 15-year implementation period.
A major problem with their economic analysis is that they used a cost of $44/kg for transportation
into the needed orbit, assuming the existence of the planned Heavy Lift Vehicle. This is an areal
mirror cost of $0.44/m2 . A vehicle that can transport at such a cost has never materialized, in fact
the current space transportation costs acan at best be around $3,000/kg and are more likely around
$10,000/kg [43]. Conservatively assuming the latter cost and keeping all the NASA assumptions the
same, the price of a 1 km diameter orbiting mirror would rise to $80.3 million, i.e. $102.30/m2 (40.1
times higher) and would thus raise the price of generated electricity in similar proportion and make
it highly uncompetitive. It is noteworthy that the cost of transportation to orbit is also the major
obstacle for economic deployment and use of the SPS, and the only way to have competitive space
power generation is to reduce the transportation costs to about $200/kg, i.e. by nearly 2 orders of
magnitude, achievable by frequently planned but never commercially produced reusable launch
vehicles (RLV) [10,11,14] and other methods.
We have also conducted an economic feasibility analysis of the use of a space solar mirror system
by approximating life time costs, profits and resulting revenues for three separate applications: (1)
24hr farm lighting, (2) Night-time illumination in Polar Regions, and (3) Greenhouse produce
growing in Polar Regions.
The mirror aerial weight we determined was 11 gr/m2 (an aluminized 7.62 m Kapton mirror
reflector),which turned out to be close to the 10 gr/m2 used in the NASA study [19], and
considering all of the reflector components the reflector cost was estimated at $2.05/m2 . Using the
NASA data that show that reflector cost is about 48% of the total mirror cost without transportation,
our estimate of the mirror area cost was thus $4.27/m2 .
For transportation costs, we used those for the Falcon 9 rocket in [44], at $7,143/kg. i.e $78.57/m2
of mirror. Like in NASAs analysis we assumed a functional life of the mirror as 30 years and a
discount rate of 15% but also calculated for a 5% discount rate to provide a range of values that
266
come closer to present conditions. A lifetime maintenance cost of $1.27/m2 (4.25 c/(m2 yr) was
added. Summing the mirror and its launch costs, the total areal investment cost comes to $82.84/m2
($65.03 million per mirror) and the 4.25 c/(m2 yr) recurring maintenance costs.
From Eqs. (1) and (3) and as following NASAs estimates of the mirror area shown above and in
Section 3, each m2 of the mirror illuminates 0.093 m2 of ground site area, at the planed insolation of
1.25 kW/m2 . This amounts to 116.2 W per m2 of the space mirror. The annual solar energy
incidences are thus 39.4 GJ per m2 ground area, or 3.67 GJ per m2 of the space mirror.
If the energy is used for PV electricity generation at 15% conversion efficiency and normal
incidence, this would steadily generate 187.5 W electricity per m2 of the illuminated ground area,
and thus 17.4 W electricity per m2 of the space mirror, i.e. 150 kWh/yr (0.55 GJ/yr) per m2 of the
space mirror. At the typical US electricity price of $0.012/kWh, this should generate an annual
revenue of $18 per m2 of the space mirror. The cost of the PV system is estimated to be $4/Wpeak
[42,45], and considering the above result that the power generation is 17.4 W per m2 of the space
mirror, this would be i.e. $69.60 per m2 of the space mirror and is added to the $82.84/m2 space
mirror cost for a total areal investment cost of $152.44 per m2 of the space mirror.
The present value of the initial investment into the solar space mirror system for a discount rate of
5% is $276.05 and the NPV = $276.05 $152.44 = $123.61, and the internal rate of return (IRR) is
11%, indicating a financially rather viable investment. The NPV based payback period is 14
years. If the discount rate was 15%, as assumed in the prior NASA study, then the NPV becomes
negative, -$34.53, i.e. not financially viable. Discount rates up to 11% produce positive NPV.
Since transportation into orbit is in this analysis 52% of the needed capital investment, and since it
is the cost item most likely to decrease significantly in the future, the same economic analysis was
performed for the transportation costs of $200/kg, amounting to $2.22/m2 and resulting in an overall
mirror system cost of $6.51/m2 and to total mirror+PV system cost of $76.11/m2 . Now the
investment achieves an IRR of 24% and becomes very viable even under the 15% discount rate, and
we note that the mirror system becomes only 8.6% of the total cost.
Using the solar radiation from the mirror to grow agricultural products in Polar Regions allows an
extension of the growing season from the current 4 months [46] to 12 months per year. The benefit
is obviously avoidance of the need for importation of expensive produce in the off seasons, and the
amount saving by growing locally is US$4.50/lb or $45.52 per m2 of the space mirror. This creates
annual revenues that are 2.5- fold higher than those estimated above for electricity generation.
Furthermore, while not calculated here, the investment into a tomato growing greenhouse system
may also be lower than the installation of a PV energy conversion system, with all of this pointing
to the recommendation that such agricultural use is financially most viable.
Nighttime municipal illumination failed in this analysis to break even over the lifetime of the
mirror, because of the relatively low density of needed streetlights in the selected northern regions,
but it may be a rather viable application for both civilian and limitary purposes, when the demand
justifies it.
It is noteworthy that the space mirror designs considered in the NASA analysis and also used here
were not optimized. Furthermore, the costs of materials, of space deployment and of PV electricity
generation are dropping. The economics can thus only improve, unless some unknown technical or
environmental problem arises during the more detailed system development and testing,
The estimated very high needed investments, of the order of more than $600 billion (about $40
billion per year for about 15 years), for the space mirror system become more acceptable and
appealing when compared with the expected accomplishment of providing renewable and relatively
clean energy for satisfying energy of the order of the entire global demand, with relatively minimal
global warming effects. These investments should also be compared with some other global
financial values: in 2010 the annual world and OECD GDPs were about $63,000 billion [48] and
$42,000 billion [49], respectively, and the world defense budgets were $1,437 billion (2.3% of
267
GDP) [50]. The estimated annual expenditure for the space mirrors project are thus 0.06% of the
World GDP, of the order of the World Bank subscribed capital of $44 billion for 2010.
Some other proposed high magnitude global renewable energy projects were estimated for the
Space Power Satellite (SPS) [5,9] at $908 to $15,000/kWe, which for generating the current global
power capacity of 4.4 TWe [2] would require an investment of $4,000 to $65,000 billion f, and for
the DESERTEC project at close to $600 billion to supply by 2050 only 700 TWh per year of
electricity from the Saharan deserts [51]. The space mirror concept is predicted to incur much
lower costs.
8. System Sustainability
A huge and basically untested project like this one requires a very careful formal scientific
sustainability analysis from the very start. Founded on the commonly used economic,
environmental, and social pillars of sustainability, such an analysis in quantitative form is beyond
the scope of this paper, but some major issues are identified and discussed, as follows.
same time, the associated launch capabilities developed for the space mirror system should
advances possibilities for space astronomical laboratories [19]. A response to the strong
criticisms about the effects of light that projects like Novey Svet ([24-26], Section1) may
create was posted by the Russian Space Regatta [24]; they indicate that the problem must
indeed be studied and any projects were and should be implemented with minimal damaging
impact, but also list the large advantages that such lighting can provide.
Generation of a large amount of space debris, and possible risks of their fall to earth
the Moon and other celestial bodies shall be used exclusively for peaceful purposes;
astronauts shall be regarded as the envoys of mankind;
States shall be responsible for national space activities whether carried out by governmental
or non- governmental entities;
States shall be liable for damage caused by their space objects; and
States shall avoid harmful contamination of space and celestial bodies.
As of October 2011, 100 countries and states are parties to the treaty, while another 26 have
signed the treaty but have not completed ratification. The treaty remains, however, very
incomplete and lacks some essential detail and clearly effective enforcement measures.
Amongst the obvious omissions are intellectual property of space research, and space pollution
[56,57]. A much more solid treaty must be developed to ensure internationally fair and safe
deployment and use of the space mirrors project.
Acknowledgments
My former students Kamal Shair, Betsy Rosenblatt, and Travis Schlegel have conducted the
experiments presented in Section 6.2 and contributed to other parts of this paper. I am grateful to
William P. Gilbreath and Stuart W. Bowen, authors of some the seminal NASA papers on space
mirrors, for their answers to my questions. Iris Chu and Eri Mizukane have been very helpful in the
literature search for this study.
270
References
[1] Lior, N. The ECOS 2009 World Energy Panel: An introduction to the Panel and to the present
(2009) situation in sustainable energy development, Energy 36 (2011) 3620-3628.
[2] Lior, N. Sustainable energy development: A brief introduction to the present (May 2011)
situation, accepted for publication in Energy, 2012,
[3] Glaser, P.E. Power from the Sun: Its Future, Science, 162 (1968) 857-861.
[4] Glaser, PE, Davidson, FP, Csigi, KI, editors. Solar power satellites, the emerging energy option.
New York: Ellis Horwood, 1993.
[5] Mankins JC. Space solar power: a major new energy option? Journal of Aerospace Engineering
2001;14(2):38-45.
[6] Mankins, J.C. (USA), Kaya, N. (Japan), Editors. Space Solar Power: The First International
Assessment of Space Solar Power: Opportunities, Issues and Potential Pathways Forward.
August 2011, International Academy of Astronautics, 6 rue Galile, BP 1268-16, 75766 Paris
Cedex 16, France, ISBN/EAN 978-2-917761-11-3
[7] Criswell DR, Thompson RG. Data envelopment analysis of space and terrestrially-based large
scale commercial power systems for earth: a prototype analysis of their relative economic
advantages. Solar Energy 1996;56:119-31.
[8] Brown WC. The history of wireless power transmission. Solar Energy 1996;56:4-21.
[9] Woodcock, G.R. Solar Satellites - space key to our power future. Astronautics and Aeronautics,
Vol. 15, July- Aug 1977, 30-43
[10] NASA, Solar Power Satellite Concept Evaluation. vol. 1, NASA JSC-12973, July 1977
[11] Lior, N. Power from space. Energy Conversion & Management J. 2001;42(15-17):1769-1805.
[12] Tarlecki Jason, Lior Noam, Zhang Na. Analysis of thermal cycles and working fluids for
power generation in space. Energy Conversion & Management2007;48:2864-78.
[13] Zidanek, A., Ambroi , M. Milfelner, M. Blinc, R. Lior, N. Solar orbital power: Sustainability
analysis. Energy 36 (2011) 1986-1995.
[14] National Space Society, http://www.nss.org/settlement/ssp/index.htm [accessed 2/19/2012].
[15] Oberth, H. Wege zur Raumschiffahrt. Berlin: Oldenburg Verlag, 1929 (Translation, NASA TTF-662).
[16] Buckingham, A.G, Watson, H.M. Basic Concepts of Orbiting Reflectors," J. Spacecraft, vol. 5,
May 1968, pp. 852-853.
[17] Billman, K.W., Gilbreath, W. P., Bowen, S.W. Introductory Assessment of Orbiting Reflectors
for Terrestrial Power Generation. NASA TM X-73,230, 1977.
[18] Billman, K.W., Gilbreath, W. P., Bowen, S.W. Satellite Mirror Systems for Providing
Terrestrial Power: System Concept. Paper AAS-77-240, 23rd Meeting of the AAS, San
Francisco, Calif., Oct. 1977.
[19] Billman, K.W., Gilbreath, W. P., Bowen, S.W. SOLARES-A New Hope for Solar Energy?
Alternative energy sources; Proceedings of the Miami International Conference, Miami Beach,
Fla., December 5-7, 1977. Volume 1. (A79-33984 13-44) Washington, D.C., Hemisphere
Publishing Corp., 1978, p. 233-255. IISSN: 0278-1662
[20] Gilbreath, W. P, Billman, K.W. Bowen, S.W. Enhanced solar energy options using earthorbiting mirrors. Proc. 13th Intersociety Energy Conversion Engineering Conf. San Diego,
California August 20 - 25, 1978, Paper 789186, Vol. 2, 1528-1534 Published by the Society of
Automotive Engineers, Inc. 400 Commonwealth Drive, Warrendale, PA 15096.
[21] Ehricke, Krafft A. Kraftsoletta - eine Industrie-Sonne fur Europa - Moglichkeiten
kostenrealistischer Solarenergie. Astronautik, vol. 3, 1977, 85-87.
[22] Ehricke, Krafft A. Space light: space industrial enhancement of the solar option. Acta
Astronautica Vol. 6. pp. 1515_1633, 1979.
[23]
Krafftehricke.com
Krafft
A.
Ehricke
and
America
in
Space.
[http://www.krafftaehricke.com/ehricke_power_relay_satellite_2.php] [accessed 2/19/2012].
271
[24] Sistach, L. A. The moon orbital mirror.The International Academy of Astronautics, and the
International Institute of Space Law, 54th International Astronautical Congress of the
International Astronautical Federation, 29 Sept. 2003, Bremen, Germany.
[25] Space Regatta Consortium. An open letter to astronomers professional and amateur.
http://src.space.ru/page_30e.htm [accessed 2/19/2012].
[26] Ashurly, M.Z. Space-based mirror reflector of solar radiation. J. Opt. Technol. 67 (6), June
2000 577-579.
[27] S.P. Korolev Rocket and Space Corporation Energia. http://www.energia.ru/english/
[accessed 2/19/2012].
[28] Shpakovsky, N. "Space Mirror." Space Mirror. The TRIZ Journal. <http://www.trizjournal.com/archives/2002/06/e/index.htm>. [accessed 2/19/2012].
[29] Garner, C., Diedrich, D., Leipold, M. A Summary of Solar Sail Technology Developments and
Proposed Demonstration Missions, Jet Propulsion Laboratory report JPC992697.
[30] Freiman, Chana. "Mirror, Mirror, up in Space? - Space Mirror for Providing Light." CBS
Interactive,
3
Dec.
1993.
Web.
10
Dec.
2011.
<http://findarticles.com/p/articles/mi_m1590/is_n7_v50/ai_14719235/>. [accessed 2/19/2012].
[31] Steen, M. (1999, August 2). UK Astronomers Attack Russian Space Mirror Plan.
http://www.planetark.com/avantgo/dailynewsstory.cfm?newsid=1073 [accessed 2/19/2012].
[32] NASA. History: Glenn Contributions to Deep Space 1. May 21, 2008 update, NASA.
<http://www.nasa.gov/centers/glenn/about/history/ds1.html>. [accessed 2/19/2012].
[33] Wong, Wayne. "SBIR Inflatable Concentrator Tested for Space." Aerospace Technology
Innovation. Sept.-Oct. 2001. NASA. http://nctn.hq.nasa.gov/innovation/innovation95/6smallbiz3.html [accessed 2/19/2011]..
[34] Fink, D. Biersack, J. P., Stdele, M. On the performance and lifetime of solar mirror foils in
space. Space Solar Power Review. Vol. 5, pp. 91-100, 1985.
[35] Gulino, D.A. The Survivability of Large Space-Borne Reflectors Under Atomic Oxygen and
Micrometeoroid Impact. NASA Technical Memorandum 88914, Prepared for the 25th
Aerospace Sciences Meeting sponsored by the American Institute of Aeronautics and
Astronautics, Reno, Nevada, January 12-15, 1987
[36] Kutz, Myer. Handbook of Environmental Degradation of Materials. (pp. 479-487). William
Andrew Publishing. 2005.
[37] Griffin, Michael D.; French, J. R. Space Vehicle Design (2nd Edition). (pp. 80, Table 3.5 on p.
87,
p.431).
American
Institute
of
Aeronautics
and
Astronautics.
<http://knovel.com/web/portal/browse/display?_EXT_KNOVEL_DISPLAY_bookid=1597&V
erticalID=0>. [accessed 2/19/2012].
[38] Lichodziejewski, D., Derbs, B., West, J., Reinert, R., Belvin, K., Pappa, R. Bringing an
Effective Solar Sail Design Toward TRL 6. Paper AIAA 2003-4659, 39th AIAA
Joint Propulsion Conference, 2003.
[39] Brown, M.A. Mechanical Design of a Kilometer-Sized Solar Sail AIAA 2010-2586 12 15
April 2010, Orlando, Florida 1010-1023
[40] Datashvili, L., Baier, H.
Building blocks of large deployable precision membrane
reflectors", Proceedings of the 42nd AIAA/ASME Structures, Structural Dynamics and
Materials Conference, AIAA paper no. 2001-1478, April 2001, Seattle
[41] Datashvili, L., Baier, H. Wehrle, E., Kuhn, T., Hoffmann, J. Large Shell- Membrane Space
Reflectors. 51st AIAAlASME/ASCE/AHS/ASC Structures, Structural Dynamics, and
Materials Conference, 18th AIAA 2010-2504 12 15 April 2010, Orlando, Florida.
[42] Solarbuzz, Module prices - Retail Price Summary - February 2012 Update.
http://solarbuzz.com/facts-and-figures/retail-price-environment/module-prices [accessed 2/19/2012].
[43] Koelle, D.E. Cost efficiency as design and selection criterion for future launch vehicles Acta
Astronautica Volume 57, Issues 28, JulyOctober 2005, Pages 623629
[44] Spacex Space Exploration Technologies. Falcon 9 Overview. March 2009.
http://www.spacex.com/falcon9.php [accessed 2/19/2012].
272
[45] REN 21 (Renewable Energy Policy Network for the 21st Century). Renewables Global Status
Report 2011 Update, http://www.ren21.net/default.aspx?tabid=5434 [accessed 2/19/2011].
[46] Nield, Jeff. Growing Food in the Arctic Circle. August 29, 2008.
http://www.treehugger.com/files/2008/08/growing- food- in-the-arctic-circle.php
[accessed
2/19/2012].
[47] Cook, Roberta, Calvin, Linda. North American Greenhouse Tomatoes Emerge as a Major
Market
Force.
Amber
Waves.
April
2005.
http://www.ers.usda.gov/Amberwaves/April05/Features/GreenhouseTomatoes.htm [accessed
2/19/2012].
[48]
World
Bank.
World
GDP.
http://data.worldbank.org/indicator/NY.GDP.MKTP.CD/countries/1W?display=graph
[accessed 2/19/2011].
[49] OECD. Gross domestic product in US dollars. OECD 2012.
http://www.oecdilibrary.org/sites/gdp-cusd-table-2012-1en/index.html;jsessionid=cld88aof9i8.delta?contentType=&itemId=/content/table/gdp-cusdtable-2012-1-en&containerItemId=/content/table/gdp-cusd-table-2012-1en&accessItemIds=&mimeType=text/html [accessed 2/19/2012].
[50] Stockholm International Peace Research Institute (SIPRI). http://milexdata.sipri.org/ [accessed
2/19/2012].
[51] DESERTEC Foundation, Clean power from deserts. http://www.desertec.org/ [accessed
2/19/2012].
[52] Longcore, T., Rich, C. Ecological light pollution. Frontiers in Ecology and the Environment,
Vol. 2, No. 4, May, 2004, 191-198.
[53] Laframboise, J. G., Chou, B. R.. Space Mirror Experiments: A Potential Threat to Human
Eyes. Journal of the Royal Astronomical Society of Canada, 94, 237-240, 2000, June 27.
[54] Chepesiuk R 2009. Missing the Dark: Health Effects of Light Pollution. Environ Health
Perspect 117:A20-A27. http://dx.doi.org/10.1289/ehp.117-a20
[55] U.N. The outer space treaty, http://www.oosa.unvienna.org/oosa/SpaceLaw/outerspt.html
[accessed 2/19/2012].
[56] European Centre for Space Law (ECSL). Space policy documents and useful readings on
regional
and
national
space
legislations.
http://www.esa.int/SPECIALS/ECSL/SEMDTNLTRJG_0.html [accessed 2/19/2012].
[57] Oosterlinck, Rene <Who owns and regulates space and why are space laws necessary?.
European Space Agency (ESA) http://www.youtube.com/watch?v=ADTg5CfNNUA [accessed
2/19/2012].
273
Universit de Pau et des Pays de lAdour, LaTEP, Pau, France, maxime.perier-muzet@etud.univ-pau.fr (CA)
b
Universit de Pau et des Pays de lAdour, LaTEP, Pau, France, pascal.stouffs@univ-pau.fr
c
Universit de Pau et des Pays de lAdour, LaTEP, Pau, France, jean-pierre.bedecarrats@univ-pau.fr
d
Laboratoire de Physique Et Ingnierie Mathmatique pour lEnergie et lenvironnemeNT (PIMENT),
Ile de la Runion, France, jean.castaing-lasvignottes@univ-reunion.fr
Abstract:
A heat driven thermoacoustic cooler consists of a thermoacoustic engine that converts heat into acoustic
waves, coupled to a thermoacoustic cooler that converts this acoustic energy into cooling effect. These
machines have simple structures without moving parts. The coupling of a solar concentrator and a heat
driven thermoacoustic cooler seems to be an interesting alternative to the electrically driven compression
vapour cycle. As the other solar refrigeration systems, even if the cooling demand generally increases with
the intensity of the solar radiation, one of the major difficulties is to insure a frigorific power supply when
there is no or low solar radiation. In our prototype, in order to guarantee a sufficient cooling capacity to face
to refrigeration loads in spite of the production fluctuations, a latent cold storage has been considered. The
aim of the work presented here is to investigate the behaviour of this key element under several design and
operative conditions. A description of the future prototype is done insisting on the thermoacoustic
refrigeration and the cold storage system. A modelling of the main elements of the prototype is developed.
The results of simulations under real solar radiation as well as a parametric study considering the main
design and operative parameters of the cold thermal storage system are presented.
Keywords:
Solar energy, Solar refrigeration, Thermoacoustic refrigerator, Cold latent thermal storage.
1. Introduction
Energy use for refrigeration has risen sharply in recent years. Nowadays, the major part of this
production is provided by electrically driven vapour compression machines. Globally, the
refrigeration devices consumed roughly 15% of the world electricity production. The forecasts for
the refrigeration indicate an increase in the number of units in operation over the coming years. To
ensure the refrigerating production in the next years while responding to environmental challenges
(emissions of greenhouse gases and ozone layer) the future cooling machines should not be primary
energy intensive and should use environmentally friendly refrigerants [1, 2].
Thanks to numerous possible combinations between solar thermal collectors and heat driven
cooling machines, solar cooling equipment seems to be an interesting alternative [3]. Among these
possibilities, the coupling of a solar concentrator with a heat driven thermoacoustic refrigerator is a
promising technology. Compared to the other refrigeration technologies, heat driven thermoacoustic
machines have several advantages with no moving part and environmentally friendly gas as
working fluid like helium, argon or air [4]. During the last decades, experimental investigations
have been carried out on these machines. They have proven their ability to reach very low
temperatures compatible with the liquefaction of gases such as, natural gas, nitrogen or hydrogen
[5-7]. Other studies have demonstrated the possibility to uses thermoacoustic refrigeration for
higher temperature applications like for food refrigeration [8]. The couplings of thermodriven
274
thermoacoustic refrigerators and solar concentrators have also already been experimentally studied
thanks to low power prototypes [9, 10].
Although the cooling demand is generally higher when the sun is shining, the intermittent nature of
solar energy is one of the major constraints for using solar cooling systems. To meet the timedependency of the primary energy supply and end-use requirement, a thermal energy storage has to
be used. Various configurations of the energy storage can be achieved; storing the hot energy for
continuously supply the refrigeration system or storing the produced cool energy. This energy can
be stored in the form of sensible heat (in liquid or in solid), latent heat of a Phase Change Material
(PCM ), or by chemical reaction. The choice of the storage system depends on various criteria like
the amount of energy that has to be stored or the storage temperature. However for cooling
application, the latent storages have numerous advantages: the technology is well known with a
high energy storage density; the stored and retrieved energy is at a quasi constant temperature
which corresponds to the phase change transition of the PCM [11].
The studied prototype consists in a one kW scale solar thermodriven thermoacoustic refrigerator
supplied in primary energy by a solar dish concentrator. To ensure a low variability of the
availability of refrigerating capacity, a cold latent thermal storage is coupled to the refrigerator.
In the first part of this paper the project and the main parts of its development are presented.
Secondly, the lumped model that has been built is described. Then, the first simulation results for a
long period including days with various solar radiation conditions are presented and discussed.
Finally, the results of the parametric study are presented, considering the above-mentioned different
cold storage design characteristics and storage strategies.
2. Prototype description
The device considered here consists in a solar concentrator, a solar flux modulator, a solar receiver,
a thermoacoustic machine composed firstly of a thermoacoustic prime mover linked to an acoustic
resonator and secondly of a thermoacoustic refrigerator linked to a cool thermal energy storage
(Fig. 1).
Thermoacoustic
refrigerator
Solar
T
receiver cg
Tfr
Loads
Thermoacoustic
prime mover
Cool thermal
storage
Solar
flux modulator
Solar concentrator
The solar driven thermoacoustic refrigerator heat fluxes represented in Fig. 2 are described in the
following sections. The direct solar radiation is collected, reflected and concentrated by a parabolic
mirror. M ore details on this element can be found in [12]. A solar flux modulator is placed between
the concentrator and the receiver cavity. This element regulates the solar power entering in the
absorbing cavity; it thus allows controlling the temperature of this latter. The concentrated solar
radiation is collected by an absorber situated in the receiver cavity. This latter transfers
approximately 4 kW to the working fluid (helium at about 4 M Pa), while being at a temperature
close to 500C. M ore details on this component can be found in [13, 14].
Concentrator optical losses
and
modulator losses
Solar receiver
heat losses
Collected
solar radiation
C oncentrated
solar radiation
Generator
heat losses
Heat rejected
at Tam b
Resonat or
l osses
Heat absorbed
at T cold stor age
Refrigerator
heat losses
Aborbed solar
radiation
Heat rejected
at T amb
Acoustic work
Solar concentrator
and
solar flux modulator
Thermoacoust ic
prime mover
Thermoacoustic
resonator
Thermoacoustic
refrigerator
3. Modelling
To predict the future performance of the prototype a lumped model of the entire plant has been
developed. The time variation of the energy source and the frigorific power consumption imposes a
transient approach. The solar part and the cold storage are treated by simplified transient models.
For the thermodriven thermoacoustic refrigerator, a quasi-stationary approach it used. We describe
the equations and the principal assumptions of this model in the following sections.
(1)
276
Where DNI is the Direct Normal Irradiation and Acollector_aperture is the collector aperture area.
Non ref lective band
Useful area
Qsol modul
Qs ol refl
Amodul_opened
Amodul_aperture
The Thermoacoustic machine, the flow pipes, the frame of the concentrator, etc shadow a part of
the dish (Fig. 3). The sun tracking system is considered to be accurate so that the useful area of the
concentrator remains constant all along the day. A constant and uniform mirror reflectivity is
assumed. According these hypotheses, the reflected solar power is expressed by:
Qsol _ ref lect
Where
(2)
Acollector_useful is
Acollector _ useful
A uniform solar concentrated flux is assumed in the modulator plan (Fig. 4). It is also considered
that it has no impact if it is totally opened. The solar modulated power is determined by:
Qsol _ mod ul
Qs ol _ ref lect
Amod ul _ apert ur e
Amod ul _ opened
(3)
Where Amodul_aperture is the area of the modulator aperture and Amodul_opened is the area of the
modulator aperture when it is totally opened.
modul
Qloss_or
Qsol_abs
Dint_cavity
Q sol
Dext_cavity
Qoverflow
Qech hel
Lext_cavity
The cavity absorbs the solar concentrated flux. Thanks to the hot exchanger, a large part of this heat
flux is transferred to the working fluid, the rest is exchanged with the surroundings (Fig. 5). The
277
cavity walls are considered as an equivalent mass of the hot exchanger. The energy balance of these
elements can be written as:
dU cavity
hg
Qsol _ abs
dt
Ql oss _ cavi ty
Qhel
wall _ hg ,
(4)
Where Qsol _ abs is the absorbed solar power. The absorber cavity absorbs the major part of the solar
power which passes through the solar flux modulator. The rest is the solar loss by overflow. This
loss is due to the diameter difference between the solar focus point and the cavity orifice. If the
diameter of the focal point is larger than the one of the cavity aperture (as it is the case for the
studied prototype, in order to limit the thermal losses through the orifice) a part of the solar
radiation cannot enter in this later. In this model, the solar overflow losses are evaluated considering
a solar overflow rate ( sol_overflow). This rate is equal to a constant value when the solar flux
modulator is totally opened and equal to zero when the modulator is partially closed. The absorbed
solar power is calculated by:
Qsol _ abs
Qsol _ mod ul . 1
overflow
(5)
The cavity exchanges with the surrounding by the orifice, the lateral wall and the back wall. The
cavity thermal losses Qloss _ cavity can be determined by:
Qloss _ cavity
Qloss _ or
Qloss_ lateral _ wall and Qloss_ back _ wall are calculated considering that radiative heat exchanges are
negligible compared to convective ones from the insulating material surface. The convective heat
exchange coefficient between the orifice and the surrounding is estimated from [17].
The two others cavity losses are determined using the various thermal resistances except the thermal
resistance of contact between the cavity wall and the insulation which is neglected.
Q evac
Qech hel
Qloss_wall
Fig. 7. Heat fluxes in the ambient and cold heat exchangers
The cold exchanger of the refrigerator as well as the heat exchangers at intermediate temperature,
(Fig. 7) transfer on one side with the helium and on the other with a heat transfer fluid. The cold one
is submitted to heat transfer with surroundings while such an exchange is neglected for the other
ones because of the low temperature difference between these exchangers and ambient air.
Applying the first law to the wall and to the heat transfer fluid of these elements gives:
dU wall _ exch
dt
dU fluid _ exch
dt
Q fluid
wall _ exch
Qhel
wall _ exch
Qloss _ exch
,
Qwall
fluid _ exch
(7)
The convective heat transfers between the heat exchanger walls and the fluids are given by:
Q fluid
wall _ exch
T fluid
Twall _ exch ,
(8)
Where the heat exchange coefficients ( fluid-wall) are calculated thanks to a correlation for a fluid
flowing across a bank of tubes proposed in [18].
278
Pexch
m flui d _ exch
, (9)
flui d
Where Pexch are the pressure drops inside the exchangers. They are calculated thanks to a
correlation for a fluid flowing across a bank of tubes determined by [18] and fluid is the density of
the fluid.
Each equation relative to the exchanger walls has a term of transfer with helium. A stationary model
is used to determine these heat fluxes at each calculation time step of the quasi-stationary model
[19].
wall _ pipe
dt
Qvisco _ pipe ,
(10)
The viscous dissipations Qvisco _ pi pe , are calculated by (11) with the pressure drop determined by:
Ppipe
V fluid 2
fluid
L pipe
Dint_ pipe
(11)
With K the local coefficients of pressure drop for valves, changes of direction, and
coefficient of pressure drop.
the linear
(12)
Where Qel ec _ heater is the electrical power which feeds the heater.
A constant efficiency ( pump) is considered for the pump. Considering these limitations, the energy
balance of the fluid inside the circulating pump is expressed by:
hfluid _ output _ pump hfluid _ input _ pump
Pcircuit
pump
(13)
fluid
279
C
ry
n
tio
sa
l li
mfluid
Tfluid int
So lid
ng
Liqui d
Tnod
C ry st a llisa t io n a f te r a
inc o mpl et e m elt ing
M el tin g a f t er a
inc o mp let e c ry st a llisa t io n
So lid
mfluid
Tfluid out
Cr
ys
tal
lis
a
Liquid
io
sta
In order to simplify the physical model of the latent heat storage, the following assumptions were
made:
The tank is vertical with flow from the bottom to the top for the charge mode and the discharge
mode;
The flow in the tank is axial and incompressible;
Variation of temperature of the heat transfer fluid occurs only along the axial direction, i.e. as
checked experimentally [16], the temperature is independent of the radial position;
The insulation of the tank is considered perfect;
Heat transfer by conduction is neglected in the heat transfer fluid;
Kinetic and potential energy changes are negligible;
The tank is divided in several control volumes according to the height (Fig. 9);
The nodules are considered as exchangers. The energy flux exchanged is proportional to the
difference of temperature between the fluid and the interior of the spherical nodule;
The supercooling phenomenon is taken into account.
The variation of the internal energy of the tank elements (metal wall pipe, flow diffusers, etc..) is
neglected.
The pressure losses are considered for the flow inlet and outlet, the flow diffusers and the nodule
bed thank to local coefficients of pressure drops.
elt
in
Solid
el
ti
Liquid
N n od
m fluid h fluid in
h fluid out
Q nod
fluid ,i
i 1
mn od
du n od ,i
dt
Q no d
fluid ,i
(15)
With
280
u nod ,i
uref
Tref
uref
Tref
uref
Tref
if
xnod ,i
xsolid L
if
if
xnod ,i
x nod ,i
(16)
Even when the heat transfer fluid temperature is considered uniform in each layer, all the nodules of
each layer do not simultaneously pass through the phase change at the melting temperature Tmelt
during cooling because of the supercooling and the erratic character of the crystallisation. The
nodules can be in different states (non-crystallised, entirely crystallised or partly crystallised)
according to their own value of the beginning of the crystallization (Fig. 10).
Applying the nucleation laws, the number of new crystallisations and the corresponding fluxes can
be calculated at each time t [20].
Considering a nodule of inner radius rint (Fig. 11). Uniform cooling of its surface will result in a
spherically symmetric crystallisation-front, r = rinterface (t) the inner radius of solid PCM , propagating
inwards from r = rint with liquid at Tmelt for 0 r rinterface (t) and solid for rinterface (t) r rint.
Assuming constant thermal properties in each phase, the steady-state solution of the heat conduction
in the solid phase has the form ( is the temperature of the solid PCM ):
r
(t )
1 interface
r
(17)
(r , t ) Tm elt T fluid (t ) Tmelt
rinterface (t )
k PCMsolid
kP CMsolid k PCMsolid rinterface
1
1
kenv
rint
k env
rext
ex trext
The interface conditions here have the standard form:
kP CMsolid
( r, t )
r
r ri nt erface ( t )
Qnod fluid ,i
2
4 rinterface
( t)
(18)
r ext
Liquid
Solid
r int
S ol id
r interface
Liquid
Fig. 11. Crystallization inside a nodule (left) and melting inside the nodule (right).
The determination of Q nod fluid before crystallisation starts and after crystallisation is finished is
done considering the uniform PCM temperature and it is possible to write that the internal energy
variation in the PCM is equal to the flux that leaves the nodule.
The heat transfer coefficient between the nodule and the fluid is determined by a correlation [20]
and so depends on the flow rate and on the fluid temperature.
Supercooling occurs only upon crystallisation but never upon melting. So, all the nodules from each
layer simultaneously pass through the phase change at the melting temperature Tmelt.
According to a simplifying assumption, the melting-front is considered to be concentric (Fig. 11)
and equations for Q nod fluid are the same than for crystallisation. During the melting process, heat is
transfered by natural convection and conduction. Only the heat conduction equation is kept into
consideration but an apparent thermal conductivity [21] is used in order to take the natural
convection into account.
281
4. Results
From the model described in the previous section, numerical simulations of the experimental plant
have been carried out assuming a constant ambient temperature of 20C. The solar radiation
conditions that have been used are the DNI measured between the 1st and the 7th of July 2006 by
PROM ES laboratory at Odeillo (South of France) (Fig. 12). The main parameters that have been
considered for this simulation are summarised in Table 1.
-2
DNI (W.m )
1000
800
600
400
200
Time (h)
0
0
24
48
72
96
120
144
168
13.5
3.5
0.92
0.1
0.1
30
1
1
350
77
500
400
300
200
100
Time (h)
0
0
24
48
72
96
120
144
168
As shown in Fig. 13, when the sun is shining, the temperature of the hot exchanger of the generator
increases extremely rapidly to reach the temperature threshold for wave generation. Actually the
machine starts only if a sufficient temperature gradient exists between the two heat exchangers of
the engine cycle. According to previous experiences, the starting hot temperature (and also stop
temperature when the exchanger temperature decreases) of the generator has been set at 600 K. The
solar power available is very large compared to the one consumed by the acoustic wave generation
and the thermal losses so that the temperature of the hot exchanger continues to grow rapidly to
reach the regulating temperature. This regulation is ensured by the control of the absorbed solar
radiation thank to the solar flux modulator. When the DNI is important, it can shut up to 45% of the
reflected solar power. At the end of the day or during cloudy periods, the hot exchanger is cooled
due to the power consumed by the acoustic wave generation (if its temperature is higher than 600
K) and due to the radiative and convective heat losses. The hot exchanger behaviour is roughly the
same for each simulated days with variations in function of the solar radiation fluctuations.
Qelec_h eater (W)
400
300
200
100
Time (h)
0
0
24
48
72
96
120
144
168
Solid fraction
Temperature (C)
So lid f raction
Tfr
0.8
0.6
20
15
10
5
0
-5
-10
-15
-20
0.4
0.2
0
0
24
48
72
96
120
144
Time (h)
-25
-30
168
Fig 15. Solid fraction in the storage tank and temperature of the cold exchanger of the refrigerator
The operative conditions corresponding to the frigorific loads have been set as follows: a constant
heating power if the fluid temperature at its outlet stays lower than -18C and no heating power if
this temperature is higher than this limit. To ensure the cooldown of the storage and the loop
elements the heater is turned off during the first 30 hours (Fig. 14 and 15). During this period, when
the refrigerator works, the temperature of the cold exchanger of the refrigerator decreases to reach a
temperature close to the melting temperature of the PCM (Fig. 15). Due to the adiabatic assumption
and no heating power, the temperature of the storage remains constant during the first night. At the
beginning of the second day (close to 32 hours after the beginning of the simulated
experimentation) the tank temperature continues to decrease below the melting temperature of the
PCM . Due to thermal resistances between the fluid and the nodule and due to the supercooling
283
effect, the temperature of the fluid has to be a few degrees lower than the melting temperature to
initiate the crystallisation. Even if the refrigerator continues to produce a frigorific power, the
temperature of the cold exchanger remains quasi constant after the initiation of the crystallisation
thank to the isothermal behaviour of the nodules during this phase. At the end of the second day, the
majority of the PCM is solid. During the following night, a large part of the solid PCM is melting to
supply the loads. During the next four days, the system reaches a quasi periodically steady state
condition, the storage tank solid fraction evolves between the same levels from one day to the other
and the temperature of the cold loop stays close to the melting temperature of PCM . Due to a low
insolation during the two last days, the solid fraction of the tank decreases and there are periods
where the storage cannot provide the cooling capacity within the imposed temperature condition.
20
10
Ti me (h)
0
Standard case
0
24
48
72
96
-10
120
144
168
Without supercooling
-20
-30
Fig. 16. Temperature of the cold exchanger of the refrigerator with and without PCM supercooling
Volume of the storage tank: The variations of the tank volume have significant impact on the cold
exchanger temperature (Fig 17) and the PCM solid fraction (Fig 18). With the smallest tank there
are periods where all the PCM into the tank is solid, thus the residual energy have to be stored as
sensible heat. This generates very low temperatures in the cold circuit. Considering the largest tank,
due to higher mass of fluid and PCM , a larger part of the energy is stored as sensible heat. The solid
fractions are thus in this case lower than the other one. Due to this, with this configuration, there is a
large period of time between the two last days where the storage cannot provide the cooling
capacity.
284
Temperature (C)
20
10
CS
Standard case
0
0
24
48
72
96
120
144
Ti me (h)
= 0.15 m
VVtank
= 0.15m3
168
= 0.25 m
VVtank
= 0.25m3
-10
-20
-30
-40
Fig. 17. Temperature of the cold exchanger of the refrigerator for different tank volumes
Sol id fracti on
1
0.8
0.6
St
Srie4
andard case
3
0.4
VSrie2
ta nk = 0.15 m
0.2
VSrie5
ta nk = 0.25 m
Ti me (h)
0
0
24
48
72
96
120
144
168
Fig. 18. Solid fraction in the storage tank for different tank volumes
Heat transfer fluid mass flow rate: Simulations have been made for different heat transfer fluid
mass flow rates. The obtained results show that in the simulation conditions, this parameter has only
a slight effect on the operation of the cold circuit (Fig. 19).
Temperature (C)
20
10
Ti me (h)
0
Standard case
CS
0
24
48
72
96
-10
120
144
168
-1
mfluid
= 0.05
kg.s
m
= 0.05
kg.s-1
-1
mfluid
= 0.2
kg.s
m
= 0.2
kg.s-1
-20
-30
Fig 19. Temperature of the cold exchanger of the refrigerator for different mass flow rates
Melting temperature of the PCM: For these simulations the thermophysical properties of the
PCM (except the melting temperature) are considered identical. Because of the low temperature
difference between the expected temperature at the heater and the melting temperature of the
highest PCM melting temperature, the major part of the energy is stored as sensible heat (Fig. 20.
and Fig. 21). Due to the difficulty of the refrigerator to reach a low temperature to initiate the
crystallisation of the lowest melting temperature PCM , the variation of the solid fraction in this case
is low. In these two cases the temperature stabilisation which is expected with latent energy storage
is not assured (Fig 21).
285
Solid fraction
1
0.8
0.6
Standar
Standardcas
case
e
0.4
Tm elt==-18C
-18.3C
Tm
Tm elt==-26C
-26.2C
Tm
0.2
Time (h)
0
0
24
48
72
96
120
144
168
Fig 20. Solid fraction in the storage tank for different PCM melting temperatures
Temperature (C)
20
10
Ti me (h)
0
0
24
48
72
96
-10
120
144
168
CSandard case
St
Tm
T
-18.3C
mel =
t =-18C
T
-26.2C
Tm
mel =
t =-26C
-20
-30
-40
Fig 21. Temperature of the cold exchanger of the refrigerator for different PCM melting
temperatures
4. Conclusion
A solar driven thermoacoustic cooler which consists of a thermoacoustic engine that converts solar
energy into acoustic waves, coupled to a thermoacoustic cooler that converts this acoustic energy
into cooling effect, has been studied using a numerical model. The designed prototype would be
able to deliver a refrigerating capacity of about 1 kW at -30C.
In order to guarantee a sufficient cooling capacity to face to refrigeration loads in spite of the
production fluctuations, a cold storage using encapsulated PCM has been considered. It stores the
excess cooling power when the sun is highly shining and restitutes it at night or during cloudy
hours.
The developed model permits to determine the future performances of the prototype but also to
choose the best configuration of the storage tank which has to be adapted to the loads. The main
design parameters for the cool storage are the tank volume, the crystallization and melting
temperature of the PCM and the mass flow rate of the heat transfer fluid.
Acknowledgments
This work was carried out within the framework of the TACSOL project. It is funded by the ANR
PRECODD. We would like to thank our project partners, S. Cordillet (PROM ES), P. Duthil
(IPNO), M.X. Franois (Hekyom), T. Le Polles (Hekyom), G. Olalde (PROM ES), M . Pierens
(IPNO), J.P. Thermeau (IPNO), for their help.
Nomenclature
A
C
area, m
specific heat, J kg-1 K-1
286
DNI
h
k
K
L
L
.
N
P
Q
heat flux, W
r
radius, m
t
time, s
T
temperature, K
u
specific internal energy, J kg-1
U
internal energy, J
x
solid fraction of PCM
Greek symbols
convective heat transfer coefficient, W m-2 K-1
emissivity
reflectivity of the concentrator
linear coefficient of pressure drop
density, kg m-3
StefanBoltzmann constant, W m-2 K-4
rate
efficiency
Subscripts and superscripts
abs
absorbed
amb ambient
collect collected
collector
solar collector
env
nodules envelope
exch exchanger
ext
external
fluid heat transfer fluid
hg
hot exchanger of the generator
hel-wall
exchange between the helium and the exchanger wall
int
internal
interface
interface liquid/solid of the PCM
liq
liquid phase of the PCM
loss yhermal losses
melt melting of the PCM
modul modulated or solar flux modulator
287
nod
or
PCM
ref
reflect
sol
solid
visco
nodule
orifice
Phase Change M aterial
reference
reflected
solar
solid phase of the PCM
viscous
References
[1]
289
DETEC- Univ. of Naples Federico II, P.le Tecchio 80, 80125 Naples, Italy, frcalise@unina.it
DIT Univ. of Naples Parthenope C. D. IS.5, 80143 Naples, Italy, laura.vanoli@uniparthenope.it
Abstract:
Photovoltaic/thermal (PVT) solar collectors are based on a combination of solar thermal and solar
photovoltaic collectors. PVT systems allow one to produce simultaneously electrical energy and thermal
energy by solar irradiation. Different PVT arrangements are presently under investigation. The most common
configuration is the "tube and sheet" one in which a photovoltaic layer is encapsulated in the absorber of a
conventional flat plate solar thermal collector. PVT electrical efficiency may be even higher than the one of a
similar PV system when the fluid average temperature is relatively low. Therefore, the majority of PVT
systems presently under development produce hot streams at temperatures lower than 45 C. However, in
the last few years different types of high temperature PVT systems are also under investigation. In fact, rising
PVT fluid outlet temperature would dramatically increase the range of possible thermal applications. This
paper is focused on this specific technology and presents a design procedure and a simulation model of a
novel concentrating PVT collector. The layout of the PVT system under investigation was derived from a
prototype recently presented in literature and commercially available in order to improve its electrical
performance. The prototype consisted in a parabolic trough concentrator and a linear triangular receiver. The
bottom surfaces of the receiver are equipped with mono-crystalline silicon cells whereas the top surface is
covered by an absorbing surface. The aperture area of the parabola was covered by a glass in order to
improve the thermal efficiency of the system. In the modified version of the collector considered in this paper,
two changes are implemented: the cover glass was eliminated and the mono-crystalline silicon cells were
replaced by triple-junction cells. These modifications allow one to increase significantly the electrical
efficiency of the system especially in case of high operating temperatures. In order to analyze the
performance of the modified Concentrating PVT (CPVT) collector a detailed mathematical model was
implemented. This model is based on zero-dimensional energy balances on the control volumes of the
system. The simulation model allows one to calculate in detail the temperatures of the main components of
the system (PV layer, concentrator, fluid inlet and outlet and metallic substrate) and the main energy flows
(electrical energy, useful thermal energy, radiative losses, convective losses). Results showed that the
performance of the system is excellent even when the fluid temperature is very high (>100 C). Conversely,
both electrical and thermal efficiencies dramatically decrease when the incident beam radiation decreases
Keywords:
PVT, triple-junction, solar energy
1. Introduction
In the last few years, special attention has been paid to the renewable energy sources as a
consequence of the dramatic decrease of the availability of conventional fossil fuels and the related
increase of their costs. Among the available renewable technologies (solar, wind, hydro,
geothermal, etc.), solar energy is commonly considered one of the most viable options. This is due
to large availability of solar radiation for the all over the world and to the recent development of its
conversion technologies. On the other hand, solar energy technology is still suffering of high capital
costs and low power density. However, during the last few years the capital cost of solar collectors
significantly decreased - specially in case of electrical ones - and their cost is expected to further
decrease during the next few years. Conversely, dramatic improvements of the power density are
not realistically expected since this parameter could be increased only by raising the conversion
efficiency of the solar collectors. The solar energy can be exploited for producing both electricity
290
(by photovoltaic collectors, PV) and heat (by thermal solar collectors, SC). From this point of view,
different commercial devices are available since several decades [1-2]. However, a possible
improvement of both PV and SC technologies consists in a combination of their effects. This occurs
in photovoltaic/thermal collectors (PVT) which simultaneously provide electricity and heat. The
basic principle of a PVT collector is simple, since it can be obtained by a conventional thermal
collector whose absorber is covered by a suitable PV layer [3]. The absorbed thermal energy is
distributed to a fluid (typically air or water), whereas the PV produces electricity [1-2]. The final
result of this arrangement is the combined production of electricity and heat and a possible
improvement of PV efficiency. In fact, the PV electrical efficiency is strongly dependent on the
system operating temperature, linearly decreasing with high values of such parameter [4].
Therefore, if the PV layer operating temperature is reduced by a cooling fluid, the system efficiency
will be higher than the one of conventional PV [1-2, 4]. In order to achieve this result, the outlet
temperature of the cooling fluid should be sufficiently low (usually < 40 C). For this reason, the
heat available from PVT systems can be used only for low-temperature heat demand (e.g., domestic
hot water, floor heating, etc.). From this point of view, an interesting application is the desiccant
cooling [5-6]. The selection of the PVT operating temperature is an important key-point in the
system design. In fact, while higher operating temperatures increase the potential use of the
cogenerative heat, they decrease the electricity production [1-2, 4]. As a consequence, researchers
are performing a special effort seeking to realise a PVT collector providing medium-temperature
heat (60 80 C) at high electrical efficiency [7]. A possible alternative for increasing fluid P VT
outlet temperature without decreasing PV electrical efficiency, may consist in the use of a heat
pump (driven by PV electricity) [1-2, 7]. Although the basic idea of the PVT was developed about
40 years ago, this product is still far from a mature commercialization [8]. Thus, several
researchers are investigating several novel P VT arrangements [8-11]. For example, Zhao et al.
investigated a novel PVT where thermal and electrical sections are separated [12]. One of the key
points in the design of a P VT system is the eventual selection of a transparent cover. In fact, the
covering can improve the thermal performance (higher insulation) but reduces the electrical one
(higher reflection). This topic is still under investigation by a number of researchers analyzing
different types of covers (tedlar, DEA, glass-to-glass, etc) [13]. Similarly, different studies
investigated the optimal design of fluid channels. Different options are under investigation: the
cooling fluid may flow between the PV and the absorber or between the absorber and the insulation
(as usual in SC) [1-2]. Usually, the adopted PVT cooling fluid is water [10]. However, several
studies are focused on the use of air [11] or in a combination of air and water [9]. The use of air as
cooling fluid is very attractive in case of building integration. In fact, some companies and
researchers are developing new prototypes of PVT facades or Building Integrated PVT (BIPVT)
[14] which can: i) produce electricity; ii) provide space heating during the winter; iii) shade from
the solar radiation in summer; iv) act as a ventilated faade in summer for reducing the cooling load
[1-2]. In this case, the appropriate selection of PV technology is crucial for maximizing system
overall performance [14]. Anyhow, the majority of the studies regarding P VT are focused on the
development of devices that reduce capital costs and improve the system reliability. In this
framework, a possible system configuration consists in the adoption of Concentrating P VT
collectors (CPVT). Basically, they are simple PVT collectors placed in the focus of some reflectors
(Fresnel, parabolic, dish, etc.) [1-2, 7, 15]. Obviously, the specific cost of this system is
dramatically lower than the flat plate PVT one; this is due to the lower amount of PV employed per
unit area. On the other hand, it must be considered that concentrating solar radiation devices
determine an increase of radiative flux on PV, increasing its operating temperature and therefore
decreasing its electrical efficiency. Usually, this drop off is not too high (typically, 0.45%/K for
silicon cells), if the increase of temperature is fair (a silicon cell having an efficiency of 15 % at 25
C, will show an efficiency of 11.6 % at 75 C). Conversely, for high concentration ratios, the PVT
operating temperature and the corresponding inefficiencies may significantly increase (for silicon
cells, the voltage drops to zero at 270 C) [4]. Therefore, the use of CP VT may be improved
adopting novel P V materials, such as multi-junction solar cells which can approach a nominal
291
efficiency of 40% [15-16]. The adoption of such materials in CPVT may lead to a system operating
up to 240 C at reasonable conversion efficiency (slightly lower than 20%) [15]. The perspective of
using high temperature PVT is very interesting since it extends the number of possible applications.
An example consists in the use of the high-temperature heat provided by the PVT to drive a heat
engine [17] or an Organic Rankine Cycle (ORC) [18] or a Solar Heating and Cooling system [15].
However, commercially or pre-commercially available CPVT systems are typically a small amount
of the PVT under development [8]. Similarly, the availability of theoretical and experimental
studies investigating CP VT performance is scarce. In particular, M ittelman et al [15, 19-20]
performed some experimental and theoretical works dealing with CPVT systems. In reference [20]
presented a novel miniature CPVT based on a dish concentrator (0.95 m2) and a silicon PV cell. A
thermal model for that system was developed in order to predict its performance. The system
provides 140-180 W of electricity and 400-500 W of heat. The cost of the system was evaluated at
2.5 $ per peak electric Watt. The operation at high temperature of similar CPVT systems was also
analyzed both for solar cooling [15] and water desalination [19] applications. A Parabolic Trough
CPVT prototype was experimentally investigated by Coventry [21]. In this work the author pointed
out that one of the major challenges in designing CPVT systems is to achieve an acceptable
radiation flux distribution. The concentration ratio of the system under investigation was 37,
thermal and electrical efficiencies were rated respectively at 58 % and 11 %. In this study the author
also concluded that the thermal efficiency of CPVT is higher than the one of flat plate solar
collectors only when radiation is scarce and/or ambient temperature is low. In fact, flat plate
collectors convert both beam and diffuse radiation, whereas CPVT can only convert the beam one.
Parabolic Trough CPVT were also investigated by Li et al. [22-24] comparing the electrical and
thermal efficiencies of the system varying the PV technology, for different concentration ratio.
Authors concluded that GeAs cells increase electrical efficiency with respect to silicon cells.
However, the thermal efficiency of GeAs results lower than the one of silicon cells. Authors also
pointed out that the cost of unit area of the GeAs is 3067.16 $/m2 versus the 131.34 $/m2 of the
silicon cell. A similar work was performed by Bernardo et al. analyzing a low concentrating
parabolic trough CPVT for the Swedish climate [25]. This study presents a thermal model of the
PVT subsequently validated by experimental data. The system is based on a combination of a one
axis tracking parabolic concentrator and silicon cells. This prototype is distributed by the Swedish
company Absolicon. Additional specific applications of CPVT technology are also investigated in
literature. Xu et al. [26] analyzed the integration of CP VT in the evaporator of an heat pump
system. Al-Alilli et al. [27] investigated the use of a CPVT in a desiccant cooling system. Finally,
Rosell [28] et al. investigated a low concentrating P VT system based on a linear Fresnel receiver.
the system was analyzed by a mathematical model subsequently validated by experimental data.
The rated thermal efficiency was about 60 %.
This paper is focused on the technology of concentrating parabolic trough PVT solar collectors and
it presents a design procedure and a simulation model of a novel concentrating PVT collector. The
layout of the PVT system under investigation was derived from a prototype recently presented in
literature and commercially available [25]. The design of this prototype was modified with the
scope to improve its electrical performance and it consists in a parabolic trough concentrator and a
linear triangular receiver. The bottom surfaces of the receiver are equipped with mono-crystalline
silicon cells whereas the top surface is covered by an absorbing surface. The aperture area of the
parabola was covered by a glass in order to improve the thermal efficiency of the system. In the
modified version of the collector considered in this paper, two modifications are implemented: the
cover glass was eliminated and the mono-crystalline silicon cells were replaced by triple-junction
cells allowing one to increase significantly the electrical efficiency of the system especially in case
of high operating temperature. In order to analyze the performance of the modified Concentrating
PVT (CPVT) collector a detailed mathematical model based on zero-dimensional energy balances
on the control volumes of the system was implemented. The simulation model calculates in detail
the temperatures of the main components of the system (PV layer, concentrator, fluid inlet and
outlet and metallic substrate) and the main energy flows (electrical energy, useful thermal energy,
292
radiative losses, convective losses). The input parameters of the model include all the weather
conditions (temperature, insolation, wind velocity, etc) and the geometrical/material parameters of
the systems (lengths, thermal resistances, thicknesses, etc.). The model was used in order to
evaluate both electrical and thermal efficiency curves related to the beam incident radiation, fluid
inlet temperature and external temperature.
Absorber
Fluid Channel
Substrate
Concentrator
Receiver
PV layer
PV layer
The two sides of the triangle facing the parabolic concentrator are equipped with triple-junction PV
layers, whereas the top side of the receiver is equipped with a thermal absorber. The triangular
receiver includes an inner channel where the fluid to be heated flows. Therefore, the concentrated
solar irradiation is converted simultaneously in electricity by the PV layer and in thermal energy by
the cooling fluid. Note also that the top side of the triangular receiver is capable to absorb both
beam and diffuse radiation, whereas PV layers can only convert the concentrated beam radiation. In
summary, the system is basically the same as the one shown in references [25, 29] differing from
that for two reason: i) there is no covering glass; ii) the P V layer is based on InGaP/InGaAs/Ge
triple-junction solar cells [16]. These two modifications allow one to increase significantly the
electrical efficiency of the system with respect to the values rated in references [25, 29]. In fact, the
covering glass is used to increase the thermal efficiency of the system since it reduces convection
and radiation losses. However, the glass also reduces the radiation incident on the PV layer,
determining a decrease of the electrical efficiency of the system. Then, the triple-junction cells are
significantly more efficient than silicon ones and they are also less sensitive to the variation of the
operating temperature.
293
Although simplified models for the calculation of CPVT performance are available in literature
[33], they cannot be applied to the system under investigation due to the use of concentrating
systems and triple junction cells. Therefore, an appropriate model, based on energy balances, has
been developed in order to design and simulate the CP VT under investigation. This is a 0-D model
since the final scope of this work is to create a new Type to integrate in TRNSYS environment.
Therefore, the model should be sufficiently fast for being used in a quasi-stationary yearly
simulation. Therefore, 1-D models have not been considered since they are too computationalintensive for the scope of that work.
The general assumptions adopted for the model are: thermodynamic equilibrium, steady state,
kinetic and gravitational terms negligible in the energy balances, radiation uniformly concentrated
along PV area. In addition the small thickness of the PV layer and the high conductivity in the metal
substrate allow one to assume negligible temperature gradients in the PV film and in the substrate.
In other words PV and substrate temperature are assumed uniform.
The system was assumed to operate below 100 C, since it is safer for the reliability of PV cells,
although the system can theoretically operate up to 240 C [15]. In this case, the CP VT could drive
a double effect ACH, significantly increasing the overall efficiency of the system. However, this
possibility must still be explored by experimental tests. Therefore, water was assumed as cooling
fluid. Nevertheless, several types of cooling fluids can be implemented in the model. The
thermodynamic and thermo-physical properties of the fluids, namely air and water, were calculated
using the appropriate routine included in TRNSYS.
The concentration ratio is defined as the ratio between the area of the receiver, A PVT, namely the
two PV triangular sides, and the aperture area, A ap, of the concentrator:
C PVT
APVT
Aap
(1)
The optical efficiency ( opt) of the concentrator is assumed being constant [15]. Therefore, the
radiation indent on the PV surface is:
GPVT
APVT Ib CPVT
opt
IAMth
(2)
As usual in concentrating systems, in the previous equation only the beam incident radiation (Ib) is
considered. Such radiation is also corrected considering both the optical efficiency of the receiver
and the Incidence Angle M odifier (IAM ) [34], the last considering that the radiation decreases when
the angle of incidence increases. The IAM , related to the thermal production is evaluated on the
basis of the data experimentally calculated by Bernardo et al.[25, 29]:
60
IAM th 1 b0,th
60 IAM th
1 b0,th
1
cos
1
cos
1
1
1
60
30
(3)
Atop I tot
(4)
top
294
In this case, the top surface area, A top, can convert both beam and diffuse radiation, i.e. the total
radiation (Itot) since the insolation incident on that surface is not concentrated.
Assuming the top surface area as gray surface and considering that the area of the top surface is
much lower than the one of the sky, the radiative heat transfer between the top absorber and the sky
can be calculated as follows [34]:
Qtop
sky
Atop
top
Ttop4 Tsky4
(5)
Here, the sky equivalent temperature (T sky ) is calculated using TRNSYS routine. T top is the
temperature of the top surface.
Similarly, assuming the area of the concentrator much larger than the one of the PVT receiver and
assuming both PVT and concentrator as gray surfaces, the radiative heat transfer between the PVT
and the concentrator [34]:
QPVT
conc
APVT
4
TPVT
PVT
4
Tconc
(6)
T PVT and T conc are respectively PVT and concentrator surfaces temperatures.
The convective heat transfer between the PVT and the air is calculated as follows [35]:
Qconv,P VT
AP VT hc,P VT TP VT
Ta
(7)
The convective heat transfer coefficient, hc,PVT, is calculated taking into account that the wind
velocity is typically around 4-5 m/s. Therefore, the convection mechanism is definitively a forced
convection. Therefore, the corresponding heat transfer coefficient is calculated using the following
correlation, relating the Nusselt, Reynolds and Prandtl numbers [35]:
1
Nu
0.664Pr 3 Re 2
(8)
In this equation the characteristic length is the length of the surface in the wind direction, assumed
parallel to the CPVT longitudinal axis, i.e. L tube . The same correlation is used to calculate the heat
transfer coefficient for the forced convection between the top absorber and the air, hc,top. The
corresponding heat flow is [35]:
Qconv,top
(9)
CPV T AP VT Ib
opt
PV
IAM el
(10)
Note that this energy is calculated considering the concentrated beam radiation (corrected by the
concentrator optical efficiency and by the IAM coefficient) incident on the PV layer, corrected by
the electrical efficiency of the PV, PV. The electrical efficiency of the triple-junction PV ( PV) is
experimentally related to the concentration ratio and to the temperature [15].
PV
TPVT
298
(11)
Note that this equation returns ultra-high values of electrical efficiency, also approaching 40 %, as
usual in III-V PV cells. The IAM el is also evaluated on the basis of the experimental data provided
by Bernardo et al.[25, 29]:
60
IAM el
60 IAM el
1 b0,el
1 b0,el
1
cos
1
cos
1
1
1
(12)
60
30
The net power produced by the system is reduced of the amount of electricity lost in the module
connections and in the inverter, considering the corresponding conversion efficiencies ( mod and
inv) [15].
PPVT ,net
PPVT ,gross
(13)
mod inv
mf hout
hin
(14)
In the previous equation, the enthalpies of the inlet and outlet cooling fluid (hin and hout) are
calculated by the thermo-physical property subroutine discussed above.
Therefore, the overall energy balance on a control volume including the entire triangular receiver is:
APV T I bC PVT
APV T I bC PVT
opt
IAMth
opt
IAMth
Ta
Atop Itot
PVT
mf hout
top
Atop
top
hin
4
Ttop
Tsky4
CPV T APV T I b
APVT
PVT
opt
TP4VT
PV
IAM el
4
Tconc
(15)
Note that in this energy balance the left side is representative of the energy flows entering the
control volume, whereas the terms at the right side of the equation are the energy flows exiting the
control volume. Among these terms at the right side, the first one is the useful thermal energy, the
second one is the electrical power produced and all the remaining terms are losses. As discussed in
the following section, this energy balance is dominated by the radiative terms being the convection
losses low, due to the low receiver area, as typical in concentrating solar collectors
A second energy balance considers the control volume including the metallic substrate and the fluid
channel (also including the fluid flowing inside). In this study, this control volume can be
considered as a heat exchanger. In particular, it is here assumed that the temperature of the metallic
substrate is homogeneous along both radial and circumferential directions. In this case, the primary
side of the heat exchanger is at constant temperature equal to the temperature of the metallic
substrate. This assumption can be considered acceptable as a consequence of the high thermal
conductivity of the metallic substrate [15, 19-20, 25, 29] and allows one to develop a 0-D model of
the CPVT. According to the 0-D approach here implemented, the performance of the heat
exchanger can be calculated using the well-known -NTU technique [36]. For the case under
consideration, the NTU number is:
296
1
1
rsub
h fluid
NTU
AHEX
(16)
mf c f
The heat exchange area, A HEX, is the lateral area of the fluid channel. The thermal resistance of the
metallic substrate, rsub, is typically orders of magnitude lower than the one of the fluid. The fluid
heat transfer coefficient, hfluid, is calculated using the following correlation [35]:
4
0.023Re 5f Pr f5
Nu f
(17)
NTU
(18)
Defined T sub the temperature of the metallic substrate, the energy balance for the considered heat
exchanger is:
mf hout
hin
mf c f Tsub Tin
(19)
Note that, for the given boundary conditions (inlet temperature and mass flow rate, beam and total
radiations and relative angle of incidence, ambient and sky temperature, ambient pressure and wind
velocity), the unknowns are five, namely: PVT temperature, substrate temperature, fluid outlet
temperature, temperature of top receiver surface (facing the sky) and temperature of the
concentrator. Therefore, three further equations, in addition to eqs. (15) and (19), must be
considered.
The third of the required five equations is derived from an energy balance on a control volume
including the PVT layer, and the metallic substrate.
APV T
TPVT Tsub
rPVT sub
m f hout
hin
Atop
Tsub Ttop
(20)
rtop
In other words, the previous equation is showing that the conductive thermal flow coming from the
PVT layer is partly used to increase the outlet temperature of the cooling fluid and partly is
conductively exchanged with the top side of the triangular receiver. Note that the top thermal
resistance, rtop, is the conductive resistance calculated considering both the metallic substrate and
the top absorbing surface included between the fluid channel and the top surface.
A fourth energy balance can be considered with respect to the control volume including the top side
of the substrate and the top surface of the triangular receiver:
Atop
Tsub Ttop
rtop
Atop I top
Atop I top
top
Atop
top
Ttop4 Tsky4
(21)
Finally, the last energy balance considers the control volume including only the parabolic
concentrator.
297
AP VT
PVT
4
TPVT
4
Tconc
I tot Aconc
c onc
Aconc
conc ,back
4
4
Tconc
Tsky
(22)
In this case, the left side of the energy balance includes the radiative heat transfer with the PVT and
the radiative energy absorbed by the concentrator surface. Conversely, the terms at the right side of
the equations are respectively: the radiative heat transfer of the back surface (external side of the
parabola) of the concentrator, the convective losses at the front surface (facing the sun) of the
concentrator, and the convective heat loss at the back surface.
Eqs (15), (19), (20), (21), (22) are a system of five equations in the above mentioned five
unknowns. This system of equations is highly non linear as a consequence of the radiative terms
included in the energy balances and of the correlations for the calculations of heat transfer
coefficients. This system must be solved by conventional numerical iterative techniques.
Note also that the model discussed above lies on the assumption of steady state. However, this
assumption can be easily removed, simply adding the capacitive terms in the five energy balances
described above. In that case, the algebraic system of equations turns in a system of differential
equations that can be easily solved using the tool included in TRNSYS package.
Table 1- CPVT design parameters
Parameter
S ymbol
Value
Unit
Aap
12
m2
Atop
0.60
m2
PV layer area
Fluid channel diameter
Fluid specific heat
Rated fluid flow rate
Top surface absorptance
Concentrator absorptance
Back surface concentrator emissivity
Top surface emissivity
PV reflectance
PV emissivity
IAM electrical coefficient
IAM thermal coefficient
APVT
d
cf
mf
0.12
0.03
4.1877
0.15
0.90
0.03
0.30
0.20
0.03
0.20
0.28
0.14
m2
m
kJ/ kg K
kg/s
top
conc
conc
top
PVT
PVT
b0el
b0th
The overall performance of the CPVT is often evaluated using the well-known thermal and
electrical efficiencies, which are conventionally related to the incident beam radiation and to the
collector aperture area:
CP VT , th
CP VT , el
mf hout hin
Aap I b
CP VT APVT Ib opt
Aap Ib
PV
(23)
IAM el
(24)
298
CPVT design parameters are reported in Table 1[15, 19-20, 25, 29]. For the design parameters
assumed in this table, the concentration ratio is 10.
Tin Tout
2
Ib
Ta
Tf ,avg
Ta
(25)
Ib
The result of this analysis is shown in Fig. 2. Here, it is clearly shown that the correlation between
the thermal efficiency and the above mentioned parameter is very good. In fact, the value of the
correlation index is very high even using a linear interpolating curve. However, it is also clear that
the markers are not perfectly aligned on the interpolating curve. In fact, the thermal efficiency curve
of a solar thermal collector is a second-order polynomial when constant heat transfer coefficients
are considered [34]. Therefore, the deviation of the performance calculated by the model discussed
in the previous section and the interpolating curve is basically due to the fact that, in the proposed
model, the overall heat transfer coefficients is not constant and varies with the CPVT operating
conditions.
el,PVT,
0.25
0.6
0.24
0.5
0.23
0.4
0.3
el,PVT
t,PVT
0.7
el,PVT
t,PVT
0.21
0.2
0.1
0.19
0.1
0.22
0.2
0
0
0.8
0.18
20
40
60
80 100
T mean ,f [C]
120
140
160
Conversely, there is not any good correlation between the electrical efficiency and the parameter
shown in eq. (25). In fact, the electrical efficiency of the CPVT basically depends on the operating
temperature of the PV layer and on the concentration ratio, as shown in eq. (11). In the case of the
electrical efficiency it is a common use to show this parameter as a function of fluid average
temperature [21, 25, 29]. This graph is shown in Fig. 3 where the linear interpolating curve is also
displayed. Such Figure shows that the electrical efficiency basically linearly depends on the fluid
mean temperature since such temperature directly affects the PV layer temperature. Note also that
the electrical efficiency scarcely depends on the incident beam radiation. In fact, a reduction of the
beam radiation, for a given fluid average temperature, only determines a slight decrease of PV layer
temperature, consequently causing a slight increase in CPVT electrical efficiency. The plots of Fig.
2 and Fig. 3 also show that the efficiency curve of the proposed CPVT is slightly different from the
original model of the Bernardo et al. [25, 29] due to: the use of triple-junction PV and to the
absence of the covering glass. As a consequence, the optical and the electrical efficiencies are
higher. In fact, this last efficiency ranges between 20 % and 25 % for the considered operating
points. The use of a covering glass reduces the amount of radiation available for both electrical and
299
thermal conversions but also reduce thermal losses. Note also that a higher electrical efficiency is
often counterbalanced by a lower value of the thermal efficiency. However, the thermal
performance of the CPVT considered in this work is also very good due to the excellent radiative
properties of the considered surfaces. Therefore, it can be concluded that the modifications in the
layout of the CPVT considered in this work, allow one to increase both thermal and electrical
efficiency with respect to the values published by Bernardo et al. [25, 29]. Obviously, this increase
in efficiency is counterbalanced by a significantly higher capital cost due to the use of triplejunction PV layer. Therefore, the system considered in this work may be considered a suitable
option for the next future when the cost of triple-junction PV is expected to dramatically decrease.
In the following paragraphs a brief parametric study is performed, with the scope to analyze the
performance of the CPVT under different operating condition and varying some of its main design
parameters. The study was performed considering the set of boundary conditions shown in Table 2
This parametric analysis aims at evaluating the effect of the variation of the design parameters on
the CPVT thermal and electrical performance.
Table 2- CPVT boundary conditions
Parameter Value Unit Parameter Value Unit
T in
70
C
wa
5
m/s
T sky
25
C
Itot
1000 W/m2
Ta
25
C
Ib
800 W/m2
pa
101
kPa
0
deg
As mentioned above, the layout of the system under investigation was derived by some prototypes
previously developed and discussed in literature [8, 21, 25, 29] and has been here modified in order
to improve the electrical performance of the system and to achieve a good thermal performance at
operating temperatures up to 90-100 C. To this scope, in the following figures some of the main
geometrical parameters were varied from the initial configuration, in order to assess their impact on
the overall performance of the system. First, the axial length - Ltube - of the CPVT (including the
axial lengths of the triangular receiver and of the parabolic concentrator) was varied in very large
range. Fig. 4 shows both thermal and electrical efficiencies as a function of this parameter. Here, it
is clearly displayed that both efficiencies are very slightly affected by the increase of CPVT length,
showing a slight decrease for higher CP VT length. However, thermal and electrical efficiencies
drops are lower than 1%. Obviously, an increase of CP VT length also determines a proportional
increase of CPVT aperture area and of PV layer area. This linear relationship is clearly shown in
Fig. 5, where both thermal and electrical powers are plotted as a function of the CP VT length. Note
also that the slight decrease of efficiencies shown in Fig. 4 also determines a deviation of the plots
of Fig. 5 from an ideal proportional relationship. In fact, for a CP VT length of 4 m, thermal and
electrical powers are respectively 8320 kJ/h and 4531 kJ/h, whereas at 16 m these values increase
up to 33403 kJ/h and 15178 kJ/h, which are values slightly lower than the ones expected in case of
constant thermal and electrical efficiencies. The reason of this decrease may be explained by the
temperature profiles shown in Fig. 6. Here, it is shown that higher lengths of the CPVT system also
determine a slight increase of all the temperatures. In fact, an increase in the length also causes an
increase of fluid outlet temperature that consequently determines an overall increase of the
temperatures of the system. In particular, the increase in P V temperature determines the decrease of
the electrical efficiency. M oreover, the increase of PV and top side temperatures also determines an
increase of both convective and radiative losses causing the slight reduction of thermal efficiency
shown in the previous figure. Therefore, from this analysis it could be concluded that, for
maximizing the efficiency of the CPVT the length of the system should be small and the
temperature increase of the fluid should be also small. This result is in accordance with the expected
behaviour of the system since an increase of the fluid mean temperature generally determines a
decrease of both thermal and electrical performances.
300
e l,PVT
t,PVT
0.218
0.217
0.216
0.215
0.214
4
12 [m]
Ltube
16
x 10
4.5
16
4
3.5
3
P el
Qt
12
2.5
2
1.5
8
4
8
12
Ltube [m]
16
Q t [kJ/h]
el,PVT
0.219
x 10
20
Pel [kJ/h]
0.22
0.609
0.608
0.607
0.606
0.605
0.604
0.603
0.602
0.601
0.6
20
t,PVT
0.221
1
0.5
0
20
83. 5
Ttop
29.35
TPVT
To
29.3
Tconc
29.25
29.2
79
29.15
12
Ltube [m]
16
0. 2182
0.218
0. 2178
0. 6053
0. 605
0. 6048
0.217
0. 2168
0.01
29.05
70
0
0. 6058
0. 6055
0. 2176
0. 2174
0. 2172
29.1
74. 5
0. 2186
0. 2184
29
20
el,PVT
88
29.4
Tsu b
Tconc [C]
92. 5
e l,PVT
0. 6045
0. 6043
0. 604
t,PVT
0.02
0.03
0.04
d [m]
0.05
t,PVT
0. 6038
0. 6035
0.06
electrical
The variation of the diameter of the fluid channel does not significantly affect the overall
performance of the system, being important only for the internal balance of the system. In fact, Fig.
7 and Fig. 8 clearly show that an increase in channel diameter determines a slight decrease of
electrical and thermal efficiencies. A similar decrease is also observed for both thermal and
electrical powers produced by the CPVT. Conversely a variation in the diameter shows a significant
impact on the temperatures of the layers of the CP VT, as shown in Fig. 9. Here, it is clearly shown
that an increase of the diameter determines a general increase of the temperature of all the solid
layers of the CPVT, whereas the outlet temperature of the fluid does not significantly vary. In fact,
the lower the diameter, the higher the velocity of the fluid also determining a corresponding
increase of its heat transfer coefficient. Simultaneously, the higher the diameter, the higher the heat
exchange area. The two contrasting effects determine the temperature plot shown in Fig. 9.
2.093
2.092
9.3
2.091
9. 29
9. 28
9. 27
2.09
P el
2.089
Qt
2.088
9. 26
9. 25
0.01
Qt [kJ/h]
Pel [kJ/h]
9. 31
86
2.087
0.02
0. 03
0.04
d [m]
0.05
85
84
To
TPVT
Tc onc
29.22
Tsub
29.2
83
29.18
82
29.16
81
29.14
80
79
0.01
2.086
0.06
29.24
Ttop
Tconc [C]
x 10
2.094
x 10
9. 32
0.02
0.03
0.04
d [m]
0.05
29.12
0.06
e l,PVT
0. 22
0.215
0. 21
0.205
0.2
0.04
0.08
0.12
Lrec [m]
0.16
9.8
Pel
2.16
9.6
Qt
2.14
2.12
9.4
2.1
9.2
2.08
2.06
8.8
2.04
8.6
0.04
31.5
30.5
82
30
81
29.5
80
29
79
78
0.04
0.08
0.12
L rec [m]
0.63
0.22
0.218
0.16
0.216
el,PVT
0.625
t,P VT
0.62
0.615
0.61
0.214
0.605
0.212
28.5
0.21
0.6
28
0.2
2.02
0.2
31
83
0.16
el,PVT
84
0.12
Lrec [m]
0.222
32
Ttop
T pl
TP VT
To
Tc onc
85
T conc [C]
86
0.08
Qt [kJ/h]
t,PVT
x 104
2.18
t,PVT
el ,PVT
0.225
x 103
10
P el [kJ/h]
0. 23
0.63
0.625
0.62
0.615
0.61
0.605
0.6
0.595
0.59
0.585
0. 2
t,PVT
0.235
0.6
0.8
1.8
0.595
2
A further design parameter is the length of each of the three sides of the triangular receiver, Lrec.
The variation of such parameter is very important since affects: the concentration ratio, PV area and
top absorber area. The length of the receiver significantly affects both thermal and electrical
efficiencies, as shown in Fig. 10 displaying that an increase in receiver length determines a decrease
of the electrical efficiency and a simultaneous increase of the thermal efficiency. Similar trends are
also shown for the thermal and electrical powers (Fig. 11). An increase of the length of the receiver
determines a proportional decrease of the concentration ratio, reducing the radiative flow incident
on the PV layer. Such phenomenon determines, as expected, a decrease of P V layer temperature
(Fig. 12). A reduction of the PVT temperature would suggest an increase of the electrical efficiency,
in contrast to the trend shown in Fig. 10. This is due to the fact that the electrical efficiency of the
triple-junction PV layer also depends on the concentration ratio, as shown in eq. (11). In particular,
a reduction of the concentration ratio also determines a decrease of the of the PV electrical
efficiency. This effect is dominant over the decrease of PV temperature, determining the overall
result of a significant decrease of P V electrical efficiency, as shown in Fig. 10. Conversely, the
thermal efficiency increases faster than the electrical efficiency decreases. In fact, the thermal
efficiency increases for two simultaneous effects: i) a lower electrical efficiency makes more heat
available for thermal conversion; ii) an increase of top absorber area improves the utilization of the
total radiation. This is also clear by Fig. 12, showing that an increase of the temperature length
determines an increase of the top surface temperature, causing also an increase of the outlet
temperature of the fluid. Therefore, it may be concluded that the length of the receivers plays an
important role in the design of the CPVT system. In case the goal is the maximization of the
electrical efficiency this parameter should be as low as possible. Conversely, it should very high if
the goal is the increase of the thermal efficiency.
302
92.5
Qt
2.5
10
2
1.5
6
4
0.6
Q t [kJ/h]
Pel [kJ/h]
12
0.8
1.2 1.4
L ap [m]
1.6
1.8
1
2
0.2181
el,PVT
el,PVT
0.2181
t,PVT
0.218
0.218
0.2179
0
2.2
4.4
6.6
rtop [(m 2*C)/W]
8.8
TPVT
To
Tsub
92
70
0
4.4
6.6
2
rtop [(m *C)/W]
8.8
1.6
28.8
2
1.8
2.08
Pe l
Qt
2.06
2.04
9304
9298
0
2.02
2
1.98
2.2
4.4
rtop
6.6
[(m *C)/W]
2
1.96
11
8.8
0.5 4
0.5 2
0.5
29.14
2.2
1.2 1.4
Lap [m]
9308
0.5 6
29.15
x 10
2.1
29.16
Tcon c
114
0.8
9310
0.5 8
29.15
29
29.16
29.16
29.2
76.5
t,PVT
136
Ttop
29.4
80.5
9300
Tconc [C]
158
29.6
9302
29.8
Tconc
9306
t,PVT
0.2182
0.61
0.605
0.6
0.595
0.59
0.585
0.58
0.575
0.57
0.565
11
Tsub
84.5
72.5
0.6
88.5
To
TP VT
Qt [kJ/h]
P el
14
30
Ttop
Tconc [C]
x 10
3.5
Pel [kJ/h]
x 10
16
600 W /m2
t ,P VT, It ot =Ib=
W/m 2
400
300 W/m2
200 W/m2
0.4 8
29.14
11
0.4 6
0
4
6
2
r to p [ (m *C)/W]
10
Opposite results are achieved varying the aperture length, i.e. the length of the aperture area in the
radial direction. In fact, an increase of the aperture length determines a decrease of the thermal
efficiency and a simultaneous increase of the electrical efficiency (Fig. 13). In addition, both
thermal and electrical powers increase proportionally with the aperture length (Fig. 14). Note that
the decrease of the thermal efficiency is much larger than the increase in the electrical one. In fact,
the thermal efficiency basically decreases as a consequence of the significant increase of both
radiative and convective losses, due to the increase of both PV and top temperatures (Fig. 15). A
temperature increase of the layers of the CPVT receiver is due to the increase of the radiative flow
to the higher concentration ratio. This increase of the concentration ratio would also determine an
increase of the electrical efficiency, as discussed before. However, this positive effect is partly
303
counterbalanced by the increase in PVT temperature (Fig. 15), resulting in a slight overall increase
of the electrical efficiency. Note also that this particular trend is achieved as a consequence of the
high efficiency PV layer considered in this work. If a conventional silicon P V layer would be
considered, the electrical efficiency would have dramatically reduced.
A possible alternative to the CPVT layout discussed in this paper may consist in a thermal
insulation of the top surface of the receiver [20]. This arrangement could reduce thermal losses but
would also reduce the radiative heat absorbed by that surface. In order to analyze this possibility,
the top thermal resistance was varied in a very wide range, whose lower bound could be
representative of no insulation, whereas the upper bound is achieved only in case of good thermal
insulation. The results of this analysis showed that both thermal efficiency and thermal power
dramatically decrease with higher insulation (Fig. 16 and Fig. 17). Conversely, the variations of
electrical efficiency and electrical powers are marginal (Fig. 16 and Fig. 17). This is due to the fact
that an increase of top insulation determines a dramatic increase of top surface temperature (Fig. 18)
whereas the temperatures of the remaining layers of the CPVT do not significantly vary (Fig. 18
shows only a slight decrease). In fact, a higher insulation inhibits the heat transfer from top surface
to the cooling fluid, determining only an increase of the top surface temperature. As a consequence,
increasing the insulation, the fluid is not capable to absorb the radiative heat coming from the top
surface. This results in a local increase of the temperature in the top surface and in a general
decrease of the CPVT temperatures (including the fluid outlet temperature) determining the
decrease of thermal efficiency shown in Fig. 16. Therefore, the results of this study would suggest
that, for the boundary conditions assumed in this study, the option of insulating the top surface is
not profitable for the layout under investigation. However, these results dramatically change if
different values of irradiation are considered. This is clearly shown in Fig. 19, where the thermal
efficiency is plotted for different values of radiation and top resistance. Here, it is clearly shown that
for low irradiation, the thermal efficiency increases in case of higher insulation whereas the
opposite trend occurs in case of higher solar radiation. However, this type of CPVT is typically used
for locations in which the average solar radiation is high, where the top insulation would not be
profitable.
4. Conclusions
The results of the case study presented prove the technical feasibility of the novel arrangement of a
concentrating parabolic trough P VT solar collector based on triple-junction technology. Results
showed that:
Both thermal and electrical efficiencies are very high for a wide range of operating
conditions.
The system is particularly sensitive to the available beam radiation since both thermal and
electrical efficiencies dramatically decrease for low beam insolation. Therefore, the results
of the model suggest that this kind of device can be profitably adopted for the locations in
which the average beam radiation is sufficiently high.
An increase in CPVT length and/or in fluid channel diameter determines a decrease in both
thermal and electrical efficiencies, due to the increase of CPVT operating temperature,
causing an increase of thermal losses and in P VT electrical efficiencies.
Decreasing receiver length and/or increasing aperture length determine an increase of the
concentration ratio. This is favourable for the electrical efficiency which increases for higher
concentration ratios. Conversely, such modification also determines a general increase in
CPVT operating temperature, negatively affecting its thermal efficiency.
Insulating the top surface rather than using a high-absorbitivity surface is recommended for
increasing the CP VT electrical efficiency. On the other hand, a simultaneous decrease of the
304
thermal efficiency must be taken also into account, due to lower radiation absorbed by the
collector.
The analysis has been performed by a simulation model which allows one to change the design
parameters of the system and to evaluate the corresponding thermodynamic performance. The
model also allows one to perform dynamic simulations and system optimization, defining the set of
the design parameters maximizing system performance.
On the other hand, it must be pointed out that the system investigated in this study is very expensive
due to the use of triple-junction PV cells. Therefore, the use of this technology is presently scarcely
profitable from an economic point of view. However, a possible market penetration of this
technology would make this CPVT an excellent system for providing electricity and heat for several
applications, such as: industry, residential, solar heating and cooling, integration in ORC cycles, etc.
In the Part II of this paper a specific application of this system will be investigated, i.e. the
integration of the CPVT in a Solar Heating and Cooling systems providing space heating, space
cooling, domestic hot water and electricity to the user.
Future developments of this work will include the investigation of dish concentrating PVT systems
for higher temperature applications.
Nomenclature
Area [m2]
Specific Heat [J/kg/K]
CPV T
Concentration ratio
D
GP VT
hc
hf
Ib
I tot
IAM
k
L
mf
Nu
p
Nusselt Number
Pressure [kPa]
PPVT
Pr
Prandtl Number
Q
r
Re
305
T
UA
Temperature [K]
Overall heat transfer coefficient [W/K]
w
velocity [m/s]
Greek Symbols
Absorptance
Emittance
Density [kg/m3]
PV T
PVT Reflectance
Stephan-Botzmann constant
Viscosity [kg/m/s]
inv
Inverter efficiency
mod
M odule efficiency
PV
PV efficiency
opt
Optical efficiency
th
Thermal efficiency
el
Electrical efficiency
ex
Exergetic efficiency
Subscripts
a
ap
atm
conc
el
f
hor
in
Ambient
Aperture
Atmospheric
Concentrator
Electrical
Fluid
top
Horizontal
Inlet
Outlet
Receiver
Top surface
th
Thermal
out
rec
References
[1] Chow, T.T., A review on photovoltaic/thermal hybrid solar technology Applied Energy, 2010.
87(2): p. 365-379.
[2] Zondag, H.A., Flat-plate PV-Thermal collectors and systems: A review Renewable and
Sustainable Energy Reviews, 2008. 12(4): p. 891-895.
[3] Calise, F., Palombo, A., Vanoli, L., , Design and dynamic simulation of a novel solar
trigeneration system based on photovoltaic/thermal collectors, in ECOS 2011. 2011: Novi Sad,
Serbia.
[4] Skoplaki, E., Palyvos, J.A. , On the temperature dependence of photovoltaic module electrical
performance: A review of efficiency/power correlations Solar Energy, 2009. 83: p. 614-624.
306
[5] Fong, K.F., Chow, T.T., Lee, C.K., Lin, Z., Chan, L.S., Advancement of solar desiccant
cooling system for building use in subtropical Hong Kong Energy and Buildings, 2010. 42: p.
2386-2399.
[6] Beccali, M ., Finocchiaro, P., Nocke, B., Energy and economic assessment of desiccant cooling
systems coupled with single glazed air and hybrid PV/thermal solar collectors for applications
in hot and humid climate Solar Energy, 2009. 83: p. 1828-1846.
[7] Ibrahim, A., Othman, M ., Ruslan, M .H.,/ M at, S., Sopian, K., Recent advances in flat plate
photovoltaic/thermal (PV/T) solar collectors Renewable and Sustainable Energy Reviews,
2011. 15: p. 352-365.
[8] Zhang, Xingxing, Zhao, Xudong, Smith, Stefan, Xu, Jihuan, Yu, Xiaotong, Review of R&D
progress and practical application of the solar photovoltaic/thermal (PV/T) technologies
Renewable and Sustainable Energy Reviews, 2011. In Press, Corrected Proof,Sep 2011
[9] Tripanagnostopoulos, Y., Aspects and improvements of hybrid photovoltaic/thermal solar
energy systems Solar Energy, 2007. 81(9): p. 1117-1131.
[10] Daghigh, R. , Ruslan, M .H. ,Sopian, K., , Advances in liquid based photovoltaic/thermal
(PV/T) collectors Renewable and Sustainable Energy Reviews, 2011. 15(8): p. 4156-4170.
[11] Kumar, Rakesh, Rosen, M arc A., , A critical review of photovoltaicthermal solar collectors
for air heating Applied Energy, 2011. 88(11): p. 3603-3614.
[12] Zhao, J., Song, Y., Lam, W.H., Liu, W., Liu, Y., Zhang, Y., Wang, D.Y., Solar radiation
transfer and performance analysis of an optimum photovoltaic/thermal system Energy
Conversion and M anagement, 2011. 52: p. 1343-1353.
[13] Joshi, A.S., Tiwari, A., Tiwari, G.N., Dincer, I., Reddy, B.V., Performance evaluation of a
hybrid photovoltaic thermal (PV/T) (glass-to-glass) system International Journal of Thermal
Sciences, 2009. 48: p. 154-164.
[14] Daghigh, Ronak, Ibrahim, Adnan, Jin, Goh Li, Ruslan, M ohd Hafidz , Sopian,
Kamaruzzaman, Predicting the performance of amorphous and crystalline silicon based
photovoltaic solar thermal collectors Energy Conversion and M anagement, 2011. 52(3): p. 411747.
[15] M ittelman, G., Kribus, A., Dayan, A., Solar cooling with concentrating
photovoltaic/thermal (CPVT) systems Energy Conversion and M anagement, 2007. 48: p. 24812490.
[16] Nishioka, K., Takamoto, T., Agui, T., Kaneiwa, M ., Uraoka, Y., Fuyuki, T., Annual output
estimation of concentrator photovoltaic systems using high-efficiency InGaP/InGaAs/Ge triplejunction solar cells based on experimental solar cell's characteristics and field-test
meteorological data. Solar Energy M aterials and Solar Cells, 2006. 90: p. 57-67.
[17] Vorobiev, Y. , Gonzlez-Hernndez, J. , Vorobiev, P. , Bulat, L., Thermal-photovoltaic
solar hybrid system for efficient solar energy conversion Solar Energy, 2006. 80: p. 170-176.
[18] Kosmadakis, G. / M anolakos, D. / Papadakis, G., Simulation and economic analysis of a
CPV/thermal system coupled with an organic Rankine cycle for increased power generation
Solar Energy, 2011. 85: p. 308-324.
[19] M ittelman, Gur / Kribus, Abraham / M ouchtar, Ornit / Dayan, Abraham, Water desalination
with concentrating photovoltaic/thermal (CPVT) systems Solar Energy, 2007. 83(8): p. 13221334.
[20] Kribus, Abraham / Kaftori, Daniel / Mittelman, Gur / Hirshfeld, Amir / Flitsanov, Yuri /
Dayan, Abraham, A miniature concentrating photovoltaic and thermal system Energy
Conversion and M anagement, 2006. 47(20): p. 3582-3590.
[21] Coventry, Joe S., Performance of a concentrating photovoltaic/thermal solar collector Solar
Energy, 2005. 78(2): p. 211-222.
[22] Li, M ., Li, G.L. , Ji, X. ,Yin, F. , Xu, L., The performance analysis of the Trough
Concentrating Solar Photovoltaic/Thermal system Energy Conversion and M anagement, 2011.
52(6): p. 2378-2383.
307
[23]
Li, M ing, Ji, Xu , Li, Guoliang,Wei, Shengxian , Li, YingFeng , Shi, Feng, Performance
study of solar cell arrays based on a Trough Concentrating Photovoltaic/Thermal system
Applied Energy, 2011. 88(9): p. 3218-3227.
[24] Li, M . , Ji, X. , Li, G.L., Yang, Z.M . , Wei, S.X. , Wang, L.L, Performance investigation
and optimization of the Trough Concentrating Photovoltaic/Thermal system Solar Energy,
2011. 85(5): p. 1028-1034.
[25] Bernardo, L.R., Perers, B.,Hkansson, H. ,Karlsson, B., Performance evaluation of low
concentrating photovoltaic/thermal systems: A case study from Sweden Solar Energy, 2011.
85(7): p. 1499-1510.
[26] Xu, Guoying / Zhang, Xiaosong / Deng, Shiming, Experimental study on the operating
characteristics of a novel low-concentrating solar photovoltaic/thermal integrated heat pump
water heating system Applied Thermal Engineering, 2011. 31(17-18): p. 3689-3695.
[27] Al-Alili, A., Hwang, Y., Radermacher, R. / Kubo, I., A high efficiency solar air conditioner
using concentrating photovoltaic/thermal collectors Applied Energy, 2011. In Press,
Corrected Proof,May 2011.
[28] Rosell, J.I. / Vallverd, X. / Lechn, M .A. / Ibez, M ., , Design and simulation of a low
concentrating photovoltaic/thermal system Energy Conversion and M anagement, 2005. 46(1819): p. 3034-3046.
[29] Bernardo, Ricardo., Retrofitted Solar Thermal System for Domestic Hot Water for Single
Family Electrically Heated Houses, Development and Testing, in Faculty of Engineering. 2010,
Lund University.
[30] Calise, F., Design Of A Hybrid Polygeneration System With Solar Collectors And A Solid
Oxide Fuel Cell: Dynamic Simulation And Economic Assessment. International Journal of
Hydrogen Energy, 2011. 36(10): p. 6128-6150
[31] Calise, F., High Temperature Solar Heating and Cooling Systems for Different
Mediterranean Climates: dynamic simulation and economic assessment Applied Thermal
Engineering, 2011. 32: p. 108-124.
[32] Calise F., Dentice d'Accadia M ., Palombo, A., Vanoli L., Dynamic Simulation Of High
Temperature Solar Heating And Cooling Systems, in Eurosun 2010. 2010: Graz (AU).
[33] Florschuetz, L. W. , Extension of the Hottel-Whillier model to the analysis of combined
photovoltaic/thermal flat plate collectors. Solar Energy, 1979. 22: p. 361-366.
[34] Duffie, J.A., Backman, W.A, Solar engineering of thermal processes, ed. John Wiley and
sons. 1980.
[35] Incropera, F.P., DeWitt , D.P., Fundamentals of Heat and Mass Transfer, 5th Edition ed.
Inc. John Wiley & Sons. 2001.
[36] Kakac, S., Liu, H., Heat Exchanger Selection, Rating, And Thermal Design., ed. CRC Press.
1998.
308
DETEC- Univ. of Naples Federico II, P.le Tecchio 80, 80125 Naples, Italy, frcalise@unina.it
DIT Univ. of Naples Parthenope C. D. IS.5, 80143 Naples, Italy, laura.vanoli@uniparthenope.it
Abstract:
During the last few years, the technology of photovoltaic/thermal (PVT) solar collectors is becoming more
and more attractive. In fact, these systems allow one to produce simultaneously electrical energy and
thermal energy by solar irradiation. In particular, PVT systems are very attractive when they are able to
produce medium-temperature heat. The majority of PVT systems presently under development produce hot
stream at temperatures lower than 45 C. This is determined by the fact the electrical efficiency of the PVT
dramatically decreases for higher operating temperatures. However, medium-temperature PVT systems can
be designed using two technical devices: i) multi-junction PV collectors which can achieve very high
electrical efficiencies even when their operating temperature is high; ii) concentrating devices which can
concentrate the beam radiation on the PV panel. The combination of these two devices allows one to design
an efficient medium-temperature PVT solar collector. In particular, in this work the collector considered,
consisted in a parabolic-trough concentrator and a triangular receiver, placed on the focus of the parabola,
on which a multi-junction PV panel is laminated. The triangular receiver is equipped with an internal tube for
the cooling fluid. The system is also equipped with a one-axis tracking system, typical of Parabolic-Trough
Solar Thermal Collectors. The system was accurately simulated by means of detailed energy balances and
calculations of heat transfer coefficients. The model of this innovative device was subsequently integrated
with a transient model of a solar tri-generation system based on PVT collectors and solar heating and
cooling technologies, previously developed by the authors. Although concentrating PVT collectors are also
capable to produce hot streams up to 240 C, while maintaining reasonable electrical efficiency, in this study
an operating temperature below 100 C was considered, since presently no PVT collector operating at higher
temperature is commercially available. The peculiarities of concentrating PVT are specially attractive for
integration in Solar Heating and Cooling system (SHC), where PVT can be used as an auxiliary system
driving the absorption chiller. The combination of concentrating PVT and SHC are studied in this work,
implementing a novel virtually zero-emission polygeneration system able to produce electricity, space
heating and cooling and domestic hot water. The polygeneration system is based on the following main
components: concentrating parabolic PVT collectors, a single-stage LiBr-H2O absorption chiller, tanks,
auxiliary heaters, balance of plant devices and the building. The PVT produces electrical energy which is in
part consumed by the building lights and equipments, in part is used by the system parasitic loads and the
rest is eventually sold to the grid. Simultaneously, the PVT provides the heat required to drive the absorption
chiller. The system was designed and then simulated by means of a zero-dimensional transient simulation
model. The economic results show that the system under investigation can be profitable, if a proper funding
policy is available. In any case, the overall energetic and economic results are more encouraging than the
ones previously published in literature for similar polygeneration systems
Keywords:
PVT, solar heating and cooling, solar energy, TRNSYS
1. Introduction
The Part I of this paper presents the technology of Photovoltaic Thermal (PVT) solar collectors,
producing simultaneously thermal and electrical energy. In particular, Part I of the paper presents a
simulation model of a novel concentrating P VT (CPVT) based on parabolic trough concentrators
and triple-junction photovoltaic (PV) receiver. The Part II of the paper investigates a very
interesting application of the CPVT collector presented in Part I, consisting in its integration in a
309
Solar Trigeneration system. In fact, the final scope of this work is the analysis of the possible
integration of P VT in Solar Heating and Cooling (SHC) systems, designing a novel solar
trigeneration system producing heat, cool and electricity for an educational building.
A SHC system is conventionally based on solar thermal collectors producing heat. This heat is used
during the winter for space heating and domestic hot water, whereas in the summer it is converted
in cooling energy by a thermally-driven (absorption, adsorption, etc.) chiller [1-6]. The majority of
the SHC systems investigated in literature are based on the combination of evacuated tube solar
collectors and single-stage absorption chillers [1-16]. However, in the last few years, the
combination of concentrating solar collectors and double-stage chillers is becoming more and more
attractive [17-22]. Replacing solar thermal by PVT collectors caused an additional production of
electricity by the PVT.
As discussed above and in Part I, literature review showed a large number of papers investigating
separately PVT and SHC systems. However, the analysis of these two technologies in a single
polygeneration system is scarcely investigated. In particular, the literature review revealed only two
works investigating the theoretical feasibility of the integration of SHC and P VT. The first work
was presented by Vokas et al. [23]. Here, the theoretical feasibility of flat plate PVT, producing
both space heating and cooling (by an absorption chiller), for a domestic user is investigated. This
preliminary work was based on a simplified approach (F-charts) for the system simulation; in
addition, suitable system layout and control strategies are not applied. Anyway, the considered
system may cover a significant amount of domestic heating and cooling demand [23]. A second
work on this topic was recently presented by M ittelman et al. [24]. The authors of this study
investigated the theoretical feasibility of integrating concentrating P VT, based on triple-junction
cells, with SHC based on a single-effect LiBr-H 2O absorption chiller. The presented system layout
is very simple and it does not include storage tank, auxiliary heater and/or control equipments. The
electrical efficiency ranged approximately between 19 and 23% (in function of P VT operating
temperature, varying between 65 and 120 C). On the other hand, the thermal efficiency of the PVT
was stably slightly lower than 60%. The authors concluded that the reduction of PVT electrical
efficiency, due to the increase in its operating temperature, was not significant. They also showed
that the system was economically profitable (under certain conditions) and improvable optimizing
the system layout (e.g. including a storage tank) or performing a sensitivity analysis in function of
the climatic area and energy costs [24]. An additional paper about this topic was recently presented
by the authors [25], investigating the possibility of integrating "tube and sheet" flat PVT system in
previously developed SHC layouts [19, 26]. The results of the simulations, developed TRNSYS
environment [27], showed that the selected type of PVT dramatically suffer of the high operating
temperature required to drive to absorption chiller. Therefore, the authors concluded that a possible
improvement of that layout may consist in CPVT system, based on multi-junction PV modules,
which are less sensitive to the increase of temperature. The use of a concentrating system is
recommended in order to reduce the high capital cost per unit area of the multi-junction modules.
This is improvement has been here implemented by the authors, aiming at designing and analysing
a novel trigeneration solar system based on CPVT technology equipped with multi-junction PV
modules. In fact, apart from the works of M ittelman [24] and the one recently presented by the
authors [25], none of the papers found in literature analyze the possibility of integrating CP VT or
PVT in SHC systems. In particular, this work is an improvement of the one recently presented by
the authors [25], in which the flat plate tube and sheet PVT has been replaced by a parabolic CP VT
including multi-junction PV cells, with the scope to improve its thermal and electrical performance
at the operating temperature demanded by the SHC. As discussed in Part I, a new thermodynamic
model, based on mass and energy balances, has been implemented by the authors for the simulation
of a novel configuration of parabolic CPVT. This model has been integrating in TRNSYS
environment creating a new type that is included in the layout discussed in reference [25]. Finally,
310
the study also includes energy and cost analyses in order to evaluate the conditions for system
convenience.
2. System Layout
In the following, for sake of brevity, the investigated polygeneration system is named P VSHC. The
considered system layout is schematically shown in Fig. 1. Here six different loops are shown for
the considered fluids: Solar Collector Fluid, SCF (water); Hot Fluid, HF (water); Hot Water, HW;
Cooling Water, CW; Domestic Hot Water, DHW; Chilled or Hot Water, CHW.
CP VT: Concentrating Parabo lic Trough Pho tovoltaic Thermal Collector
TK: Tank
ACH: Single-stage Abso rption Chiller
AH: Auxiliary gas-fired heater
CT: Cooling Tower
HE: Heat Exchanger
HS: Hydraulic Separator
SCF
HF
CW
CHW
DHW
HW
TK1
CPVT
P1
HE1
D
AH
ACH
D
HE2
P2
from zone
fancoil
to zone
fancoil
P3 P4
P4
CT
A1
A2
A3
B1
C1
C2
C3
A1
A2
A3
B1
C1
C2
C3
HS
TK2
a plate-fin heat exchanger (HE2) in the HW loop, transferring the heat from the HF to the hot
water (CHW) to be supplied to the fan-coils during the winter;
a plate-fin heat exchanger in the solar loop, used to produce Domestic Hot Water (HE1) when
the solar irradiation is higher than ACH (or HE2) thermal demand;
some Balance of the Plant (BOP) equipments (the majority not displayed in Fig. 1, for sake of
simplicity), such as pipes, mixers, diverters, valves, and controllers required for the system
operations.
A number of additional mandatory components (not displayed in Fig. 1) have been also
implemented in order to run the simulations and to process the data, such as: controllers (feedback,
proportional and on/off), schedulers (daily and seasonal), weather databases, printers, integrators,
etc.
Details regarding the operating principle and the controls strategies of the system are extensively
provided in [25]. Note also that the system investigated in this paper is based on a Parabolic Trough
CPVT equipped with a single axis system, similar to the one discussed in references [19, 26, 28].
Note also that this is a concentrating system. Therefore, it can convert only the beam fraction of the
total solar insolation. This is only partially compensated by the one axis tracking system [19, 26,
28].
The considered PVSHC provides electricity, space heating and cooling and domestic hot water to
the building. The same building test case (university building with thermal demand for a fitness
centre) used in previous analyses [4, 26, 29] was here considered. Thus, it is possible to compare
the results of this study with those achieved by the previously analyzed systems [4, 26, 29]. Data
regarding building walls, occupancy, equipments and loads are diffusely discussed in reference [4],
whereas the design data of the components are given in reference [19]. The duration curve of the
building electric load is given in reference [29]. The considered DHW daily demand was set at 25
m3/day at 45 C.
3. Simulation model
The PVSHC polygeneration system described in the previous section was dynamically simulated by
TRNSYS, which is a well-known software diffusely adopted for both commercial and academic
purposes. The TRNSYS software includes a large library of built-in components, often validated by
experimental data [27]. As above mentioned, the P VSHC layout investigated in this paper was
originated from the layout developed in previous works [3-5, 19, 26, 29], where the models of both
built-in and user-developed components are described in detail. As discussed in references [1-5, 19,
29], the majority of the models of the components (e.g. pumps, mixers, diverters, valves,
controllers, auxiliary heater, absorption chiller, cooling tower, plate-fin heat exchanger, building,
etc.) were taken from TRNSYS library whereas some new models (defined types, in TRNSYS)
were developed by the authors in Fortran and then linked to TRNSYS (e.g. DHW heat exchangers,
Hydraulic Separator, Fan coils, Primary Energy Calculator, Economic Costs Calculator, CPVT,
etc). In particular, a new TRNSYS type has been created by the authors in order to simulate the
CPVT under investigation. This new model of the CPVT is discussed in detail in Part I of the paper
where details regarding the modelling approach are also provided. As regards the remaining new
TRNSYS types developed by the authors, the reader is referred to references [3-5, 19, 26, 29] for
further details. Conversely, the present part of the paper (Part II) is focused on the integration of the
CPVT in a SHC layout.
A complete validation of the whole trigeneration system is not possible since, in authors'
knowledge, a similar prototype has never been experimented. However, all models of the
components included in the systems are very reliable. In fact, the components included in TRNSYS
312
library are validated against experimental data. Similarly, all the components previously developed
by the authors are based on experimental data provided by the manufacturers.
1.78E+08
Qc,ACH
6.48E+08
2.25E+09
QAH,s
1.35E+08
PE
2.95E+09
QHE2
8.85E+07
Qrej
Qh
7.77E+08
QCT
1.46E+09
7.50E+07
QPVT,s
1.85E+09
As mentioned in the previous sections, the simulation tool can provide results on whatever time
basis. In particular, yearly results are graphically reported from Table 1 to Table 3. Table 1 shows
some important annual energy flows. Here, the total amount of thermal energy produced by the
CPVT (QPVT) is 2.25 109 kJ/year. Such thermal energy is basically produced during the summer
period (QPVT,s :1.85 109 kJ/year) when the thermal performance of the CPVT is significantly better
as a consequence of the higher availability of beam radiation. Note that the thermal performance of
the CPVT is slightly better than the one of the PVT considered in reference [25] which produced a
total thermal energy of 1.84 109 kJ/year (in summer 1.39 109 kJ/year). It could be argued that the
better performance of the CPVT is due to the one-axis tracking system which allows one to pick up
a higher amount of irradiation. In fact, the flat plate PVT (South, 30 deg slope) can convert 6.23
kJ/m2 year of total insolation (3.64 kJ/m2 year of beam radiation), whereas the total radiation
incident on the CPVT is 6.93 kJ/m2 year (4.23 kJ/m2 year of beam radiation). However, it must be
considered that CPVT can convert only the beam radiation (except for the small top absorber area)
whereas the flat plate PVT collectors can convert the total radiation. Therefore, it can be concluded
the insolation available for conversion is lower for the case of the CPVT. Thus, the higher thermal
production by the CPVT is basically due to its high thermal efficiency which is significantly better
than the one of flat plate PVT. This circumstance is well known in literature and it is due to the
313
lower surface per unit thermal power of the CPVT (lower thermal losses) and to the lower average
angle of incidence occurring in the CPVT due to the one-axis tracking system. This better thermal
performance also determines a lower demand of thermal energy by the auxiliary heater AH. In fact,
the thermal energy produced by the AH (Q AH) is 1.78 108 kJ/year (1.35 108 kJ/year in summer,
QAH,s), which is significantly lower than the value calculated in reference [25] (2.31 108 kJ/year). As
a consequence, a better primary energy saving (2.95 109 kJ/year, PE) is achieved with respect to
the value calculated in [25] (2.29 109 kJ/year). Note also that this increase in energy saving is lower
than expected since the additional thermal energy produced by the CPVT is not completely
consumed since a large amount of this heat must be rejected. This is basically due to the phase shift
between solar radiation and space heating/cooling demand. This phase shift also determines a large
amount of heat (QHE1) produced by the CPVT field and employed for DHW. In fact, HE1 produces
DHW only when PVT outlet temperature is higher than the set point. In other words, HE1 produces
DHW when CPVT thermal energy is not demanded by the building and when TK1 storage tank is
full. The system equipped with CPVT produces a larger amount of DHW (Q HE1) with respect to the
flat plate PVT (1.47 109 kJ/year vs 1.11 109 kJ/year). However, this additional DHW is generally
dissipated. In fact, the amount of heat rejected (Q rej ) by the CPVT (7.77 108 kJ/year) is significantly
higher than the one calculated in [25] (4.22 108 kJ/year). Finally, as expected, the calculated values
of space cooling (Qc ) and heating (Qh) demand, Cooling Tower thermal energy (Q CT), ACH cooling
energy (Qc,ACH) and HE2 thermal energy (QHE2) are the same as the ones calculated in reference
[25]
Table 2- Annual results: electrical energy (kJ/year)
Eel,PVT
Eel,aux
Eel,+
Eel,-
Eel,+,F1
Eel,+,F2
Eel,+,F3
Eel,-,F1
Eel,-,F2
Eel,-,F3
8.79E+08 6.98E+07 4.60E+08 3.11E+08 2.28E+08 1.12E+08 1.20E+08 2.76E+08 3.48E+07 6.28E+04
The electricity annual balance is shown in Table 2. Here, it is clearly displayed that the electricity
produced by the PVT (Eel,PVT) is very high (8.79 108 kJ/year), also determining a large amount of
electricity in excess and sold to the grid (Eel,+ ). The overall electrical energy produced by the CPVT
is significantly higher than the one produced by the flat plate PVT (6.04 108 kJ/year) [25], as a
consequence of the high nominal efficiency of the triple-junction PV included in the CP VT
collector. Consequently, the electrical energy in excess (Eel,+ ) produced by the CPVT is larger than
one calculated in reference [25] (4.60 108 kJ/year vs 2.01 108 kJ/year). Similarly, in case of CPVT
collectors, the electrical energy bought from the grid (Eel,-) decreases (3.11 108 kJ/year vs 3.24 108
kJ/year).
Table 3- Annual results: performance indexes
PES
7.70E-01
t,PVT
3.26E-01
el,PVT
1.27E-01
t,PVT
5.29E-01
el,PVT
2.07E-01
Fsol,el
Fsol,s
Fsol,w
FDHW
7.39E-01
9.32E-01
9.01E-01
6.56E-01
The overall thermal and electric energy balances can be summarized by some indexes, shown in
Table 3. The Primary Energy Saving ratio (PES) was significantly higher than the one achieved by
the flat plate system (77.0 % vs 70.2 %). This is mainly due to the higher values of both thermal and
electrical efficiencies of the CPVT compared with the flat plate PVT. In fact, the CP VT thermal
efficiency ( t,PVT), calculated on the basis of the total incident radiation, is 32.6 % whereas the same
value was about 29.4 % in case of flat plate PVT. This is typical of concentrating systems whose
heat transfer area is significantly lower than the one of non-concentrating systems, determining
lower losses toward the environment. This is much clear if the efficiency is calculated with respect
of the sole beam radiation ( t,PVT): in this case the thermal efficiency is slightly lower than 53 %.
Similarly, the use of high performance triple-junction PV layer allows one to achieve ultra-high
electrical efficiency. In fact, the electrical efficiency, calculated with respect to the total radiation
314
( el,PVT), is 12.7 %, whereas the value increases up to 20.7 % ( el,PVT) when the sole beam radiation
is considered. Such values are significantly high, considering the typical operating temperature of
the PV layer. In fact, the electrical efficiency calculated in [25] was 9.7%. As a consequence, the
electrical solar fraction, F sol,el, i.e. the ratio of the overall electrical energy demand produced by the
PVT, is significantly higher than one rated in [25] (73.9 % vs 65%). Similar results are also
achieved for the summer solar fraction, Fsol,s (93.2 % vs 88.2 %). Conversely, the winter solar
fraction, Fsol,w ,of the CPVT is slightly lower than the one calculated in [25] (90.1 % vs 90.8%).
This is also typical of concentrating systems whose thermal performance dramatically decreases
during the winter as a consequence of the lower availability of beam radiation. Finally, Table 3
shows that the about 65.6 % (FDWH) of the heat produced by PVT is used by HE1 for the production
of DHW. This high value is not surprising since the PVT area is significantly high and the TK
volume relatively low. Therefore, there are long periods in which the P VT thermal production is
higher than system space cooling/heating demand and the storage capacity of TK is full. In addition,
PVT thermal energy is always used exclusively for DHW during the Sundays and during the
periods in which both space cooling and heating systems are switched off [4].
Table 4- Annual results: costs (/year)
Cop
CDHW
Cel,-
Cel,+
Cel,+,RS
Crej
Cft,ee
Cft,pe
I0/FA
From the energetic point of view the system equipped with CPVT collectors shows a better
performance than the one based on planar PVT. This was a quite trivial result due to higher
technological level of CP VT. However, this energetic performance must be evaluated
simultaneously with the economic one. In this study, according with the literature review discussed
in the Part I of the paper, a capital cost of 600 per m2 of aperture area of CPVT was assumed. This
cost could be achieved in the next future due to the increasing commercialization of triple junction
PV layers. In spite of this optimistic assumption, the capital cost of the whole system is
dramatically high, due to the high capital cost of CPVT collector. In fact, Table 4 shows that the
annual owning cost ( I0/FA), defined in [25], of the PVSHC system is dramatically higher than the
one calculated for flat PVT (6370 /year vs 47910 /year). On the other hand, CPVT based system
can achieve better values for all the cash flows ( Cop) and eventual feed-in tariffs (Cft,pe and Cft,ee ,
proportional respectively to the primary energy saved and to the electrical energy produced). The
overall result is a lower economic performance with respect to the ones of reference [25]. In fact,
Table 5 shows that all the Simple Pay Back Periods are slightly higher the ones of reference [25]
(respectively 16.8, 11.7, 5.4 and 3.6 years). Therefore, it can be concluded that the cost assumed for
the CPVT is still high to recover the higher energetic performance shown by this kind of
photovoltaic system. Note that in
Table 5 four different Simple Pay Back Periods were considered [25]:
SPB: no public funding
SPBcc: Capital cost contribution of 30% of the extra cost of the PVSHC with respect to the cost
of the RS
SPBft,ee : Feed-in tariff equal to 0.45 per electricity kWh (similar to the one presently adopted
for PV in Italy) produced by the PVT
SPBft,pe: Feed-in tariff equal to 0.207 per primary kWh saved by the PVSHC with respect to the RS
SPBcc
1.30E+01
SPBft,ee
5.21E+00
315
SPBft,PE
3.74E+00
The tool also allows one to display the trends of the state-points temperatures and components
energy flows, during the dynamic simulation. The related plots can be displayed in whatever time
basis multiple of the simulation time step (0.040 hours). For example yearly plots are useful for
detecting the general trends whereas daily or weekly plots can display in detail the typical dynamic
oscillations of temperatures and energy flows. Some of these analyses, regarding both building and
system performances, were presented and discussed in previous works [4, 25-26, 28, 30]. Therefore,
in the following the discussion will be focused only on the differences of the CPVT system
considered in this work with respect to the layouts previously analyzed. Fig. 2 and Fig. 3 display
some key-points temperatures of the system under investigation. Similarly, Fig. 4 and Fig. 5 show
respectively the thermal and electrical energy flows, whereas PVT thermal and electrical
efficiencies are shown in Fig. 6. Such Figures are useful to detect the general trends of the
thermodynamic parameters. However, for a better interpretation of the results it is more convenient
to consider a random winter and summer week to analyse in detail the trends of the thermodynamic
parameters.
35
30
zone
T ext
Temperature ( C)
T CHW,i
25
T CHW,o
20
15
10
5
0
0
1000
2000
3000
4000
5000
Time (h)
6000
7000
8000
8760
8000
8760
T CHW,i
T CHW,o
Temperature (C)
T o,AH
70
T o,SC
T i,SC
T TK,s
60
T TK,w
50
40
0
1000
2000
3000
4000
5000
Time (h)
6000
7000
316
3.5
x 10
QAH
Qh
QSC
Qc
2.5
QD HW
I tot,SC
I b,SC
1.5
1
0.5
0
0
1000
2000
3000
4000
5000
Time (h)
6000
7000
8000
8760
x 10
PPVT
10
Pb ld
Pa ux
PP+
6
4
2
0
0
1000
2000
3000
4000
5000
Time (h)
6000
7000
8000
8760
0.6
t
el
Efficiency
0.5
0.4
0.3
0.2
0.1
0
0
1000
2000
3000
4000
5000
Time (h)
6000
7000
8000
8760
In particular, the corresponding summer plots are shown from Fig. 7 to Fig. 11, whereas the winter
week is shown from Fig. 12 to Fig. 16. Fig. 7 and Fig. 8 show some key-point temperatures for the
considered summer week. Here, it is clearly shown the typical oscillating trend of the internal zone
air temperature due to the operation of the on/off thermostat. The internal temperature is controlled
for the first six days of the week since during the Sunday the building is not used. Fig. 7 also shows
that ACH outlet temperature is often equal to corresponding set-point except for the central hours of
the Tuesday. Conversely, the CHW inlet temperature significantly varies as a consequence of the
variations of cooling demands of the building. Fig. 8 shows the temperatures of the solar loop.
Here, it is clearly displayed that the CPVT outlet temperature is stably over its set point
temperature. PVT outlet temperature is often even higher than the summer set-point (80 C). This
often occurs when the inlet temperature is also high. In fact, the variable speed pump cannot
increase the fuel flow up to its maximum capacity (0.333 kg/s per CPVT module), determining the
increase of CPVT outlet temperature shown in Fig. 7. Note that Fig. 7 also shows that TK average
temperature is often close to 80 C: only during the first two days of the week TK temperature
significantly decreases. This trend can be better interpreted analyzing Fig. 9. Here, it is clearly
shown that the thermal energy produced by the PVT is significantly low during the first two days as
a consequence of the reduction of solar beam radiation. As a consequence, during these days the
AH must provide additional thermal energy to achieve the temperature of the hot fluid required to
drive the ACH. Obviously, in those days there is no DHW production since there is no heat in
excess from solar loop. Conversely, during the other days of the week the PVT thermal energy is
significantly higher than the one demanded by the ACH and a lot of thermal energy in excess is
converted in DHW by HE1. This trend could be mitigated using larger TK storage systems,
balancing positive and negative energy flows in the solar loop. However, larger systems would also
determine larger thermal losses and a longer response of the system to the user demand. Therefore,
a possible solution to this problem could consist in a variable volume storage system which is an
option that the authors are going to explore in forthcoming studies. Fig. 10 shows the electricity
flows for the same summer week. Here, it is clearly displayed that the electrical power produced by
the PVT is generally higher than the overall electrical energy demand (CPVT auxiliary devices,
building equipments and lights). Therefore, the electrical energy in excess is typically a large
amount of the PVT production, whereas the integration from the grid is required only in case of low
beam radiation.
35
T zone
T ex t
30
Temperature (C)
T CH W,i
T CH W,o
25
20
15
10
5
0
5880 5892 5904 5916 5928 5940 5952 5964 5976 5988 6000 6012 6024 6036 6048
Time (h)
318
85
80
Temperature (C)
75
70
T CHW,i
65
T CHW,o
60
T o,AH
T o,SC
55
T i,SC
T TK,s
50
T TK,w
45
5880 5892 5904 5916 5928 5940 5952 5964 5976 5988 6000 6012 6024 6036 6048
Time (h)
3.5
x 10
QAH
Qh
3
Thermal Energy (kJ/h)
QSC
Qc
2.5
QDHW
Itot,SC
Ib,SC
1.5
1
0.5
0
5880 5892 5904 5916 5928 5940 5952 5964 5976 5988 6000 6012 6024 6036 6048
Time (h)
x 10
PPVT
Pbl d
Paux
P-
P+
3
2
1
0
5880 5892 5904 5916 5928 5940 5952 5964 5976 5988 6000 6012 6024 6036 6048
Time (h)
319
Finally, Fig. 11 shows that the electrical efficiency of the PVT is stably higher than 20 % and it
slightly decreases during the central hours of the day due to a corresponding increase of PV layer
temperature. The thermal performance of the CP VT is also very good being typically higher than 50
% for highly insolated days. Note that both the efficiencies shown in figure are calculated with
respect to the incident beam solar radiation.
0.6
t
el
Efficiency
0.5
0.4
0.3
0.2
0.1
0
5880 5892 5904 5916 5928 5940 5952 5964 5976 5988 6000 6012 6024 6036 6048
Time (h)
The winter operation of the system is dramatically different from the summer one. For the
considered winter week, temperatures are shown in Fig. 12 and Fig. 13. Here, it shown that the
temperature of the internal zone oscillates around 20 C (winter set point) only during the first
hours of the day. Then the temperature overcomes the set-point only driven by the internal loads
(equipment and people) of the building. Therefore, the thermal energy is required only during the
first hours of the day when unfortunately beam radiation is scarce and external temperature is low.
In fact, for the selected week P VT outlet temperature is significantly lower of the corresponding set
point in the 3rd and 4th day when the beam radiation is very low (Fig. 14).
25
Temperature (C)
20
15
10
T z one
T e xt
T C HW,i
T C HW,o
0
8064 8076 8088 8100 8112 8124 8136 8148 8160 8172 8184 8196 8208 8220 8232
Time (h)
320
60
TC HW,i
TC HW,o
To ,AH
Temperature (C)
55
To ,SC
Ti ,SC
TTK,s
TTK,w
50
45
40
8064 8076 8088 8100 8112 8124 8136 8148 8160 8172 8184 8196 8208 8220 8232
Time (h)
Fig. 14 also shows that the thermal energy produced by the solar field is basically used for DHW
since for the majority of the day there is no space heating demand. The winter plot of the electrical
energy flows (Fig. 15) is completely different from the summer one. In fact, the electrical energy
produced by the PVT is very often lower than building and auxiliaries demand, determining a larger
amount of electrical energy bought from the grid. This is also determined by the lower PVT
electrical efficiency shown in Fig. 16 which is stably lower than 20 %. Such lower values compared to the summer ones- are basically determined by the lower availability of solar radiation
which also overcomes the positive effect of the lower operating temperature of the winter operation.
This lower availability of beam radiation also affects the thermal performance of the CPVT whose
thermal efficiency is significantly lower than 50 %.
x 10
QAH
Qh
QSC
1.5
Qc
QDHW
Itot,SC
Ib,SC
0.5
0
8064 8076 8088 8100 8112 8124 8136 8148 8160 8172 8184 8196 8208 8220 8232
Time (h)
321
x 10
PPVT
6
Elect rical Energy (kJ/h)
Pb ld
Pa ux
PP+
4
3
2
1
0
8064 8076 8088 8100 8112 8124 8136 8148 8160 8172 8184 8196 8208 8220 8232
Time (h)
t
el
Efficiency
0.5
0.4
0.3
0.2
0.1
0
8064 8076 8088 8100 8112 8124 8136 8148 8160 8172 8184 8196 8208 8220 8232
Time (h)
322
3.5
x 10
QHE1
QAH
2.5
QPVT
PE
Qrej
2
1.5
1
0.5
0
-0.5
0
200
400
600
800
1000
1200
A PV T ( m )
12
x 10
E el,PVT
10
E el,a ux
E el,+
E el,-
8
6
4
2
0
0
200
400
600
2
A PV T ( m )
800
1000
323
1200
1.2
P ES
,P
t VT
el,PVT
F sol,el
PES, , F
sol
0.8
F sol,s
F sol,w
0.6
0.4
0.2
0
-0.2
0
200
400
600
800
1000
1200
A PV T ( m )
Fig. 19 - Sensitivity analysis: PES, efficiencies and Solar Fractions vs PVT area
As expected, Fig. 18 shows that PVT electricity production (Eel,PVT) linearly increases with P VT
area. This increase also determines a larger amount of electricity sold to the grid (Eel,+ ) whereas the
one bought from the grid (Eel,-) obviously decreases. Fig. 19 shows that PES and all the solar
fractions increase less then proportionally with PVT area. In fact, the higher the PVT area is, the
higher the PES. In fact, for larger solar fields, the amount of thermal and electrical energy produced
by the auxiliary devices (heater or grid) reduces, also increasing the PES. Conversely, as typically
occurs, PVT thermal efficiency decreases for higher PVT area whereas the electrical efficiency is
only very slightly affected by the variation of the PVT area.
5
2.5
x 10
C op
C DHW
C el,-
Cost (/year)
C el,+
1.5
C el,+,RS
C DHW,rej
C ft,ee
C ft,pe
0.5
0
-0.5
0
200
400
600
800
1000
1200
A PV T ( m )
electricity production, linearly increases with PVT area (the electricity production is a linear
function of the PVT area). Conversely, the feed-in tariff proportional to the PE savings, shows the
same shape of the PE curve. Both feed-in tariffs are dramatically higher than all the remaining
cash flows for whatever PVT area is considered. Note also that cash flow due to the selling of the
exceeding electricity (Cel,+ ) is significantly lower than the corresponding purchase cost by the RS
(C+,RS). This is due to the differences between electricity costs and prices discussed in the previous
section.
The economic performance of the system under investigation is calculated using Simple Pay Back,
SPB (Fig. 21), Net Present Value (Fig. 22) and Profit index (Fig. 23). The minimum SPB is
achieved at 960 m2 except for the case of SPBft,ee whose minimum occurs at 1200 m2. However, the
system is never profitable without incentive since PI and NPV are always negative for whatever
PVT area. Conversely, NPV increases with PVT area considering both feed-in tariffs. The
corresponding Profit Indexes are maximum respectively at 960 m2 (PIftpe ) and 1200 m2 (PIftee ).
30
SPB
SPBcc
25
SPBft,ee
SPBft,pe
SPB (year)
20
15
10
5
0
200
300
400
500
600
700
2
A PV T ( m )
800
900
1000
1100
1200
Prof it Index
1.5
1
PI
P Ift,ee
0.5
P Ift,pe
0
-0.5
-1
-1.5
-2
0
200
400
600
2
A PV T ( m )
800
325
1000
1200
2.5
x 10
NPV
NPVft,ee
NPVft,pe
1.5
1
0.5
0
-0.5
0
200
400
600
800
1000
1200
A PV T ( m )
P Ift,pe,C PVT=200 /m
P Ift,pe,C PVT=400 /m
P Ift,pe,C PVT=600 /m
2
2
P Ift,pe,C PVT=800 /m
Prof it Index
P Ift,pe,C PVT=1000 /m
P Ift,pe,C PVT=1200 /m
P I,C PVT=200 /m
P I,C PVT=400 /m
P I,C PVT=600 /m
P I,C PVT=800 /m
-1
P I,C PVT=1000 /m
-2
-3
0
P I,C PVT=1200 /m
200
400
600
2
A PV T ( m )
800
1000
1200
Fig. 24 - Sensitivity analysis: PIft,pe vs PVT area and PVT capital cost
326
2
2
x 10
NP Vft,pe,C PVT=200 /m
2.5
NP Vft,pe,C PVT=400 /m
NP Vft,pe,C PVT=600 /m
2
2
NP Vft,pe,C PVT=800 /m
1.5
NP Vft,pe,C PVT=1000 /m
NP Vft,pe,C PVT=1200 /m
NP V,C PVT=200 /m
0.5
NP V,C PVT=400 /m
NP V,C PVT=600 /m
NP V,C PVT=800 /m
-0.5
NP V,C PVT=1000 /m
-1
NP V,C PVT=1200 /m
-1.5
0
200
400
600
800
1000
2
2
1200
A PVT (m )
Fig. 25 - Sensitivity analysis: NPVft,pe vs PVT area and PVT capital cost
QA H
Q PV T
PE
Q rej
x 10
3
2
1
0
20
40
60
80
100
120
140
Q HE1
10
x 10
Eel,PV T
Eel,aux
5
Eel,+
Eel,-
20
40
Cop
CDHW,rej
x 10
2
1
20
40
60
80
vT K,s (L/m 2)
100
120
F sol, el
F sol, s
140
20
10
0
0
20
40
60
80
vTK,s (L/m2)
100
120
140
F sol, w
3
P rofit Index
SPB (year)
PES
,P
t VT
el,PV T
30
S PBf t ,pe
140
0
0
Cf t ,pe
S PBf t ,ee
120
0.5
Cf t ,ee
S PB
S PBcc
100
1
PE S, , Fs ol
Cel,+, RS
Cost (/year)
Cel,+
80
CDHW
Cel,-
60
vTK,s (L/m2)
v TK, s (L/m )
20
40
60
80
100
120
140
PI
PI f t,ee
2
1
PI f t,pe
0
-1
20
40
60
80
100
120
140
vTK,s (L/m2)
v TK, s (L/m )
preferable also from the energetic point of view. In fact, the time shift between solar radiation and
energy demand is lower and the het losses toward the environment are higher. Therefore, low TK
winter volumes should be considered according both energetic and economic criteria.
QA H
Q PVT
PE
Q rej
2
1
0
20
40
60
80
100
120
140
Q HE1
x 10
10
x 10
Eel,PV T
Eel,aux
5
Eel,+
Eel,-
20
40
C el, +
C el, +,RS
C DHW,rej
Cost (/year)
C el, -
x 10
2
1
20
40
60
80
v TK, w (L/m2)
100
120
F sol, el
F sol, s
F sol, w
140
20
40
60
80
v TK, w (L/m2)
100
120
140
20
10
0
PES
,P
t VT
el,PV T
P rofit Index
SPB (year)
SPB ft ,pe
140
0.5
30
SPB ft ,ee
120
C f t,pe
SPB
SPB cc
100
C f t,ee
80
PE S, , Fs ol
Cop
C DHW
60
v TK, w (L/m2)
vTK,w (L/m )
20
40
60
80
100
120
140
PI
PI f t,ee
PI f t,pe
0
-1
20
40
60
80
100
120
140
v TK, w (L/m2)
v TK, w (L/m )
4. Conclusions
The results of the case study presented prove the technical feasibility of the PVSHC polygeneration
system analyzed in the paper, based on photovoltaic/thermal solar collectors. The dynamic
simulation showed that the system is capable to produce electricity, space heating, space cooling
and domestic hot water all year long. The results of the dynamic simulations showed the significant
potential of energy savings of CPVT, specially when they are coupled with an absorption chiller for
the production of cooling energy. The combination of CPVT and SHC allows one to maximize the
utilization of the thermal energy produced by the CPVT, particularly during the summer. On the
other hand, results showed that CPVT thermal and electrical productions are very fluctuating in
function of external radiation and temperature. Therefore an auxiliary system is always mandatory
for a safe operation of the system. In addition, such fluctuations also determine a large amount of
peak thermal energy which cannot be stored in storage tanks. Therefore, in order to maximize
energy utilization, the simultaneous demand of DHW is crucial, since this DHW can be produced
using the peak thermal energy of CPVT, exceeding building space heating and cooling demand.
Finally, results also showed that CPVT are particularly sensitive to the ambient temperature: during
the cold winter days their outlet temperature is often significantly lower than the corresponding setpoint.
The economic analysis showed a profitability of the PVSHC considered of the same order of
magnitude than other SHC systems. In particular, the system is not profitable without public
funding policies and becomes extremely convenient in case of feed-in tariffs comparable to the ones
presently adopted for PV collectors in UE.
328
The overall energetic performance of the CPVT system under investigation is better than the one
based on flat plate CPVT previously investigated by the authors. Conversely, the economic indexes
of the system based on concentrating P VT are slightly worse than the ones of flat plate PVT. This is
mainly due to the high capital cost of triple-junction PV cells which is not adequately balanced by
the increase in its overall electrical efficiency
Future developments of this work will focus on the possibility to increase CPVT operating
temperature in order to drive a double effect absorption chiller. The aim of this improvement is to
improve the summer performance of the system, reducing the PVT area required to drive the
absorption chiller.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
Calise F., Dentice d'Accadia M ., Vanoli L., Thermoeconomic optimization of solar heating
and cooling systems. Energy Conversion an M anagement, 2010. In Press.
Calise F., Dentice d'Accadia M ., Vanoli R., Dynamic simulation and parametric
optimization of a solar-assisted heating and cooling system. International Journal of
Ambient Energy, 2010. 31(4): p. 173-194.
Calise, F., M . Dentice dAccadia, and A. Palombo, Transient analysis and energy
optimization of solar heating and cooling systems in various configurations. Solar Energy,
2010. 84(3): p. 432-449.
Calise, F., Thermoeconomic analysis and optimization of high efficiency solar heating and
cooling systems for different Italian school buildings and climates. Energy and Buildings,
2010. 42: p. 992-1003.
Calise, F., A. Palombo, and L. Vanoli, Maximization of primary energy savings of solar
heating and cooling systems by transient simulations and computer design of experiments.
Applied Energy, 2010. 87: p. 524-540.
Calise, F., M . Dentice d' Accadia, A. Palombo, and L. Vanoli. Simulation model and
analysis of a small solar-assisted refrigeration system: dynamic simulation and
optimization. in Proceedings of the ASME International Mechanical Engineering Congress
& Exposition 2008 IMECE08. 2008. Boston, USA.
Ardehali, M .M ., Shaharestani, M ., Adams C.C., Energy simulation of solar assisted
absorption system and examination of clearness index effects on auxiliary heating. Energy
Conversion and M anagement, 2007. 48: p. 864-870.
Assilzadeh, F., S.A. Kalogirou, Y. Ali, and K. Sopian, Simulation and optimization of a
LiBr solar absorption cooling system with evacuated tube collectors. Renewable Energy,
2005. 30: p. 1143-1159.
Folrides, G.A., Kalogirou, S.A., Tassou, S.A., Wrobel, L.C., Modelling and simulation of an
absorption solar cooling system for Cyprus. Solar Energy, 2001. 72(1): p. 43-51.
Folrides, G.A., Kalogirou, S.A., Tassou, S.A., Wrobel, L.C., Modelling, simulation and
warming impact assessment of a domestic-size absorption solar cooling system. Applied
Thermal Engineering, 2002. 22: p. 1313-1325.
Garcia-Casals, X., Solar absorption cooling in Spain: Perspectives and outcomes from the
simulation of recent installations. Renewable Energy, 2006. 31: p. 1371-1389.
Ghaddar, N.K., Shihab, M ., Bdeir, F., Modeling and simulation of solar absorption system
performance in Beirut. Renewable Energy, 1996. 10(4): p. 539-558.
A.Joudi, Khalid and J. Abdul-Ghafour Qussai, Development of design charts for solar
cooling systems. Part I: computer simulation for a solar cooling system and development of
solar cooling charts. Energy Conversion and M anagement, 2003. 44: p. 313-339.
A.Joudi, Khalid and Qussai J. Abdul-Ghafour, Development of design charts for solar
cooling systems. Part II: Application of the cooling f-chart. Energy Conversion and
M anagement, 2003. 44: p. 341-355.
329
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
M ateus, T. and A.C. Oliveira, Energy and economic analysis of an integrated solar
absorption cooling and heating system in different building types and climates. Applied
Energy, 2009. 86: p. 949-957.
U. Desideri, S. Proietti, Analysis of energy consumption in the high schools of a province in
central Italy Energy and Buildings, 2002. 34: p. 1003-1016.
A. El Fadar, A. M imet and M . Prez-Garca, Modelling and performance study of a
continuous adsorption refrigeration system driven by parabolic trough solar collector SOlar
Energy, 2009. 83(6): p. 850-861.
Ahmet Lokurlu, Fritz Richarts, Dirk Kruger, High efficient utilisation of solar energy with
newly developed parabolic trough collectors (SOLITEM PTC) for chilling and steam
production in a hotel at the Mediterranean coast of Turkey. International Journal of Energy
Technology and Policy, 2005. 3(1-2): p. 137-146.
Calise F., Dentice d'Accadia M ., Palombo, A., Vanoli L., Dynamic Simulation Of High
Temperature Solar Heating And Cooling Systems, in Eurosun 2010. 2010: Graz (AU).
M . M azloumi, M . Naghashzadegan, and K. Javaherdeh, Simulation of solar lithium
bromidewater absorption cooling system with parabolic trough collector Energy
Conversion an M anagement, 2008. 49(10): p. 2820-2832.
M ing Qu, Hongxi Yin, David H. Archer, A solar thermal cooling and heating system for a
building: Experimental and model based performance analysis. Solar Energy, 2009. 84(2).
Tierney, M .J., Options for solar-assisted refrigerationTrough collectors and double-effect
chillers. Renewable Energy, 2007. 32(2): p. 183-199.
Vokas, G., N. Christandonis, and F. Skittides, Hybrid photovoltaicthermal systems for
domestic heating and coolingA theoretical approach Solar Energy, 2006. 80(5): p. 607.
M ittelman, G., A. Kribus, and A. Dayan, Solar cooling with concentrating
photovoltaic/thermal (CPVT) systems Energy Conversion and M anagement, 2007. 48: p.
2481-2490.
Calise, Francesco, Adolfo Palombo, and Laura Vanoli, Design and dynamic simulation of a
novel solar trigeneration system based on photovoltaic/thermal collectors, in ECOS 2011.
2011: Novi Sad, Serbia.
Calise, F., Design Of A Hybrid Polygeneration System With Solar Collectors And A Solid
Oxide Fuel Cell: Dynamic Simulation And Economic Assessment. International Journal of
Hydrogen Energy, 2011. 36(10): p. 6128-6150
Klein, S.A. et al., Solar Energy Laboratory, , TRNSYS. A transient system simulation
program. 2006: University of Wisconsin, M adison.
Calise, F., High Temperature Solar Heating and Cooling Systems for Different
Mediterranean Climates: dynamic simulation and economic assessment Applied Thermal
Engineering, 2011. 32: p. 108-124.
F. Calise, Ferruzzi G., Vanoli L., Transient Simulation of Polygeneration Systems Based on
Fuel Cells and Solar Cooling Technologies in ECOS 2010. 2010: Lousanne, CH.
Calise, F., G. Ferruzzi, and L. Vanoli, Transient simulation of polygeneration systems based
on PEM fuel cells and solar heating and cooling technologies. Energy, 2011. In Press,
Corrected Proof,Jun 2011
330
Abstract:
Energy and environmental performances of a small-scale cogeneration system composed of downdraft
wood gasifier and reciprocating engine is theoretically examined in the paper. The gasifier is modelled using
a non-stoichiometric thermodynamic quasi equilibrium approach. Achievable power and efficiency of the
gas engine is estimated using a method based on analysis of the engine thermodynamic fuel-air theoretical
cycle. It is assumed that the relative changes of operating parameters of a selected machine are
proportional to the relative changes of the cycle parameters after the change of fuel. It is taken into account
that the gasification agent can be composed of ambient air and recirculated exhaust gas from the engine.
The proposed configuration of the system is examined in the aspect of electricity generation efficiency, total
fuel energy utilization, potential for primary energy savings and global CO2 emission reduction. It was found
that the proposed modification of a typical downdraft reactor gas engine system, that runs at a given
reactor temperature, reduces the heating value of the producer gas. Consequently the electric power output
of the engine module is derated. However, the recirculation of exhaust gas from the engine improves the
cold gas efficiency and the performance of the whole system.
Keywords:
Renewable Energy, Biomass, Gasification, Gas Engines, Energy Conversion, Cogeneration.
1. Introduction
Reductions of CO 2 emission and consumption of fossil fuels are nowadays the main driving forces
for using biomass in the energy sector. It is usually assumed that the traditional, fossil fuel based
generation technologies are replaced if renewable energy sources are introduced. The electricity
from renewable sources as well as from cogeneration is nowadays supported by different legal and
financial means. Consequently, it can be assumed that with a high probability the power replaced
within the national system will be the one from the coal fired plants. Therefore the potential for fuel
energy savings and emission reduction can be evaluated with respect to coal.
In order to demonstrate emission and primary energy saving potential of the biomass based
cogeneration two indices can be used. The first is the global reduction of CO 2 emission (under
assumption that the firing of biomass is CO 2 emission neutral). The second one is the Energy
Replacement Index (ERI), that shows the amount of non-renewable energy saved within the
regional energy system by using biomass energy in a cogeneration plant (in GJ of non-renewable
energy saved per GJ of biomass energy input).
mcoalLHVcoalWEcoal EelWEref EUF el
WEref
mCO2
WEcoal el
,
(1)
mbio LHVbio
3.6
b
mcoalLHVcoal
ERI
3.6 Eel
ref
mbio LHVbio
EUF
el
b
el
ref
331
(2)
Where the characteristics of an energy conversion plant is given by net electricity production Eel,
generation efficiency el, fuel energy utilization factor EUF and cogeneration index :
Eel
Eel, gen 1
Eel , fss ,
(3)
3.6Eel
,
mbio LHVbio
el
EUF
(4)
3.6 Eel Q
,
mbio LHVbio
3.6 Eel
Q
(5)
el
EUF
(6)
el
It can be concluded from Equations (1) and (2) that a good performance of a conversion technology
depends not only on the power generation efficiency (usually lower than the reference system
efficiency) but also on the cogeneration mode. In the Fig. 1 there are presented results obtained
under the following assumptions: reference efficiency of power generation ref,system = 0.36 [1],
reference efficiency of a heating boiler b = 0.86 [2], CO2 emission from the system power plant
WEref = 936.36 kg/M Wh and from a coal fired heating boiler WEcoal = 94.85 kg/GJ [1]. It can be
noticed that starting from relatively low values of performance parameters of biomass fired
cogeneration plant ( el at the level of 0.26 and EUF at the level of 0.5) savings of non-renewable
energy (loco plant) are bigger than the amount of biomass energy used (ERI > 1.0). Additionally
there are expected lower values of cumulative energy consumption and cumulative emission of CO 2
in the fuel production and delivery processes.
Fig. 1. ERI and global CO2 emission reduction resulted from biomass fired cogeneration
Gasification based cogeneration is an important technological alternative for biomass fired
distributed generation plants of small and medium scale electric outputs. Reciprocating engines
integrated with gasification reactors and gas cleaning devices can be widely applied in this field. On
the other hand there are no engines manufactured for producer gas utilization. Natural gas or diesel
engines are usually used. Despite this fact there is little information available in the published
literature about the effects of fuelling an engine with the producer gas. Due to the complexity of incylinder processes there are not many analytical tools available to estimate the parameters of a
machine running in the producer gas mode.
In this paper there is presented the study of a downdraft gasifier - gas engine cogeneration system
integrated with biomass gasification. The gasification process is analyzed using a nonstoichiometric approach with minimization of the Gibbs free energy. Such approach is reported in
the literature as a satisfactory for general plant studies (Baggio et al.[3], Schuster et al.[4], Ruggiero
332
et al.[5]). In this work the quasi-equilibrium approach is used. The performance of a piston engine
after the change of fuel there is estimated using the model based on the spark-ignition engine fuelair theoretical thermodynamic cycle.
2. Plant configuration
The cogeneration plant consists of typical fixed bed downdraft gasifier, gas cleaning equipment and
gas engine. The scheme of the plant is shown in Fig. 2.
Biomass is converted into a combustible gas in the reactor using a mixture of air and engine exhaust
gas as the gasifying agent. The temperature of the agent can be elevated using waste heat from the
raw producer gas. The exhaust gas from the engine is recirculated into the gasifier as well as it is
used for feedstock drying and network water heating. Scheme of the drying system is given in
Fig. 3. The low temperature heat from engine cooling cycle is used for water heating.
The prime mover selected for the analysis is the GE J320 GS gas lean fuel-air mixture burn engine.
The operation of this type of engine fuelled with producer gas has been already demonstrated by
Papagiannakis et al. [33]. According to Herdin et al. [6] many problems have been so far
encountered in this type of plants. The biggest one is currently a low reliability. The authors claim
that a great research and development work must be done to improve the performance. There are
also required commercial applications to demonstrate the economic effectiveness. Nevertheless the
technology is available and it can be applied in many locations [7].
The type of feedstock is spruce in the form of chips. Characteristics of the wood is given in Table 1.
Table 1. Characteristics of the feedstock (spruce chips)
Proximate analysis (%wt, dry basis) Ultimate analysis (%wt, dry basis)
Fixed carbon
28.3%
Carbon C
51.2%
Volatile mater
70.2%
Hydrogen H
6.1%
Ash
1.5%
Oxygen O
40.9%
HHV, MJ/kg
20.1
Nitrogen N
0.3%
3. Biomass drying
It is assumed that the initial biomass water content is 40% (wet basis). The feedstock is dried to
10% of water content using a rotary drum drier. The drying medium is a mixture of the engine
exhaust gas after the heat recovery boiler and ambient air. If it is necessary a small amount of
producer gas can be burned in order to provide the heat required for drying. The temperature of the
drying medium at the inlet of the drier is 140C. According to Bolhr-Nordenkampf et al. [8] using
a low temperature heat source gives the significant advantage of a low exergetic input into the dryer
as well as low organic emissions.
The total mass balance of the drying process takes the form:
mwood, db 1 wdb mdm 1 X dm, in mwood,db 1 wdb,max mdm 1 X dm,out ,
(7)
The balance of moisture is as follows:
mbio, dbwdb mdmX dm,in mbio,db wdb,max mdm X dm,out ,
(8)
H bio,in
mdm
i 1
g i hi Texh
H bio,out
mdm
To u t
gi hi T
i 1
X dm,in hH 2 O
Td m , z H 2O
X dm,out hH 2O
in
pdm
To u t , z H 2 O
out
,
pdm
333
Qout
(9)
(10)
The dry drying medium is assumed to be composed of N 2, O2, CO2 and Ar. The heat capacity of wet
wood is calculated using correlation given by Regland et al. [9]:
cdry 4.19wdb
cwet
0.02355Twood 1.32wdb 6.191 wdb ,
(11)
1 wdb
cdry 0.1031 0.003867Twood ,
(12)
Fig. 2. Scheme of the proposed downdraft gasifier-gas engine cogeneration plant (G - gasifier;
C - cyclone; GC1 - raw gas cooler; VS - Venturi scrubber; WS1, WS2 - water separators;
GF - producer gas fan, FL - flare; F1, F2, F3 - filters; GC2 - final cooler; BT - buffer tank;
CR - carburetor; TC - turbocharger; MC - mixture cooler; JC - jacket water cooler; EC - exhaust
gas cooler; G - electric power generator; P - pump)
Fig.3. Schematic diagram of drier system (FGF flue gas fan; RDD rotary drum drier)
It is assumed that the temperature of wood at the outlet of the drier is equal to the dew point
temperature for the given partial pressure of water. Heat loss from the drier is assumed to be 3% of
the total inlet enthalpy. Relative humidity of the drier outlet gas is controlled to be lower than
100%. From this condition the demand for an additional heating using a portion of producer gas is
calculated. Emission of organic compounds and corresponding loss of carbon are, basing on [8],
neglected due to a relatively big size of wood chips and low temperature of the heating medium.
334
As presented by Gerun et al. [23] the tar is modeled as a mixture of benzene C6H6 and
naphthalene C10H8. The ratio of benzene to naphthalene is 0.35/0.65 [23].
M ass flow of carbon passing into the reduction zone is equal the flow of fixed carbon indicated
by the proximate analysis.
Gaseous products of the process are: CO, CO 2, H2, CH4, H2O, N 2, SO2.
M ethane from the pyrolysis is partially converted in oxidation zone. Remaining CH 4 stays
inactive in reduction reactions and passes to products of the process.
Pressure losses in the gasifier bed can be neglected [19].
The assumed global stoichiometric reaction of the process is:
n'C CH a Ob N c Sd n' H 2 O H 2O (l ) nH 2 O H 2 O (g )
nO 2 , min O2 nCO2 CO2 nN 2 N 2
,
(13)
n"CO CO n"CO2 CO2 n"H2 H 2 n"CH4 CH 4 n"H 2 O H 2 O (g ) n"N 2 N2 n"SO2 SO2 n"tar Cx H y
The process involves also ash and argon from the atmospheric air, that are regarded as inert
substances. The total mass composition of biomass is:
C H O S N w ash 1 ,
(14)
The stoichiometric formula of biomass in (13) can is determined from the ultimate analysis:
2n' H 2
2 n'O2
2n' N2
2n'S
a
,b
, c
, d
,
n'C
n'C
n'C
n'C
(15)
The composition of gas phase after oxidation and reduction respectively is determined assuming the
state of thermodynamic equilibrium in which the Gibbs free energy reaches its minimum. The
objective function takes the following form:
ls
lf
n ji h0ji T
Ts0ji T , p
RT ln z ji
min ,
(16)
i 1 j 1
(17)
(18)
b3: n'O2
nO2 ,min
nCO2
n"CO2 n"SO2
1
n"
n"
2 CO H 2 O
0,
(19)
In order to estimate the amount of methane from pyrolysis the assumptions made by Ratnadhariya
et al. [10] have been initially adopted. It was however found that if half of the hydrogen, that is not
associated with oxygen, is released as CH 4 and then it is not involved in the calculation of
equilibrium, the final content of methane in producer gas is to high in comparison with
experimental results. It was found that the additional assumption that 50% of CH 4 from pyrolysis is
involved in calculation of equilibrium in oxidation zone leads to a reasonable level of accuracy.
The total number of moles of the raw producer gas at the output of the gasifier is:
n"g n"H 2 n"CO n"CH4 n"CO2 n"H 2 O n"N2 n"SO2 n"Ar n"tar ,
(20)
Experimental studies show that a small amount of char leaves the gasifier unconverted. Therefore
the carbon conversion efficiency is defined:
n'C n"C
,
(21)
C
n'C
336
Where nC contains two sources of carbon. The first one is the assumed value of nC _ 1 that is not
available for the calculation of equilibrium. Remaining carbon nC _ 2 results from the equilbrium:
n"C n"C _ 1 n"C _ 2 ,
(22)
Construction of the solid particle removal system of downdraft reactors limits the maximum value
of C. In this study it is assumed that the achievable carbon conversion efficiency is 0.95.
M inimization of the objective function (16) can be performed by the method of Lagrange
multipliers [24]. The modified objective function to be minimized takes the form:
ls _ g
gi0 T , p
RT
n"i
i 1
ln n"i ln n"g
n"C _ 2
gC0 T , p
RT
min ,
k k
(23)
k 1
M inimization procedure requires a solution of eight equations with eight unknowns ( nH 2 , nCO ,
nCH 4 , nCO2 , nH 2O , 1, 2, 3). Five equations result from the condition of zeroing the partial
derivatives of function F with respect to unknowns:
F
0,
(24)
n"i
The final form of Equations (24) for the gas components takes the form:
n"i n"g exp
gi0 T , p
RT
3 ls _ g
k
akj ,
(25)
k 1 j 1
For solid carbon present in reduction zone Equation (24) takes the form:
gC0 T, p
0,
(26)
The remaining three equations are substance balances (17), (18) and (19).
The procedure is solved twice - for first and second respectively. This approach allows control of
the temperature after the oxidation process. As the equilibrium calculations are justified for high
temperatures, the temperature at the outlet of the first zone is controlled to be equal 1273 K.
The process temperature T is determined from the energy balance of the individual zones. For 1 kg
of wet biomass at input temperature 298 K the balance of the first zone takes the form:
Ti
LHVdb (1 w) wr
i
n" j LHV j
hj
ni hi 298K
TI
m fc LHVC
298K
mcharcchar TI
Tg
(27)
Qout
Heat capacity of char passing from first to second zone is modeled using formula [25]:
cp
7.32 *10 7 T 2 ,
1430 0.355T
kJ
,
kgK
(28)
n" j, II LHV j
j, II
hj
TI
298K
hj
TII
298K
m fc LHVC
n'C 1
mcharcchar TI
C
M C LHVC
Tg
TII
hC 298K
,
m ashcashTII
Qout
Heat capacity of ash is modeled using Kirovs correlation [26], that for ash takes the form:
337
(29)
kJ
,
kgK
(30)
Finally the cold gas efficiency of the gasification process is defined as follows:
Vg LHVg
,
cold _ gas
mdb LHVdb
(31)
For solving the reactor model an in-house code was written in FORTRAN. The problem of Gibbs
function minimization is of the type that in many cases is difficult to converge. There are however
procedures available to solve the task. In this study there are applied the procedures of the program
CEA (Chemical Equilibrium with Applications) [24]. Thermodynamic properties of substances are
determined using the JANAF tables.
The model has been verified against experimental data presented in the published literature [20, 27].
All the conditions of the process applied for the verification of the model are the same as given in
the cited papers. A representative example of results of the validation is given in Table 2. It was
found that the model generates typical results reported for equilibrium based approach. It can be
noticed in Table 2 that in comparison with the experiment there are overestimated carbon
conversion efficiency, product gas flow, yields of hydrogen and carbon dioxide. The yields of
carbon oxide and nitrogen are significantly underestimated while the yield of methane is in
relatively good agreement with the measured data. It was finally concluded that the model generates
the results that are over optimistic. Therefore the modified approach has been tested. It has been
assumed in the modified model that a certain amount of H2O present at the reduction zone is not
available for the calculation of the equilibrium. It was found that if it is at the level of 25 - 35% the
model gives better results of carbon conversion efficiency and product gas flow. Also composition
of the gas is corrected.
Table 2. Verification of the model against experimental data
Experiment Experiment M odel
(run 8)
(run 9)
(run 8)
[20]
[20]
Input data
0.339
0.323
0.339
Air excess ratio
Total heat loss, % of
n.a.
n.a.
8.8
chemical energy input
Tar yield, g/kgwb
6
5
5
Results:
Throat temperature, C
1130
1206
1129
3
Product gas flow, Nm /h
9.18
9.48
9.86
Product gas composition, %vol
N2
53.95
54.24
44.20
H2
11.86
13.83
18.94
CO
19.89
18.41
12.76
CH4
2.47
1.98
2.2
CO2
11.25
11.12
17.04
Ar
n.a.
n.a
0.56
C2H2
0.26
0.19
0
C2H6
0.12
0.02
0
89.1
88.9
93.39
Carbon conversion C
338
M odel
(run 9)
M odel
modified
(run 8)
M odel
modified
(run 9)
0.323
2.7
0.339
8.8
0.323
2.7
1208
10.37
1129
9.18
1208
9.48
41.98
20.16
15.38
2.0
15.59
0.54
0
0
96.61
47.44
16.15
15.22
1.55
14.68
0.61
0
0
85.72
45.97
16.41
18.96
1.35
12.35
0.59
0
0
87.52
120 00
100 00
p, kPa
80 00
60 00
40 00
20 00
0
7
6
-20 00
0
0,1
0,2
Dis0,3
pla ceme nt0,4volu me 0,5
0,6
V, m
0,7
0,8
According to Heywood [34] such cycle can be composed of the following processes: reversible
adiabatic compression of mixture of air, fuel and cylinder residual gas; combustion at constant
volume without heat loss, to burned gas in chemical equilibrium; reversible adiabatic expansion of
the burned gas; ideal adiabatic exhaust gas blowdown and ideal intake with adiabatic mixing
between residual gas and fresh mixture. This approach takes into account:
variations of density of the mixture and volumetric efficiency that result from mixing with a
residual amount of gas in the cylinder,
variations of maximum temperature and pressure of the cycle,
different properties of working fluid in compression and expansion processes due to combustion.
It is assumed that the engine is turbocharged and intercooled. A working fluid is a mixture of ideal
gases. Compression of fresh mixture in turbocharger is described by an isentropic process equation:
TTC ,out, s
c p ,m
T0
TTC ,in
dT
T
T
c p, m T
T0
dT
T
R ln
pTC ,out
pTC,in
(32)
At the beginning of the in-cylinder compression (point 1 of the cycle) the fluid is composed of fresh
mixture and cylinder residual gas:
n1 nm nr ,
(33)
The amount n1 of the fluid in cylinder is calculated from the equation of state:
p1V1
n1
,
RT1
(34)
Where pressure p 1 results from turbocharger outlet pressure and pressure loss in the intake manifold
p1
pTC,out
pm ,
(35)
nm c p ,m T Tm
nr cv,b T T5
p1 V1 V6 ,
(36)
cv ,u T
T0
dT
T
T1
cv,u T
T0
dT
T
R ln
1
,
CR
(37)
V1
,
V2
(38)
Pressure of the fluid after the ideal adiabatic compression is calculated using the equation of state:
T
p2 p1 2 s CR ,
(39)
T1
After the compression the mixture is burned at constant volume V2 = V3 (v2 = v3) to the state of
thermodynamic equilibrium. Non-stoichiometric approach using minimization of Gibbs free
enthalpy is applied:
U3 p3V3 T3 S3 min ,
(40)
340
The mass of each chemical element remains unchanged before and after the combustion. The mass
balances are the constraints of the equilibrium.
The peak temperature T 3 is calculated from the energy balance of the adiabatic combustion process
at constant volume:
n1 LHVu
cv,u T T2
n3 LHVb
cv ,b T T3 ,
(41)
cv,b
T0
dT
T
T
T3
cv,b T
T0
dT
T
R ln CR ,
(42)
(43)
Parameters of the residual gas result from the pressure of the blowdown process, that is limited by
turbocharger operation. Residual gas pressure p r is calculated due to current power demand for
compression of the fresh mixture. From the mass balance of the engine the flow of exhaust gas can
be calculated:
nex
nm
Mm
,
Mb
(44)
The expansion of exhaust gas in the cylinder from pressure p 4 to the final pressure p 5 = p 6 is
assumed to be also an isentropic process. Final temperature after adiabatic reversible expansion
from point 4 to the final exhaust pressure is calculated from equation:
Tex , s
dT
T
c p ,b T
T0
T4
dT
T
c p ,b T
T0
pex,TC ,out
,
p4
R ln
(45)
Then the energy balance of the turbocharger together with the equations of internal work of fresh
mixture compression and exhaust gas expansion within turbocharger assuming irreversible process
lead to the temperature T 5:
nex c p ,b T T5
5
c p ,b T Tex ,s
mech , T
ex ,s
i, C
c p ,m TC,in TTC,in
(46)
m, C
c p ,b
T0
dT
T
T
T4
c p ,b T
T0
dT
T
R ln
p5
,
p4
(47)
The mass ratio of residual gas to total charge in the cylinder at the beginning of compression is
calculated assuming constant thermodynamic state of the gas between points 5 and 6 [34]:
xr
mr
m1
nr M b
nm M m nr M r
v2
,
v5
(48)
341
nr
p5V2
,
RT5
(49)
Calculation of the cycle parameters is an iterative process. Once it is completed the cycle work is
calculated:
Wcycle W3
W1
m1 cv ,b T T3 cv,b T T4 cv,u T T2 cv ,u T T1 ,
(50)
(51)
Wcycle
mm LHVm
cycle
cv,b T T4
4
cv,u T T2 cv ,u T T
2
1 xr
(52)
Assuming that mechanical efficiency m and electricity generator efficiency g remain unchanged
after the change of fuel the following formulas are used for estimation of electric power, efficiency
and exhaust gas temperature of an ICE module fueled with producer gas:
Wcycle
Pel Pel ,ref
Pel ,ref CP ,
(53)
Wcycle,ref
cycle
el
el , ref
C Eff ,
(54)
Tex ,ref CT ,
(55)
el, ref
cycle, ref
Tex
Tex, ref
Tex,cycle
Tex,cycle ,ref
Where the engine cycle exhaust gas temperature T ex,cy cle is:
Tex,cycle
c p ,b T T5
5
c p ,b T T5 c p ,b T Tex,s
5
ex ,s
c p ,b T
i, T
(56)
ex, cycle
It is also assumed that the ratio of exhaust gas heat to total available engine heat remains constant in
the producer gas mode. Therefore the available jacket water heat QCC can be estimated:
T
QCC
QCC,ref
ex
Hex 393
K
Hex ,ref
Tex , ref
(57)
393K
Where 393 K is the engine exhaust outlet temperature given in the technical specification.
Engine simulation has been performed using in-house built Fortran code. Properties of working
fluids were calculated using JANAF tables. As there are no detailed experimental data available in
the literature, validation of the model was performed using the specification data of the JM S 320
GS engine [36] in natural gas and biogas mode respectively. The calculated values of correcting
factors in biogas mode are as follows: CP = 1.003; CEff = 0.982; CT = 1.022. If these factors are
calculated from the data given in Table 3 the values are: CP =1.000 ; CEff = 1.000; CT = 1.033.
Another validation was made using data from tests of commercial GAS-250 power pack system
purchased from Ankur Scientific Energy Technologies Ltd. (India) [35]. The of engine used within
342
this system is Cummins GTA-1710-G. The electric power generated in producer gas mode was
within the range of 240 - 250 kW and the efficiency was 27.52% - 28.67%. It gives the range of
relative power CP is 0.789 - 0,822 while the model estimated value is CP = 0.794. Efficiency
correction factor CEff is 0.917 - 0.955 while the model gave CEff = 0.919. Finally it was concluded
that the modeling results are in relatively good agreement with the measured parameters.
Table 3. Technical specification of GE J320 GS engine [36]
Engine manufacturer
GE Jenbacher
Engine fuel
Natural gas
Biogas
Fuel gas LHV, kJ/Nm3
34200
23040
Bore, mm
135
Stroke, mm
170
Number of cylinders
20
Compression Ratio CR
12.5
Aspiration
Turbocharged-intercooled
Electric power (at p.f. = 0.8), kW
1054
Fuel energy input (LHV based), kW
2607
M ax cooling water intercooler inlet temperature, C
40
50
3
Combustion air flow rate, Nm /h
4249
4006
Exhaust gas flow rate, Nm3/h
4491
4389
1.695
1.603
Exhaust gas temperature at 100% load, C
427
450
Recoverable thermal output** (water; 90/70C), kW
1197
1147
Electricity generation efficiency (at p.f. = 0.8),%
40.4
EUF in cogeneration mode
0.863
0.844
3
Reference fuel LHV
36.2 MJ/Nm
* value calculated; ** Natural gas - exhaust to 120 C, Biogas - exhaust to 150 C.
343
500 0
300 0
0 .40
200 0
0 .30
0 .20
0 .50
400 0
0 .60
0.2
0 .25
0.3
0.35
0 .4
0.45
0.5
400 0
0 .60
0 .50
300 0
0 .40
200 0
0 .30
100 0
0 .10
0 .00
0.1 5
0 .70
0 .20
100 0
0 .10
500 0
0 .80
0 .70
0.1
600 0
b)
0 .90
0 .80
1 .00
0. 55
600 0
a)
0 .90
1 .00
Cold ga s ef ficiency
0 .00
0.6
0
0.1
0.1 5
0.2
0 .25
0.3
0.35
0 .4
0.45
0.5
0. 55
0.6
900
0. 5
100 0
b)
900
0.5
800
700
0. 4
600
0. 3
500
400
0. 2
300
200
0. 1
Net E UF
0
0.1
E UF and efficie nc y
800
E lectric output, kW
E UF a nd ef f icie ncy
0.6
1000
a)
600
0.3
0.2
0.25
0.3
0.35
400
0 .45
0.5
0.55
300
200
0.1
100
0.4
500
0.2
Net EUF
100
0
0. 15
700
0.4
E le ctri c output, kW
0. 6
0
0.1
0.6
0.15
0 .2
0.2 5
0.3
0.35
0.4
0 .45
0 .5
0.5 5
0.6
10 0
a)
0.9
90
b)
95
85
90
0.9
80
75
0.7
70
75
0.8
70
0.7
65
80
0.8
85
65
0.6
60
ER I
55
0.5
50
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
60
0.6
0.5 5
0.6
ERI
50
0.1
55
0.5
0.15
0 .2
0.25
0.3
0.3 5
0 .4
0.45
0.5
0.5 5
0 .6
solutions. The best value obtained in the analysis is 24.4%. On the other hand the demand for
biomass drying reduces the value EUF for the gasification based technology. Therefore the global
non-renewable energy saving and CO 2 emission reduction are lower than expected in the case of
plant with combustion boiler, where EUF is usually higher than 0.8 (see Fig. 1). The advantage of
the integrated gasification technology is a smaller power range that makes the technology a
potential candidate for distributed plants, if there are no offers for small steam boilers.
7. Conclusions
The complex analysis of downdraft biomass gasifier reciprocating engine system is presented in
the paper. Two mathematical models of gasification and engine processes have been developed.
The models though relatively simple and based only on thermodynamic calculations give results of
a relatively good quality. Therefore these have been used in order to examine the effectiveness of
modification of a typical reactor-engine system, that assumes recirculation of the engine exhaust gas
back into the gasification process. The model gave an encouraging results that need an experimental
verification. There is also an economic analysis required in order to judge a financial attractiveness
of the technology. The profitability of a potential investment projects is nowadays the most
important decision parameter of distributed generation plants.
Acknowledgments
This work was carried out within the frame of research project no. N N513 004036, titled: Analysis
and optimization of distributed energy conversion plants integrated with gasification of biomass.
The project is financed by the Polish M inistry of Science.
The paper has been also supported by the RECENT project (REsearch Center for Energy and New
Technologies) supported by 7th Framework Programme, Theme 4, Capacities.
Nomenclature
ash
ash content in wet biomass
a number of particles of element (C, O, H) in gas component
b constraint equation identifier
c specific heat, J/(kmolK)
C carbon; mass content of carbon in biomass
E energy
EUF fuel energy utilization factor
ERI energy replacement index
g partial free enthalpy of pure component, kJ/kmolK
h specific enthalpy, kJ/kmol
H hydrogen; mass content of hydrogen in biomass; enthalpy, kJ
H
enthalpy flow, kW
m mass, kg
m mass flow rate, kg/s
LHV lower heating value, kJ/kg,
ls number of components,
ls_g number of gaseous compounds
lf number of phases,
M molecular mass, kg/kmol.
n number of moles of substrate
345
fss
related to fuel supply system
g related to gas
gen
at generator output
g-s
temperature difference between gas and solid phases
i component identifier; internal
in input
j phase identifier
m mixture; mechanical
out
output; loss
p at constant pressure
P power
r residual gas
ref reference value
T temperature, turbine
TC
turbocharger
u unburned mixture
v at constant volume
wb wet basis
0 standard state
References
[1] National Fund for Environmental Protection and Water M anagement: M ethodology for
calculation of CO2 emission reduction in the action 9.1 of the EU Operating Program
Infrastructure and Environment 2007 - 2013. Report prepared by National Agency for Energy
Savings KAPE S.A. (in Polish, available at: http://pois.nfosigw.gov.pl).
[2] Decree of the M inister of Economy issued on 26 September 2007 in the matter of calculation of
data within the application form for the certificate of electricity origin from cogeneration...
Journal of Law no. 185/2007 pos. 1314. (in Polish).
[3] Baggio P., Baratieri M ., Grigiante M .: Equilibrium simulation of small-scale gasification
biomass CHP plant. Proceedings of 3rd Int. Conference "Bioenergy 2007". Jyvskyl, Finland.
[4] Schuster G., Lffler G., Weigl K., Hofbauer H.: Biomass steam gasification - an extensive
parametric modeling study. Bioresource Technology 77 (2001), pp. 71-79
[5] Ruggiero M ., M anfrida G.: An equilibrium model for biomass gasification process. Renewable
Energy Vol. 16 (1999), pp. 1106-1109.
[6] Herdin G., Robitschko R., Klausner J., Wagner M .: GEJ Experience with Wood Gas Plants.
Printed
materials
of
GE
Jenbacher
AG,
Jenbach,
Austria,
2003,
<media.godashboard.com/gti/IEA/9revGEJenbacherRauschengl.PDF> accessed 17.02.2011.
[7] Warren T.J.B., Poulter R., Parfitt R.I.: Converting Biomass to Electricity on a Farm-Scale using
Downdraft Gasification and Spark-ignition Engine. Bioresource Technology Vol. 52, 95 - 98.
[8] Bolhr-Nordenkampf M ., Hofbauer H.: Biomass gasification combined cycle thermodynamic
optimisation using integrated drying. Proceedings of ASM E Turbo Expo 2004: Power for Land,
Sea, and Air. June 1417, 2004, Vienna, Austria. ASM E Paper GT2004-53269.
[9] Ragland K. W., Aerts D. J., Baker A. J.: Properties of Wood for Combustion Analysis.
Bioresource Technology Vol. 37 (1991) 161-168.
[10] Ratnadhariya J.K., Channiwala S.A.: Three zone equilibrium and kinetic free modeling of
biomass gasifier a novel approach. Renewable Energy 34 (2009) 10501058.
347
[11]
Altafini C.R., Wander P.R. Barreto R.M .: Prediction of the working parameters of a wood
waste gasifier through an equilibrium model. Energy Conversion and Management 44 (2003)
27632777.
[12] Sharma A.Kr.: Equilibrium modeling of global reduction reactions for a downdraft
(biomass) gasifier. Energy Conversion and M anagement Vol. 49 (2008), pp. 832 842.
[13] Zainal Z.A., Ali R., Lean C.H., Seetharamu K.N.: Prediction of performance of a downdraft
gasifier using equilibrium modeling for different biomass materials. Energy Conversion and
M anagement 42 (2001) 1499-1515.
[14] Babu B.V., Sheth P.N.: M odeling and Simulation of Downdraft Biomass Gasifier.
Proceedings of ChemCon04. M umbai, India 2004.
[15] M elgar A., Perez J.F., Laget H, Horillo A.: Thermochemical equilibrium modelling of a
gasifying process. Energy Conversion and M anagement 48 (2007), 5967.
[16] Jarungthammachote S., Dutta A.: Thermodynamic equilibrium model and second law
analysis of a downdraft waste gasifier. Energy, 2007, vol. 32, no. 9, pp. 1660-1669.
[17] Centeno F., M ahkamov K., Silva Lora E.E., Andrade R.V.: Theoretical and experimental
investigations of a downdraft biomass gasifier-spark ignition engine power system. Renewable
Energy 37 (2012) 97-108.
[18] Baratieri M ., Baggio P., Fiori L., Grigiante M ., Biomass as an energy source:
Thermodynamic constraints on the performance of the conversion process. Bioresource
Technology, Volume 99, Issue 15, October 2008, 7063-7073.
[19] Gbel B., Henriksen U., Jensen T.K., Qvale B., Houbak H.: The development of a computer
model for a fixed bed gasifier and its use for optimization and control. Bioresource Technology
98 (2007), 20432052.
[20] Dogru M ., Howarth C.R., Akay G., Keskinler B., M alik A.A.: Gasification of hazelnut
shells in a downdraft gasifier. Energy 27 (2002), 415427.
[21] Tinaut F.V., M elgar A., Prez J.F., Horrillo A.: Effect of biomass particle size and air
superficial velocity on the gasification process in a downdraft fixed bed gasifier. An
experimental and modelling study. Fuel Processing Tech. Vol. 89 (2008), 1076 1089.
[22] M ilne T.A., Evans R.J., Abatzoglou N.: Biomass Gasifier Tars: Their Nature, Formation,
and Conversion. Report prepared under Task No. BP811010. NREL/TP-570-25357, National
Renewable Energy Laboratory, Golden, Colorado, November 1998.
[23] Gerun L., Bellettre J., Tazerout M ., Gobel B., Henriksen U.: Investigation of the O xidation
Zone in a Biomass two-Stage Downdraft Gasifier, Confrence ECOS 2005, 1525-1532
(Trondheim, 20-23 June 2005).
[24] Gordon S., M cBride B. J.: Computer Program for the Calculation of Complex Chemical
Equilibrium Compositions with Applications; I. Analysis. NASA Reference Publication 1311
(1994) (available at: http://www.grc.nasa.gov/WWW/CEAWeb/RP-1311.pdf)
[25] Larfeldt J., Leckner B., M elaaen M .C.: M odelling and measurements of the pyrolysis of
large wood particles. Fuel 79 (2000) 1637-1643.
[26] Kirov N.Y.: Specific Heath and Total Contents of Coals and Related M ateriale At Elevated
Temperatures. BCURA M onthly Bulletin 1965, pp. 29 33.
[27] Jayah T.H., Aye L., Fuller R.J., Stewart D.F.: Computer simulation of a downdraft wood
gasifier for tea drying. Biomass and Bioenergy 25 (2003), pp. 459 469.
[28] Sridhar G., Sridhar H.V., Dasappa S., Paul P.J., Rajan N.K.S., Mukunda H.S. Development
of producer gas engines. Proceedings of the Institution of M echanical Engineers, Part D:
Journal of Automobile Engineering 219 (2005) 423-438.
[29] Sridhar G., Paul P.J., M ukunda H.S.: Biomass derived producer gas as a reciprocating
engine fuel - an experimental analysis. Biomass and Bioenergy 21 (2001) 6172.
348
[30]
Baratieri M ., Baggio P., Bosio B., Grigiante M ., Longo G.A.: The use of biomass syngas in
IC engines and CCGT plants: A comparative analysis. Applied Th. Eng. 29 (2009) 33093318.
[31] Tinaut F.V., M elgar A., Horrillo A., Diez de la Rosa A.: M ethod for predicting the
performance of an internal combustion engine fuelled by producer gas and other low heating
value gases. Fuel Processing Technology 87 (2006) 135 142.
[32] Dasappa S.: On the estimation of power from a diesel engine converted for gas operation a
simple analysis, Proceedings of the Seventeenth National Conference on IC Engines and
Combustion. India. December 2001 <cgpl.iisc.ernet.in> accessed 17.02.2011.
[33] Papagiannakis, R.G., Rakopoulos, C.D., Hountalas, D.T. and Giakoumis, E.G.: Study of the
performance and exhaust emissions of a spark-ignited engine operating on syngas fuel, Int. J.
Alternative Propulsion, Vol. 1 (2007), No. 2/3, pp.190215.
[34] Heywood J.B.: Internal Combustion Engine Fundamentals. M cGraw-Hill International
Editions. Automotive Technology Series. USA 1988. ISBN 0-07-100499-8.
[35] Ankur Scientific Energy Technologies Pvt. Ltd. (www.ankurscientific.com): Results of test
measurements of commercial Gas-250 power pack system. Baroda, India, 2006.
[36] Jenbacher AG: Technical specifications of JM S 320 GS-N.L Natural gas engine module of
1.065kWel and JM S 320 GS-B.L Biogas engine module of 1.065kWel. Jenbach, Austria, 22
January 2003.
349
Abstract:
2
The islands of Malta are located in the Mediterranean basin enjoying 5.3kWh/m /day of solar insolation, at a
latitude of 35 50N. Electricity generation for the islands is dependent upon imported fossil fuels for
combustion. The available solar res ource could be exploited to offset the current generation of electricity
using solar photovoltaic technology (PV). Due to the limited land availability onshore, the offs hore
environment surrounding the Malt ese islands could be considered for the installation of PV floating on the
sea surface. The output from s uch an installation would have to be integrated with t he existing conventional
electricity generation infrastructure, which relies on gas and steam turbine technology. Gas and steam
turbines have a variable efficiency output depending on the load which they have to deliver, with highest
efficiency delivered at the specified full load of the turbine. Displacing fossil fuel derived electricity with that
from variable and intermittent photovoltaic systems reduces the formers load factors and hence efficiency,
which must be taken account of in feasibility studies.
To assess the feasibility of floating PV being integrated with the existing fossil plant, monthly trend
consumption dat a for Malta were analysed. The change in gasoil and heavy fuel oil (HFO) consumption
resulting from the part load efficiency variation and the displacement of electricity generation from the PVs
were estimated. A cost analysis was prepared for the system integration analysis specifically accounting for
the reduction in combustion of fossil fuels at the power station and the capital expenditures and operating
costs due to the floating PV installation. Aside from the basic cost-benefit of a floating PV installation, CO 2
are also considered.
Keywords:
Floating PV, Offshore solar, Malta PV.
1. Introduction
The islands of Malta are located in the Mediterranean basin enjoying an average daily insolation of
5.29kWh/m2 , at a latitude of 35 50 N, with minimal land space availability due to its 316km2 small
size.Electricity generation for the island is solely dependent upon fossil fuels which are high in CO 2
emissions. According to the EU Directive [1] of the European Parliament and of the Council,on the
promotion of the use of energy from renewable energy sources, it requires member states to achieve
a share of energy of the gross final consumption from renewable energy. As per Annex I of this
directive, Malta is required to produce 10% of its energy from renewable energy sources by 2020.
Malta must find alternative feasible solutions for production of electricity, which is challenging
because the amount of land that can be allocated to energy production is low due to the high
population density.
Solar energy is the primary source of renewable energy available in Malta. Two methods of
generating electricity from solar energy are 1) through concentration of direct sunlight to generate
superheated fluid which is then passed through a heat engine power cycle, and 2) using
photoelectric properties of P-N junctions within photovoltaic (PV) panels. Solar thermal electricity
generation relies on direct sunlight suited to climates with high proportion of clear skies. PV panels
350
convert both direct and diffuse sunlight into electricity, potentially making it more appropriate for
the site.
For high performance of solar energy resource dependent technology such as PV, maximum
exposure to the sun is necessary. This requires a large open space for multi MW installations,
approximately 6,500m2 area for a 1MW crystalline array. This is not easily attainable in a small
country such as Malta, where the land resources are so restricted that uses alternative to power
production, for example agriculture, may take higher priority. PV installed offshore, may offer an
attractive alternative in such situations.
Figure 1: Map of the Maltese Islands showing Territorial Waters (Blue) and Fishing Zone (Red)
[2]
There are no pipeline or cable connections yet available with any neighbouring countries (e.g. Italy,
Libya). Plans are in place for a cable interconnection to Sicily commencing 2013 [3]. Malta has
territorial waters spanning over 3,000 km2 and control over approximately 61,000 km2 of the
Mediterranean Sea [4]. The ratio of area of territorial waters to area of land mass for Malta is
approximately 10, implying that the resource potential from solar electricity generation is
significantly larger offshore than it is onshore.
351
Figure 2: Map of the central Mediterranean showing the southern and western maritime
boundaries of Malta [4]
The waters surrounding the island are almost landlocked by the countries surrounding the
Mediterranean Sea, with ocean access through the Suez Canal and the Strait of Gibraltar water level
variation due to the tidal is negligible around the Maltese coastline and tidal currents are slight.
Also the characteristic waves within the Mediterranean basin tend to be of less destructive nature
compared to ocean waves, due to the shorter fetches over which the wind acts.
The total annual solar radiation on the horizontal surface in the waters offshore of the Maltese
islands (at a Lat. 35.9 and Long. 14.6) is 1932.9kWh/m2 (5.29kWh.m2 /d x 265d) according to the
data given in Table 1.
Table 1: Malta solar and technical PV (for horizontal oriented panels) resource averages including
climatic conditions [7]
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sept
Oct
Nov
Dec
Annual
The calculations undertaken in preparation of Table 1assume that the panels lie perfectly horizontal
on the surface of the lake and that the PV panels generate electricity all year round. Yield estimates
in Table 1 assumed miscellaneous losses of 10% and depict an efficiency of 8% for thin film arrays
and 14% for crystalline arrays as reported by RETScreen [8] and packing density is not accounted
for when calculating the specific yield from either technology.
Marsa Station(Steam)
Delimara Station(Steam)
Marsa Station(Gas)
Delimara Station(Gas)
Delimara Station(CCGT)
Total
Generating
Units
7
2
1
2
1
13
Installed Capacity
(MW)
220
120
37.5
75
110
562.5
A 100km inter-connector power line with Sicily is planned for and is expected to be operational by
the end of 2013, which will increase the available capacity by 200MW and provide opportunity for
Accounts for variation in solar insolation arising fro m overcast conditions and variability.
354
WE WD
Table 3: Base and peak loads for monthly weekday and weekend demand profiles
Enemalta monthly demand profiles for 2010 [3]
Jan Feb Mar Apr May Jun Jul Aug Sep Oct
Base Load (MW) 162 163 161 158 165 181 235 230 216 189
Peak Load (MW) 293 294 302 271 268 293 352 350 335 316
Base Load (MW) 156 165 148 162 167 184 226 203 208 194
Peak Load (MW) 259 290 272 242 238 271 305 291 282 286
based on
Nov
164
286
165
276
Dec
168
323
165
308
From recorded load data supplied by Enemalta [3], the base load for the Maltese power system can
be seen to be in the early morning hours, between 1 5am. The main trend for the demand profile
in winter is an increase in electricity consumption until 8am (at which time a plateau is achieved),
an increase to peak conditions in the evening (~ 6pm) and then a decrease back to the base load. In
summer, the peak is shifted to morning hours, reflecting usage of air conditioning systems. The base
load is higher in summer (c. 225MW) than it is in winter (c. 170MW), when the greatest influx of
tourists is measured on the islands. For the weekends, the base load is generally lower, with two
peaks occurring, the first just before noon and the second in the evening.
Figure 4: Weekday (2010) demand profiles January and August, for Malta [3]
Figure 5: Weekend (2010) demand profiles January and August, for Malta [3]
355
For the analysis of integration of PV, the demand profiles are based on a typical weekday and
weekend profiles for each month. Collectively these represent an annual electrical consumption for
the Maltese Islands of 2,158,817 MWh (Table 4), implying an aggregate load factor for the islands
power stations of 43.8%, where this is considered the ratio of the actual electricity generated to the
maximum possible electricity that could be produced.
Table 4: Representative monthly electricity consumption data according to demand profiles for
2010 [3]
Days
MWh/Day
Total
Month
WEs
WDs
WEs
WDs
(MWh)
Jan
10
21
4966
5604
167344
Feb
8
20
5369
5656
156072
Mar
8
23
5028
5647
170105
Apr
8
22
5060
5415
159610
May
10
21
5078
5593
168233
Jun
8
22
5657
6019
177674
Jul
9
22
6579
7193
217457
Aug
9
22
6015
7002
208179
Sep
8
22
5991
6700
195328
Oct
10
21
5787
6389
192039
Nov
8
22
5252
5510
163236
Dec
8
23
5589
6036
183540
Total (MWh)
2,158,817
Figure 6: Efficiency curve for a single unit 37.5MW gas turbine with part load range 19 32%
(left); aggregate efficiency curve for 3 such 37.5MW gas turbines units (right)
356
The 37.5MW gas turbine units have efficiency ranging from 19% at minimum load to 32% at
maximum load [3]. Using this range as a baseline, a curve representing the varying performance of
the turbine with the load could be depicted (Figure 6 - right).
For three of these units running in a coordinated fashion, with maximum output of 112.5MW, the
part load efficiency calculation was split into three stages:
i) Load: 0 37.5MW : 1 unit working
ii) Load: 37.5 76MW : 2 units working, 1 at full load ( = 32%)
iii) Load: >76MW : 3 units working, 2 at full load ( = 32%)
Table 5: CCGT and GTs peak loads for monthly weekday and weekend
Enemalta monthly demand profiles for 2010 [3]
CC + GTs
Jan Feb Mar Apr May Jun Jul Aug
Weekday (MW) 64
95 118 100 118
85
90 100
Weekend (MW) 39
55
95
86
89
78
68
68
Oct
74
39
Nov
51
51
Dec
69
58
The 8 steam turbine (ST) generators producingelectricity have three different installed capacities
and part load efficiencies (60, 30 and 10MW turbines). The resulting combined steam turbine part
load efficiency is illustrated in Figure 7.
Figure 7: Efficiency curve for an 8 series combination of 10, 30 and 60MW unit steam turbines with
part load efficiency range from 20to 25%
consumption for the conventional generators the PV array integration could be compiled, which is
also indicated in Table 6.The scenarios are the set at 4 and 8MW, in order to allow large scale
integration of the offshore PV technology and hence applicability for commercial economics also
8MW is the threshold at which crystalline PV is somewhat economical.
Steam Power
Station
CCGT Po wer
Station
MWh
MWh
29334
25921
3413
MWh
MWh
MWh
29334
23956
5378
MWh
MWh
MWh
10673
7822
2851
MT
MT
MT
10673
7499
3173
MT
MT
MT
MWh
MWh
MWh
MT
MT
MT
MWh
MWh
MWh
MT
MT
MT
408658
403722
4935
82967
82085
882
1744373
1745500
-1127
506430
506270
161
408658
398529
10128
82967
82092
875
1744373
1745437
-1064
506430
506286
145
MWh
MWh
MWh
MT
MT
MT
MWh
MWh
MWh
MT
MT
MT
This wave energy converter (WEC) was used as a reference in comp iling the economic case, due to it being the only
commercial W EC installation to date and would signify expected costs for mooring arrangements of an offshore large
scale PV installation.
358
current prices, means the cost of a 2.25MW wave park would be $12.38M. Moorings, installation,
grid connection and project management account for a total of 24% of the total investment [14] in a
wave farm. Hence the infrastructural and installation costs can be taken to be approximately
$1320/kWp for a marine floating PV array of the same rating. The cost per Wp for crystalline PV
are taken to be $2.07 and $1.03 for the thin film [15]. Thus the specificCAPEX was estimated to be
$3390/kWp and $2350/kWp for crystalline and thin film floating PV systemsrespectively. For the
operating and maintenance costs of the project, US $2700/MW/year and US $1800/MW/year were
taken to apply for both the crystalline and the thin film technologies as indicated in research
comparing large scale application of PV technologies [16]. Using an average conversion rate of
0.7533 $/ for 2010, these rates were converted to those shown in Table 7.
Table 7: Capital expenditures and O&M costs for the power station and PVs
Annuitized Costs (i = 10%)
Inc. Fuel Costs
Year
Generating Station
CAPEXper MW/year
O&Mper MW/year
2010
Malta Power Stations
131,334
107,912
2010
Offshore Crystalline
290,625
1,971
2010
Offshore Thin Film
201,466
1,314
Thesecosts for the thin film and crystalline PV were taken together with those for conventional
power generating capacity to estimate the average annual cost per unit of electricity for Scenario 1
and 2 operating in an integrated fashion. The results are illustrated in Table 10 calculated using the
costs of investment and operation in Table 8 and Table 9.
Table 8: Investment and operation cost estimates for Scenario 1 and 2
Installed Capacities
Powe r Station Costs
Power
PVs
CAPEX
O&M
Stations
Million
Million
(MW)
(MW)
EUR/year EUR/year
W/O PV Integration
562.5
73.88
60.70
W Thin Film PV
562.5
4.00
73.88
60.70
1
W Crystalline PV
562.5
4.00
73.88
60.70
W Thin Film PV
562.5
8.00
73.88
60.70
2
W Crystalline PV
562.5
8.00
73.88
60.70
PV Costs
CAPEX
O&M
Million
EUR/year
0.81
1.16
1.61
2.33
Million
EUR/year
0.0053
0.0079
0.0105
0.0158
1
2
GTs
MWh
W/O PV Integration 29,334.0
W Thin Film PV
25,920.9
W Crystalline PV 25,920.9
W Thin Film PV
23,955.9
W Crystalline PV 23,955.9
Electricity Generation
CCGT
Steam
PV
MWh
MWh
MWh
408,657.5 1,744,373.0
403,722.4 1,745,500.0 7,221.2
403,722.4 1,745,500.0 7,221.2
398,529.2 1,745,437.0 14,442.4
398,529.2 1,745,437.0 14,442.4
359
Fue l Consumed
GTs
CCGT
Steam
Mtonnes Mtonnes Mtonnes
10,672.5 82,966.8 506,430.2
7,821.6 82,084.9 506,269.6
7,821.6 82,084.9 506,269.6
7,499.0 82,091.7 506,285.6
7,499.0 82,091.7 506,285.6
Cost of Electricity
(Total)
EUR c/kWh
17.114
17.051
17.068
17.081
17.113
The costs of electricity reported in Table 10 show the averaged cost per unit of electricity at the
power stations when integrated with the two scenarios analysed. For scenario 1 using thin film,
which has the lowest cost of electricity for the scenarios considered, the annual savings which could
be made through the offsetting of fossil fuel generation is of $1.376 million. For higher installed
capacities of offshore PVs investigated the demand variability faced by the integrated system
implied that more fuel was being consumed to balance off the system utilising only part load
efficiency, and this was not compensated for by electricity generated from the PV arrays. This
limited the techno-economical viability of offshore PVs to approximately 14-15MW, at which point
the cost of electricity from the integrated PV - power station generation would be equal to the cost
of electricity from the power station alone.
Thin Film PV
kgCO2 /m2
21.3
20.1
2.3
19.9
2.3
18.4
0.13
84.43
Table 11 presents the list of processes which emit CO2 during the manufacturing, installation and
transportation of the PV arrays. According to the International Energy Agency [19] the CO 2
emissions from the generation of a kW in Malta was at 850g CO 2 /kWh for 2009, which reflectsthe
average 3.09 g CO 2 /Tonne of fuel combusted at the power stations in Malta. For the installed
capacities specified by Scenario 1 and 2, carbon savings are over twenty times higher the embedded
PV carbon for thin film and crystalline installations in both scenarios.
360
4. Conclusion
New renewable energy generating capacity in Malta must consider variable demand in technoeconomic studies, offshore PV is no exception. This paper has shown that PV integration will lead
to an overall reduction of generating costs across the system.With the inter-connector (Malta to
Sicily) in place, by the end 2013, it could provide more flexibility in the system with ability to
absorb some further excess generation from either renewable energy technologies or to operate the
current system more efficiently at full load.
The carbon savings through the adaptation of offshore PVs are considerable and a 12MW
installation would be able of providing 1% of the electricity generation in Malta, equivalent to 10%
of the total 2020 electrical renewable target as specified by the EU. Future work will address
environmental implications of such a development, as well as the actual electrical yield from a
similar system.
Acknowledgments
The authors acknowledge the provision of data supporting this research by Enemalta, as well as
valuable contributions regarding plant operation. MIRARCO is also acknowledged for its financial
support for the lead authors research.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
European Comission (EC), Directive 2009/28/EC of the European Parliament and of the
Council. Official Journal of the European Union; 2009.
UoM, PJS BENSPEFISH, University of Malta, Malta. Available at:
<http://www.um.edu.mt/science/biology/staff/profpatrickschembri/fisheries_management_zo
ne_fmz> [accessed 13.2.2012]
Enemalta Corporation, Plant system load and operation data. Personal Communication: 2011
Oct 12th .
Blake, G., Coastal state sovereignty in the Mediterranean sea: the case of Malta. GeoJournal
1997; 14(2):173 180.
Woody, T., Solar on the water. New York, US: The New York Times; 2011 April 20th .
Ueda, Y., Sakurai, T., Tatebe, S., Itoh, A., Kurokawa, K., Performance analysis of PV
systems on the water. Proceedings of the 23rd European Photovoltaic Solar Energy
Conference; 2008 Sep 1-5; Valencia, Spain. 2670-2673.
NASA. Surface meteorology and solar energy. Atmospheric Science Data Center. Available
at: <http://eosweb.larc.nasa.gov/sse/RETScreen> [accessed 28.9.2011].
RETScreen. RETScreen clean energy project analysis software. Natural Resources Canada:
PV4 Module, 2010.
Alstom, Alstom partners with Nexans for Malta-Sicily submarine electrical interconnection.
Available at: <http://www.alstom.com/grid/news-and-events/press-releases/Nexans-MaltaSicily/>[accessed 20.1.2012].
Enemalta Corporation, Electricity generation plan: 2006 2015. Marsa, Malta. Available at:
<http://www.enemalta.com.mt/enemaltastorage/images/files/archived%20news/generation%2
0plan%20%2821.06.06%29.pdf> [accessed 19.1.2011].
CEPCI. Chemical engineering plant cost index. Available at: <http://www.che.com/pcritical/>
[accessed 15.2.2011]
Platts, European marketscan January release. 2011; 43 (1).
Burman, K., Ocean Energy Technology Overview. Energy Efficiency & Renewable Energy:
US Department of Energy, 2009.
Callaghan, J., Results of the marine energy challenge: cost competitiveness and the growth of
wave and tidal stream energy. London, UK: Carbon Trust, 2006.
361
[15] SunElectronics, Solar panels and inverter price comparison. Available at:
<http://sunelec.com/> [accessed 21.11.2010].
[16] M. Ito, K. Kato, K. Komoto, T. Kichimi, K. Kurokawa, A Comparative Study on Cost and
Life-Cycle Analysis for 100MW Very Large PV Systems in Deserts Using m-Si, a -Si, CdTe
and CIS Modules, Progress in Photovoltaics: Research and Applications, 2008; 16: 17 - 30.
[17] Myrans, K., Comparative energy and green carbon assessment of three green technologies for
a Toronto roof [dissertation]. Toronto, Canada: University of Toronto; 2009.
[18] Pacca, S., Silvraman, D., Keolelan, G. A., Life cycle assessment of the 33kW photovoltaic
system on the Dana building at the University of Michigan: Thin film laminates, multicrystalline modules and a balance of system components. Michigan, US: University of
Michigan. Available at: <http://css.snre.umich.edu/> [accessed 8.3.2011].
[19] IEA, CO2 fuel emissions from fuel combustion highlights 2011 edition. International
Energy Agency. Available at: <http://www.iea.org/co2highlights> [accessed 18.1.2011].
[20] Cazzaniga, R., Rosa-Clot, M., Tina, G. M., Floating tracking cooling concentrating (FTCC)
system.
Scienza
Industria
Technologia.
Available
at:
<http://www.scintec.it/ricerca/energia/ftcE.html>[accessed 11.12.2011].
[21] Campbell, H., Solar power proposed for remote mines. Sudbury, Canada: Sudbury Mining
Solutions Journal; 2011: 8.
[22] Cangini, E., Loto progetto energia. Consorzio di bonifica della Romagna occidentale.
Available
at:
<http://www.bonificalugo.it/uploads/ba/60/ba60299310a79a5d6325208235d3ecf4/LOTOprogetto-energia.pdf>[accessed 22.12.2012].
362
Abstract:
The article presents the issue related to power production from gas received during biomass gasification
process and advantages of biomass gasification technology use in cogeneration systems. Re search
installation for power production from biomass ( 60 kWt), developed at the Institute for Chemical Processing
of Coal in Zabrze, is characterized. The installation consists of a fuel supply system, gas generator with
fixed-bed, dry gas cleaning system and the system used to produce electricity. The article describes
installation development process which was divided in three staged. The influence of four different biomass
available on polish market, on quality of received process gas i s presented as well as received reduction
efficiencies of dust and organic contaminations in proposed new conception of dry purification of gas system.
Power generation system supplied by process gas, equipped with dual fuel combustion engine with rated
power of 25 kW is pre sented. The article presents re sults of work of power generation system for
production of 8 and 15 kW of electricity, by engine supplied with diesel oil and process gas.
Keywords:
Biomass, Gasification, Research installation, Double-fuel piston engine, Electricity
1. Introduction
The process of producing electrical energy and heat in the coming years will be connected with the
taking up of, with a more intensive effort, the aim of lowering the harmful influences of the energy
industry in man's natural environment [1]. One of the interesting directions, which may bring the
desired ecological effects, may be the use of renewable sources of energy, biomass in particular.
Because of the specific properties of biomass there is the possibility of it becoming an interesting
solution in gasification technology [2,3,4]. The conversion of chemical energy contained in biomass
for energy of combustible gas is meaningful for a larger area of energetics used in biomass. The use
of processed gas as a fuel for diffused and average co-generation systems of the production of
electrical energy and heat may result in the achievement of the minimal established value of the
OZE system in the general balance of used fuel. Furthermore, these systems allow the use of local
sources of fuel in the places where they occur, which thus eliminates the cost of transport. An
important asset of these systems, based on biomass gasification technology, is likewise the
possibility of using solid fuel, often waste products e.g. from the wood or food industry.
2. Research installation
The introduction of new technological solutions onto the energy sector are usually preceded by a
series of well-used research tests on a smaller scale of the proposed technology of innovative
equipment [5, 6]. On a lesser scale it allows one to check the accuracy of the applied solution. A
similar method of proceedings was adopted in the case of the development of the installation of
biomass gasification with an innovative gas generator called GazEla [7].
363
The experimental installation was built, started up and tested by The Institute for Chemical
Processing of Coal in Zabrze.
During the research there appeared a gradual development of the position, which was connected
with, among other things, the accepted schedule of the experimental work. There also occurred
during the process of research the necessity of making changes in the construction with the aim of
improving reliability and the efficiency of both particular parts of the apparatus and of the
equipment in the composition of the installation.
The conducted research of the development of the installation may be detailed in three stages:
- Stage I - the starting of the gas reactor together with the biomass measurement system in the
processed gas combustion chamber.
- Stage II - extension of the installation in stages, through the biomass drying system of gas
purification,
- Stage III - the integration of a double-fuel piston engine with the installation of biomass
gasification equipped within the gas purification system.
Stage I: Research of the experimental installation of biomass gasification involved work connected
with the starting up and examining of the GazEla gas generator - fig. 1 [8].
Fig. 1. Outline of the examined installation during the first stage of experimental research
364
Stage II is the introduction of the researched concept of the system of the purification of drying gas
produced in the GazEla generator with the aim of attaining a level of purified gas which allows for
its combustion in a piston engine.
Processed gas is generated in the reactor in the first stage of purification which is then directed to an
inertial flash vessel. The next is the transportation of the gas to the cyclone separator with the aim
of removing minute particles of dust contamination. Initially, purified gas is directed to the gas
cooler where it undergoes the process of reducing the temperature from approx. 120C to approx.
30 40C with the aim of condensation contained in the organically contaminated gas and steam
from water. The condensed damp and contamination are gathered in lower exchanger collector. In
the next stage of purification, the gas flows to two coke filters. The filters determine the rinsing of
the coke of granulation 10 20 mm. The purified gas is then burnt in the combustion chamber, and
the remaining combustion is emitted through the chimney into the atmosphere [9].
After achieving a satisfactory level of purified gas we move on to stage III, which depends on the
integration with the previously described system of the piston engine with the aim of practically
examining the possibilities of creating in the system co-generated electrical energy together with
heat.
A current-generating system from the firm Andoria from Andrychow was used during the
experimental tests. The adaptation of the engine in the research was carried out by The Institution of
Combustion Engines, Department of Environmental Engineering and Energetics of the
Polytechnic of Silesia, Gliwice. This engine has a strength of 24,6 kW (33,5KM ), 3620 cm3 and
also 1500 rev./min. The engine was produced in 1971.
During this research there was used an engine with an automatic mixed ignition working in a twostroke system. During the setting in motion phases of the engine there was used only diesel fuel.
The produced electrical energy collected of the generator was with the help of a resonance system
of power collection, a cooled air flow. The work of the engine is dependent on the load of the
collecting system of electric power. The forced thrust of electric power on the load currentgenerating aggregate causes an automatic change in the flow of the collected fuel with the help of a
regulator of the internal engine.
After achieving the nominal work parameters of the gas generator and also the purification system
there followed on the starting up of the current-generating aggregate. The next step was the opening
of the shutter which allowed the entrance of the processed gas, together with the inducted air, to the
process of combustion in the engine. In figure 2 we have presented the final outline of the gas
purification system integrated with a two-stroke piston engine.
The final appearance of the biomass gasification system integrated with the drying system of gas
purification and an internal combustion engine consists of a bucket feeder, a biomass intermediary
tank, a conveyor worm, a GazEla gas generator, an inertial pressure reducing valve, a cyclone, a gas
cooler, a water processing tank, a cox filter, a fabric filter, a combustion chamber and a two-stroke
piston engine.
365
Fig. 2. Outline of the gas purification system integrated with a two-stroke piston engine
Test
Designation
Test I
Test II
Test III
Test IV
366
Fuel
Chips II
Pellets II
Pellets I
Chips I
Fig. 4. The rate of progress of the main ingredients of processed gas for chips II - Test I
The rate of progress of the internal temperature of the reactor is shown in figure 5.
It is noticeable that there is greater intermittent fluctuation of the gas contribution brought out by
the prescribed measurement of the fuel to the reactor and also a temporary growth in the governing
pressure inside the generator. In figure 4 there appears less fluctuations during the depicted six
367
hours of work and this is a normal phenomenon for this type of equipment. A similar situation took
place on the scale shown in figure 5.
During the test conducted with the generator, the flow of processed air was distributed to a
particular zone of the equipment. The average value sum of three flows was approx. 21,6 m3n/h.
Unit
Chips II
(Te st I)
Chips I
(Te st IV)
6,1
Wood I
pelle ts
(Te st III)
6,7
Straw II
pelle ts
(Te st II)
8,9
H2
7,5
CH4
2,1
2,5
0,9
2,6
CO
25,0
19,5
22,0
20,3
CO2
9,5
11,8
12,9
12,8
O2
0,0
0,0
0,0
0,0
N2
55,9
60,1
57,5
55,4
Wd
MJ/m 3 n
4,72
4,01
3,82
4,45
It is likewise important to notice that for some types of biomass it is essential to modify the manner
of conducting the gasification process. This is particularly important in the case of straw, for which
it was essential during the research to reduce the load in the combustion carbonised zone with the
368
aim of lowering the resulting high temperatures or the essential minimalising of the amount of
remaining ash crust. The appearance of high temperatures in the deposit of fuel during the
gasification tests of straw pellets led to the appearance of crusts in the lower parts of the system.
This was particularly problematic in the exploitation of the system resulting in the appearance of
slag. This appears in areas characterised by the highest temperatures, namely, in the area of the
suction chamber and also in the grate. The next problem then was the exceeding of temperatures;
softening; melting and the flow of slag, which confirmed the analysis conducted in the laboratory.
was at a of level 36,2 kg/h. The observed average value of fuel for the used dry gas was 3858,5
kJ/kg.
The performance of producing electrical energy was observed with the aim of grading the effects of
the energy of the work of the system el .
In describing the performance of producing electrical energy in the research of the currentgenerating unit we may highlight it using the following contingents:
Nel
(1)
el
E ch
where: Nel
E ch
The exploitation research of the combustion of processed gas in a two-stroke piston engine
conducted with three variants:
- variant A - The engine output on the basis of fuel (i.e. diesel fuel), and electrical power was level
at 8 kWel,
- variant B - The engine output on the basis of a mixed substance consisting of diesel and
processed gas of 8,3 kWel,
- variant C - The engine output on the basis of diesel and processed gas of 15,5 kWel.
In variants B and C the directed flow of processed gas to the piston engine was always 100% of the
value of the created flow in the gas generator (i.e. used permanent value of 42,45 kg/h)
During the work of the current-generating unit only on the basis of fuel (variant A) the contribution
of diesel in the fuel used was 100%. The collected flow through the fuel engine was level at 4,05 l/h
during the production of electrical power at a level of 8 kWel. The temporary performance of the
chosen system with contingents (1) was 16,54%, at the same time the value gained for the arranging
of the angle of the injection for optimal fuel for the combustion process was a mix of processed gas
and diesel. In the case of variant B the flow of diesel was level at 1,42 l/h, and the observed
electrical performance was 15,62%. For nominal conditions of the work of the engine (variant C)
with electric power at 15,5 kWel the flow of diesel was formed at a level of 2,4 l/h. The performance
of the creation of electrical energy in this case was 24,61%.
From the above results we may notice that the highest performance of the system is gained by the
burning of a mixture of processed gas and diesel fuel with the dominating advantage being from
the first of these ingredients and with nominal power from a current-generating unit.
6. Summary
The development of biomass gasification is currently possible depending on the solving of some
key problems. These are connected with its technology. Among other things, the improvement of
the gas generator reliability, the development of gas purification systems characterised by low
levels of complication, connected directly with the lowering of investment costs and the
exploitation of the installation. The level of gas purity, directly after the value of fuel, determines
one of the fundamental parameters in deciding about the possibility of the further use of the gas
process, e.g. in internal combustion engines. A vital aspect in the resolving of the above issues is
experimental research conducted on the installations characterised by a semi-technical scale
(exploitation) piloting and then a demonstration. Such research would permit the creation of a
mature technological solution, which would then allow commercial application of the abovedescribed technology.
370
During the development of the installation there were many hours of testing with the aim of
controlling the output of the whole installation. Particular attention was given to the gas generator in
order to achieve the output of the equipment without any difficulties. The value of fuel gained from
the drying gas process with a GazEla generator contained in the compartment from 3,82 to 4,72
MJ/m3n which was appropriate to its value and which is presented with the assistance of national or
international publications. For a wood chip composition of drying gas contained within the
following ranges: H2=6,17,5%, CH4=2,12,5%, CO=19,525,0%, CO2=9,511,8%,
N2=55,960,1%. In the case of pellets being the main ingredients of gas, the results were
respectively:
H2=6,78,9%,
CH4=0,92,6%,
CO=20,322,0%,
CO2=12,812,9%,
N2=55,457,5%. The change of the main ingredients of gas are not great and the acquired
concentration process of the specific ingredients are stable. The nominal flow of added fuel to the
testing reactor was measured at a level of approx. 14 15 kg/h. The experimental generator is
characterised by a power nominal of approx. 62 kWt.
The presented development of the dry-purifying system of the gas process and also its final
configuration allows for the achieving of an average performance of the removal of ash
contamination of approx. 95%, and of organic contamination of approx. 47%. Further development
of this installation will issue from the requirements of specific products of the internal-combustion
engine.
Purified gas is used as a source of chemical energy in a two-stroke piston engine in the creation of
electrical energy and also heated in the form of hot combustion gas. The flow of obtained electrical
energy during the tests depended on the load of the system's collection of power, approx. 8,3 or
15,5 kWel. This was adapted to the research of the engine, of which the basic fuel was diesel during
the gas-burning process and liquid fuel was characterised by obtaining it in relation to a low
performance in comparison with the present commercial solutions. The optimum achieved
efficiency of created electrical energy for the nominal parameters of the engine output (i.e. for
electrical power 15,5 kWel) was approx. 24,6%. At the same time gas processes determined 61,8%
of chemical energy fed into the engine, and diesel fuel 38,2%.
Acknowledgements
The strategic program of scientific research and experimental development of the National (Polish)
Centre for Research and Development: Advances Technologies for Energy Generation; Task 4.
,,Elaboration of Integrated Technologies for the Production of Fuels and Energy from Biomass as
well as from A gricultural and other Waste M aterials.
References
[1] [1] Chmielniak T.: Role of different types of Technologies in achieving emission target up to
2050, Rynek Energii (Energy M arket), 92 (2011), 3-9.
[2] [2] Chmielniak T., Skorek J., Kalina J., Lepszy S.: Integrated power systems with biomass
gasification. Gliwice, 2008. Wydawnictwo Politechniki skiej.
[3] [3] M cKendry P.: Energy production from biomass (part 3): gasification technologies.
Bioresource Technology, 83 (2002), 55-63.
[4] [4] Kirkels A.F, Verbong G.P.J.: Biomass gasification: Still promising? A 30-year global
overview. Renewable and Sustainable Energy Reviews, 15 (2011), 471-481.
371
[5] [5] Najser J., Ochodek T., Ch ond R.: Character of operations of the biomass gasification
installation and economic aspects of electricity generation. Rynek Energii (Energy M arket), 85
(2009), 68-74.
[6] [6] Plis P., Wilk K. R.: Theoretical and experimental investigation of biomass gasification
process in a fixed bed gasifier , Energy, 36 (2010), 3838-3845.
[7] [7] Polish Patent application: P-383541 (2007).
[8] [8] Sobolewski A., Kotowicz J., Iluk T, M atuszek K.: Influence of biomass type on operating
parameters of the gas generator with fixed bed, Rynek Energii (Energy M arket), 82 (2009), 5358.
[9] [9] Kotowicz J., Sobolewski A., Iluk T.: Parametric Analysis of Biomass Gasification
Installation Integrated with a Combustion Engine, Proceedings of the 24rd International
Conference ECOS 2011, 04-07th July 2011, Novi Sad, Serbia, s.3658-3667.
372
Abstract:
A Pre-Industrial Scale Anaerobic Digestion Cell, working in batch mode in the Solid State (i.e. Total Solids
content higher than 25% w/w) was built for the co-treatment of Animal and Agricultural Residues.
The anaerobic cell has a total volume of about 170m3 and is able to treat about 95 tonnes of biomass per
cycle. The aim is to perform the Anaerobic Digestion process with a high solids content to reduce, or
completely eliminate, the waste liquid fraction usually produced by other types of Anaerobic Digestion
technology. This could lead to the production of renewable electrical energy from animal manure and other
agricultural residues in a more sustainable way.
The cell was built on an existing Italian farm that raises about 135 tonnes per year of live weight swine, and
also produces other agricultural residues.
The problem of the starting the Anaerobic Digestion inside the cell without inoculum was analysed and
different strategies were evaluated. The one based on aerobic pre-treatment of the mixture before the
anaerobic process inside the cell seems to give interesting results both for the rapidity of starting the
methanogens phase as well as for the quality of the biogas produced.
The results show that 15 days of aerobic pre-treatment led to the production of a biogas with a biomethane
concentration of about 50%v/v after 10 days of anaerobic treatment. Further, the whole anaerobic process
can be considered practically finished after about 65 days.
Keywords:
Agricultural Residues, Biogas, Renewable Energy, Solid State Anaerobic Digestion, Swine manure cotreatment
1. Introduction
The production of energy from renewable sources is one of the most crucial issues to be pursued in
the EU [1], in order to achieve a more sustainable and environmentally sound development [2].
Among the different renewable sources, biogas, a biological gas rich in methane, is the one that
gives a relevant contribution to the total energy demand of different European countries, and the
one that shows a very high potential for growth in the near future [3]. Sources of biogas may be
energy crops [4], [5], various biodegradable residues, livestock manure [6], [7] and the organic
fraction of municipal solid waste [8], [9]. From the energy and ethical points of view, the most
suitable sources are from waste and residues. In particular, the exploitation of livestock manure in
Anaerobic Digestion (AD) facilities has a very high potential for spreading in the different countries
of the EU 27. The most diffused AD technologies work with a TS content generally less than 15%
w/w, causing the production of a large amount of digestate. This digestate has a high concentration
of compounds with a high COD (Chemical Oxygen Demand) and N, together with other potentially
polluting compounds such as P, S, Cu Zn and pathogens. Due to the high liquid phase in which
these compounds are dissolved, they are a serious threat both for the ground and surface water as
well as for the air (i.e. gaseous emissions) and for agricultural fields (i.e. heavy metal content). For
373
these reasons, the common practice of spreading livestock manure on agricultural fields is
becoming more difficult, also as a consequence of the last EU Directive 91/676/CEE on water
resource protection. A possible solution to the aforementioned problems could be the adoption of an
AD process able to operate at a high TS content, to reduce the amount of the liquid fraction
discharged by the process. The solid fraction arising from AD can be quite easily managed by
aerobic stabilization, to obtain, at least, a high quality organic fertilizer, hence becoming a resource
for the environment. This aim can be pursued by Solid State Anaerobic Digestion (SSAD), which is
a batch static anaerobic digestion process [8], able to work with a TS content from 25% up to 50%
w/w. The amount of liquid fraction discharged by this process is usually less than 10% w/w of the
treated mass. Livestock manure generally have a low TS content, <15%w/w, not compatible for use
in SSAD. Hence, to achieve a suitable moisture level, it is necessary to mix the manure with other
materials that come from agricultural and pruning residues. For this reason a pre-industrial scale
SSAD cell was built on an existing swine farm of about 135 tonnes/year of live weight swine.
Before co-treatment in this SSAD cell, the manure produced by the swine was first mixed with
other crop and trimming residues to reach a suitable moisture level. Another problem concerning
the SSAD cell is the production of suitable inoculums to activate the process [10], [11]. In this
study, the problem concerning the start-up of the SSAD cell and the amount and quality of the
biogas able to be generated from the mixture produced on the farm was analysed both from the
theoretical and experimental points of view.
Value
11,000
6
25
50+10
0.075
Unit
kg
kg
day
litres/day per kg (1)
(1)
kg of live weight.
Currently manure is managed by mixing it with other agricultural and trimming residues, to obtain a
mixture with a TS content of about 20-25% w/w. Then the mixture is aerobically treated in a
dedicated facility for the production of an organic fertilizer. The facility consists of a concrete
trench in which moves a crane bridge with screws able to mix and aerate the mass under treatment
(Fig. 1). At the end, the stabilized material is spread on fields for agricultural needs. All the
agricultural and trimming residues, along with the swine manure, have a high BP that is currently
completely unexploited. Hence they could be an important source of renewable energy for
increasing the efficiency and environmental sustainability of the farm.
For this aim a pre-industrial scale cell able to perform the SSAD process was built and now it needs
to be activated (Fig. 2). The total volume of the cell is about 170 m3. This means that at full
operating conditions it will be able to treat about 90 tonnes of biomass per cycle. For this reason the
problem of generating a suitable amount of inoculum able to allow a rapid production of high
methane rate is crucial. The correct inoculum is fundamental for making the process start in as short
a time as possible, together with the production of a high-energy content biogas. Furthermore, the
inoculum has to be generated by material that is produced on the farm and is classified as biomass.
374
Using inoculums from other AD facilities seems to be complicated and costly, considering that in
the area there are no AD facilities operating by the SSAD process.
8.5, followed by spectrophotometric detection. For heavy metals, samples were digested according
to the US EPA 3050B method [12]. Heavy metal concentrations were determined by flame atomic
absorption spectrophotometry using a Shimadzu AA-6800 apparatus.
Fraction
7.10
17.6
75.3
Unit
% w/w
% w/w
% w/w
The mixture was treated aerobically from 0 to 15 days, and every three days a SSAD process was
activated. The SSAD process was simulated in cylindrical, gastight reactors, of about 12 litres each,
able to process about 5 kg of the mixture with no inoculum (Fig. 3). The SSAD process temperature
was kept at mesophilic conditions (3082K) by the aid of a thermal bath. The biogas produced
during the SSAD test was stored in a gasometer device for volume and composition determination.
The CH4 and CO2 content (%v/v) of the biogas was determined by infrared sensors (1%), whereas
the H2S and O2 content (%v/v) was determined by electrochemical cell sensors (2%). Other biogas
components, such as ammonium, chlorine, and others, were included in a balance term.
The BP and BMP of the mixture were determined by a triplicate mesophilic test using a 10%TS
mixture inoculated with cow manure digestate produced in mesophilic laboratory tests. A blank test
was used to determine the contribution of the inoculum to the BP and BMP values.
376
The BP runs (Fig. 4) show a lag phase of about 20 days and a total biogas production of about 85
Nl/kg VS, calculated on the average of the three BP runs even if after about 120 days the runs had
not yet finished. There was a similar trend for the biomethane production curve (Fig. 5) that reached
a maximum value of about 55 Nl/kgVS, about 65% v/v of the entire biogas production. After about
130 days, the SSAD runs gave different results that were to some extent and in some cases greatly
influenced by the duration of the aerobic pre-treatment (Fig. 6).
Table 3. Chemical and physical characterization of the mixture.
Parameter
moisture
VS
pH
EC
C
N
Value
72.0
88.7
6.56
4.30
39.1
1.55
Unit
% w/w
% TS
mS/cm
% TS
% TS
100
90
80
Nl/kg VS
70
60
50
40
30
20
10
0
0
10
20
30
40
50
60
70
80
90
100
110
120
Day
NLCH4/kg VS
70
60
50
40
30
20
10
0
0
10
20
30
40
50
60
70
80
90
100
110
day
120
100
90
80
0 Days
3 Days
6 Days
9 Days
12 Days
15 Days
Nl/kg VS
70
60
50
40
30
20
10
0
0
10
20
30
40
50
60
70
80
90
100 110
120 130
day
0 Days
3 Days
6 Days
9 Days
12 Days
15 Days
NI/kg VS
2,0
1,5
1,0
0,5
0,0
0
10
20
30
40
50
60
70
80
90
100 110
120 130
day
0 Days
3 Days
6 Days
9 Days
12 Days
15 Days
NlCH4/kg VS
40
30
20
10
0
0
10
20
30
40
50
60
70
80
90
100 110
120 130
day
378
As expected the SSAD run of the sample with 0 days of pre-treatment had a very low biogas and
biomethane production. This is a consequence of strong inhibition phenomena due to the high VS
concentration causing an Organic Load of the process of about 280 gVS/kg. Hydrolysis and
successive alcoholic fermentation cause a high production of volatile fatty acids, leading to an
organic overload along with a reduction in pH. As a consequence the methanogens bacteria are
greatly inhibited and biomethane production was practically null in the first 60 days of the process.
The runs on samples with longer pre-treatment times showed a higher production of biogas and
biomethane. In particular, a longer aerobic pre-treatment is able to reduce inhibition during the
methanogens phase, leading to a higher production of biogas and biomethane in the first days of the
SSAD process (Figs. 7 and 8). The higher biomethane production also indicates that the
methanogens bacteria were operating in rather favourable conditions. The sample with 15 days of
pre-treatment was able to produce a biogas with a methane concentration of 50% v/v after about 10
days from the start of the process, whereas the other SSAD runs reached a comparable biomethane
concentration after about 25 to 50 days from the start of the process. Further, the biomethane
production phase of the sample with 15 days of pre-treatment can be considered practically finished
after about 65 days from the start of the SSAD process. From that time, the resulting digestate could
be used as inoculum for starting a new SSAD process.
4. Conclusion
The problem of the start up phase of the pre-industrial scale Solid State Anaerobic Digestion cell
can be solved by the approach investigated in the present work. The mixture used gave a BP of
about 85 Nl/kgVS and a BMP of about 55 Nl/kgVS. The aerobic pre-treatment gave a very positive
effect on the anaerobic phase and in particular the methanogens phase, when operating with SSAD
runs without inoculum. The longer is the aerobic pre-treatment time, the faster is the start of
methanogens activity. After 130 days the sample with no aerobic pre-treatment gave a very low
biogas production and almost no biomethane. For the samples with longer aerobic pre-treatment,
both the biogas and biomethane production were quite high, from 45 to 70 Nl/kgVS and from 25 to
40 Nl/kgVS, respectively. Fifteen days of aerobic pre-treatment was the most suitable for use in
starting up the SSAD process. In this case, the biomethane reaches a concentration of about 50%v/v
in about 10 days and the anaerobic process can be considered concluded in about 65 days.
Acknowledgments
The authors would like to acknowledge the Agricultural Office of the Umbria Region for the
economic support given for this study.
Nomenclature
AD
BMP
BP
C
EC
SSAD
T
TS
VS
Anaerobic Digestion
BioMethane Potential (Nl/kgVS)
Biogas Potential (Nl/kgVS)
Carbon
Electrical Conductivity (mS/cm)
Solid State Anaerobic Digestion
Temperature (K)
Total Solids (%w/w)
Volatile Solids (% TS)
379
References
[1] Directive 2001/77/CE of the European Parliament and of the Council of 27 September 2001 on
the promotion of the electricity produced from renewable energy sources in the internal
electricity market. Official Journal of the European Communities 27.10.2001.
[2] EUROPA- the official web site of the European Commission. Climate Action Available at: <
http://ec.europa.eu/clima/policies/package/index_en.htm>.[accessed 28.12.2011].
[3] Beurskens L.W.M., Hekkenberg M., Vethman P., ECN Renewable Energy Projection as
Published in the National Renewable Energy Action Plans of the European Members States
Available at: < http://www.ecn.nl/docs/library/report/2010/e10069.pdf>.[accessed 24.12.2011].
[4] Amon T., Amon B., Kryvoruchko V., Machmuller A., Hopfner-Sixt K., Bodiroza V., Hrbek R.,
Firedel J., Potsch E., Wagentristl H., Schreiner M., Zollotsch W., Methane production through
anaerobic digestion of various energy crops grow in sustainable crop rotations. Bioresource
Technology 2007; 98: 3204-3212.
[5] Murphy JD, Power N. Technical and economical analysis of biogas production in Ireland
utilising three different crop rotations. Applied Energy 2009; 86: 25-36.
[6] Gebrezgabher S.A., Meuwissen M.P.M., Prins B.A.M., Lansink A.G.J.M.O., Economic
analysis of anaerobic digestion A case of Green power biogas plant in The Netherlands. NJAS
Wageningen Journal of Life Sciences 2010; 57: 109-115.
[7] Di Maria F., Pavesi G., Leombruni S., Improvement of an existing anaerobic digestion plant:
technical and economic analysis. ORBIT 2008: Proceeding of the 6th International Conference
on Moving Organic Waste Recycling towards Resource Management and for the Bio-based
Economy; 2008 October 13-15; Wageningen, The Netherlands.
[8] Di Maria F., Sordi A., Micale C., Energy production from mechanical biological treatment and
composting plants exploiting solid anaerobic digestion batch: an Italian case study. Energy
Conversion and Management 2012; 56: 112-120.
[9] Murphy J.D., McKeogh E., Technical, economic and environmental analysis of energy
production from municipal solid waste. Renewable Energy 2004; 28: 1043-1057.
[10] Forster-Carneiro T., Perez M., Romero L.I., Sales D., Dry-thermophilic anaerobic digestion
of organic fraction of the municipal solid waste: Focusing on the inoculum sources. Bioresource
Technology 2007; 98: 3195-3203.
[11] Fdez-Guelfo L.A., Alvarez-Gallego C., Maruqez D.S., Romero Garcia L.I., Start-up of
thermophilic-dry anaerobic digestion of OFMSW using adapted modified SEBAC inoculum.
Bioresource Technology 2010; 101: 9031-9039.
[12] US EPA 3050 B (1996), Environmental Protection Agency.
380
Abstract:
Europe is striving for zero carbon electricity production by 2050 in order to avoid dangerous climate change.
To meet this target a large variety of options is being explored. Biomass is such an option and should be
given serious consideration. In this paper the potential role of biomass in a NW-European electricity mix is
analyzed. The situation in NW-Europe is unique since it is a region which is a fore runner in renewable
technology promotion but also an area with little sun, almost no potential for hydro and a lot of wind. This will
result in a substantial need for non-intermittent low-carbon options such as biomass. The benefits and issues
related to biomass are discussed in detail from both an environmental and an economic perspective. The
former will focus on the life cycle of a biomass pellet supply chain, from the growth of the trees down to the
burning of the pellets on site. The latter will provide detailed insights on the levelized cost of electricity for
biomass and the role of biomass as a grid stabilizer in high intermittent scenarios. During the discussion,
biomass will be compared to other competing electricity technologies to have a full understanding of its
advantages and drawbacks. We find that biomass can play a very important role in the future low carbon
electricity mix, the main bottleneck being the supply of large amounts of sustainably produced feedstock.
Keywords:
Biomass, Renewable Energy, Sustainability, Economics, Learning Rate.
1. Introduction
Biomass as a renewable energy source has been used by humans for many centuries. Today, it is
still a very important source of energy in developing countries. In 2011, biomass had a share of
about 10% in global energy demand, however, only a small amount of biomass is used for
electric ity production. Globally, about 62 GW of biomass power capacity was estimated to be
installed by the end of 2010. This is slightly higher than the installed capacity of PV, but
significantly smaller compared to global wind capacity (1). In the EU-27, biomass has been used for
energy purposes for a long time, mainly by Scandinavian countr ies. Today, Sweden and Finland
together account for 18.4 TWh of electricity production from solid biomass, namely 30% of the
total EU-27 solid biomass electricity production. This is almost entirely due to the use of CHP
plants (17.5 TWh)(2). Similarly, other countries are now looking at biomass technology too, in
order to reach their renewable energy targets imposed by the European commission (3). Especially
countries like Belgium, the Netherlands, Germany and the U. K. are trying to tap into this energy
source for renewable electricity production. Together, these four countries produced 21.1 TWh of
electric ity from solid biomass in 2009, which is about 34% of the total amount of solid biomass
electric ity production in the EU (2). Unlike the Scandinavian countries, NW-European (NW-E)
member states mainly use biomass for electricity-only purposes (15 TWh), in increasingly larger
power plants.
Despite the growing interest by governments to use biomass for electricity production in large scale
power plants, the scientific world has not focused much on this subject. Most literature is targeting
biomass use for transport (4-6). Authors that did focus on the use of biomass for electricity
381
production have often only considered small scale installations, mainly CHP (7-9), which is not so
common in NW-E. Furthermore, most papers limit the analysis to the environmental implications of
biomass use (5, 10, 11), and neglect economic aspects.
This paper aims to assess the potential role of large scale biomass pellet power plants in NW- E
from an environmental and economic perspective. The environmental assessment is based on the
Life Cycle Assessment (LCA) methodology. For the economic analys is, the Levelized Cost of
Electricity (LCOE) methodology is applied to evaluate the cost competitiveness of the technology.
Finally, the potential of biomass electric ity as a backup for variable renewables in the electricity
mix is studied.
Most of the data, such as energy prices and investment costs, on which the assessments in this study
are based, were obtained from a combination of a vast literature review and long discussions with
the industry. In the first part of this paper, the environmental impact of biomass use will be
discussed. The second part will consist of the economic analysis. We will conclude w ith an overall
discussion on the implications of increasing the role of large scale biomass in the future electricity
mix in NW-E. Even though this study is focusing on NW-E, the findings can be interesting for
policymakers, energy companies and investors worldwide.
Figure 1: Renewable capacity; globally, in developing countries, the EU-27 and top 5 regions (1)
2. Environmental analysis
2.1. Introduction
A life cycle assessment (LCA) is considered to be the most comprehensive and credible method to
evaluate the environmental impact of a good or service. In an LCA the whole lifecycle of a
technology is considered. Many scientists have used the LCA methodology to evaluate the
environmental impact of renewable and non-renewable energy technologies (5, 12-17). The
lifecycle of an energy technology typically consists of construction, fuel use, operation and
maintenance (O&M) and decommissioning. Various perspectives such as land use change, water
use, mineral extraction, energy use and human health can be included in an LCA (18). However,
due to increasing concerns about climate change, current research is mainly focusing on greenhouse
gas (GHG) emissions (17, 19-21).
Recently, biomass has received a lot of attention and is widely considered to be an essential part of
the sustainable or green economy (22, 23). This growing interest is unfortunately matched by a
rise in criticism. The impact of the development of a bio-based economy on the environment in
general, and on land use and food prices in particular is causing great concern (24). This resulted in
an abundance of studies that have primarily shown that the afore-mentioned issues are complex and
the sustainability of biomass strongly depends on specific circumstances (5, 14, 18).
The debate on the merits and problems surrounding biomass benefits from improved know ledge on
the life cycle of biomass in the green economy. LCAs can provide important ins ights in the
382
environmental impact of a biomass energy chain and help locate the most critical steps in that chain.
This information should help developers and policymakers to increase the sustainability of biomass
production and its use, and to possibly overcome some of the current issues. This study will focus
specifically on the life cycle of large scale (> 300MW) biomass power plants in NW-Europe using
Canadian wood pellets.
2.2. Methods
The environmental impact of biomass usage for energy purposes has been studied extens ively.
Unfortunately, assumptions regarding type of biomass, land use change, fertilizer usage, pesticide
usage, transport and final combustion technology vary immensely (18). It is thus difficult to find
studies that focus specifically on the life cycle of large scale combustion plants using pellets. In
fact, only one such paper was found in the literature (25). The main steps of the biomass life cycle
are shown in Figure 1. Each of these steps will be described in more detail below (see next
paragraph The NWE Case).
Figure 2: Boundaries of the large scale biomass lifecycle (based on Zhang et al. 2010)
An assumption which is crucial for this LCA is that the wood from Canadian forests is sustainably
grown which allows us to state that the emission of GHGs at the plant site are balanced by the
uptake of GHGs in the growth phase of the biomass. In other words, the net GHG emissions are
considered to be equal to zero. We chose not to consider the possible effects of biomass production
on land use or on carbon stocks, even though these two aspects could have a very big influence (14,
18, 26, 27). It is therefore important to note that if the sustainability assumption does not hold, the
results presented below will not be valid. If not sustainably produced, the use of biomass is neither
environmentally nor socially desirable.
g/(ton* mile) respectively (21, 28). Finally, in phase four, NOx emissions from the NW-E power
plant were estimated to be equal to the European limit for new large scale biomass power plants
(>300 MW), namely150 mg/Nm at 6% O2 , or 0.55 g/kWh (29). The GHG emissions on site are
considered to be compensated by the growth of new biomass, and thus they were assumed to be
zero.
Figure 3: Estimated GHG and NOx emissions for a NW-E biomass plant fired with Canadian
pellets (calculations based on (21, 25, 28, 29))
be subjected to the EU norm of 150 mg/Nm at 6% O2, gas plants will even have lower norms (100
mg/Nm at 6% O2 ).
2.3.2.2 Life cycle emissions
For gas and coal, lifecycle GHG and NOx emissions are similar to combustion emissions 1 (Table
1). In the case of biomass this is quite different; life cycle emissions for biomass are markedly
higher than combustion emissions. Nevertheless, over the whole lifecycle, biomass power plants
emit less GHG compared to coal and gas. For life cycle NOx emissions, biomass is similar to gas,
but much lower than coal.
The last row of Table 1 shows the estimated results for a future coal powered plant (2025) with
CCS technology. According to the literature (21), current biomass technology emits less GHGs than
possible future CCS technology, while NOx emissions are similar. However, by 2025, better NOx
removal technology could be in place for biomass too, which is likely to result in NOx emissions
lower than coal + CCS.
Compared to PV technology, biomass emits similar amounts of GHGs over its lifecycle, although
the NOx emissions of PV-systems are significantly lower. Overall, wind onshore and offshore emit
the lowest amount of polluting substances during the entire life cycle of all the energy technologies
considered here.
Table 1: Combustion and LCA emissions for GHG and NOx of various energy technologies (5, 10,
17, 18, 21, 25, 30-37)
Combustion emissions
GHG (g/kWh)
min
average max
0
318
454
636
780
953
1044
NOx (mg/kWh)
min
average
max
554
54
1208
2361
1162
2642
4122
GHG (g/kWh)
min
average max
2
66
122
92
103
110
116
50
99
160
4
17
40
9
13
17
360
466
720
800
1024
1800
130
190
280
NOx (mg/kWh)
min
average
max
781
923
1064
1000
1473
1587
1702
340
31
21
77
1782
4268
1285
2842
4399
863
1251
1639
For LNG gas this is different, since a lot of energy is needed for compression/transport/decompression
385
plants (<70 MW), which typically use relatively inefficient flue gas filtering technologies. These
emissions can be reduced drastically by using the most advanced equipment.
The upper limit of dust combustion emissions for large scale biomass plants in Europe can be
calculated starting from the legal PM-emission limit which is 20 mg dust/Nm at 6% of O2 , or about
67 mg/kWh (29). This is considerably lower than the emissions mentioned in the Greenpeace
report. Furthermore, since 20 mg dust/Nm at 6% of O2 is the legal limit in the EU, we can safely
state that in practice emissions are likely to be lower than 67 mg/kWh. However, over the whole
lifecycle the dust emissions will be higher, especially due to transport by truck and ship.
2.4. Conclusions
In general, considering GHG and NOx emissions, biomass is more environmentally friendly than
coal or gas. However, this statement is only valid if the biomass is produced sustainably. If this is
not the case, GHG and NOx emissions can be much higher than presented here. When compared to
other renewables, biomass appears to have a relatively high impact on the environment. The life
cycle GHG emissions of biomass are comparable to those of PV systems but about a tenfold higher
than the amount of GHGs emitted when using wind energy. The results for NOx are even worse,
with biomass releasing roughly five times as much as PV and 50 times more than wind. However,
some issues should be kept in mind. Firstly, the environmental impact of biomass could decrease if
power plants became more efficient or transport was organized in a more sustainable way.
Secondly, comparing biomass to intermittent renewables is not really fair, since the latter will not
be able to achieve high penetration as long as cheap storage technology is unavailable. Biomass
LCAs should be considered from a holistic, system wide perspective. Unfortunately, this is far from
easy since the specific properties of the electricity infrastructure need to be taken into account.
Finally, it should be stressed that the scope of this study is rather limited. Other important aspects of
the environmental impact of the power technologies would also be interesting to compare. Data on
life cycle particulate matter emissions, fossil fuel depletion or energy efficiency would result in a
broader understanding of the environmental impact of biomass electricity production.
Unfortunately, data on these issues is currently not available for large scale biomass power plants.
Further research is necessary to have a better understanding of all the steps in the whole lifecycle
and how these affect the environment.
3. Economic analysis
When comparing energy technologies many criteria need to be considered. For example, the type of
demand for which new capacity is needed base, intermediate or peaking load will determine the
most economically effective technology to supply electricity. Electricity might be a standardized
commodity, behind it lays a segmented supply side, with each segment functioning on different
economic parameters. Due to the expected growth of intermittent generation, the boundaries
between the different demand types will become less explic it in the coming years and many assets
will have multiple load purposes (even during one single day).
production cost per MWh of existing assets should not be compared to LCOE results, as the
investment cost per MWh is not included.
However, the LCOE methodology has some limitations; as it does not evaluate the aspect of risk,
which is very important when taking investment decisions, in addition, it looks at the different
technologies from a stand alone perspective. The LCOE is calculated at the plant level and
excludes system costs and system externalities. The latter factor is a major issue for variable (nondispatchable) renewable energy technologies (38) because demand and supply need to be balanced
literally every second. Basically, direct system costs should be added to the plant- level LCOE of all
technologies but this proves to be very challenging. In order to overcome this issue, a broader,
system-wide, economic assessment will complement the LCOE analysis (see section 3.2).
Finally, the LCOE should be interpreted as a social cost or the cost for society of building and
operating the generation plants. The financial impact of taxes, subsidies, portfolio standards or other
incentive schemes is therefore not considered. In this section, eight technologies will be compared:
100% biomass using pellets, biomass co-firing (50% co-firing, see Appendix A), PV, wind onshore,
wind offshore, coal, nuclear and gas.
3.1.2. Methodology
3.1.2.1. LCOE calculation
The LCOE takes into account the annualized investment cost, the operation and maintenance
(O&M) cost, fuel cost and carbon cost:
LCOE I
With
I
O&M
F
C
With
Itot
F C
(1)
And
O&M
(2)
= annuity factor
FLEOH
= full load equivalent operating hours [h]
The annuity factor for a given lifetime and discount rate has been calculated as follows 2 :
(3)
With
n
= lifetime
i
= discount rate
The assumptions regarding fuel cost, lifetime and O&M costs are based on data found in the
literature and discussions with the industry. They can be found in Appendix A.
Essential for estimating the full load equivalent operating hours (FLEOH) is the load factor (LF).
This indicates the percentage of time that a technology produces electricity at maximal capacity. It
shows how much electricity can be produced with an installation of a certain capacity in one year.
Annuity factor is commonly used to calculate the present value of future series of cash flows (Richard et al,
Quantitative Investment Analysis, 2007).
387
FLEOH
24h 365d
= load factor [%]
= full load equivalent operating hours [h]
LF
With
LF
FLEOH
(4)
Unlike fossil fuelled technologies, renewable intermittent technologies such as wind and PV have
the disadvantage of only produc ing electricity when the weather is favorable. Luckily, the NW-E
region is blessed with respect to wind, having relatively high average annual wind speeds compared
to the rest of Europe. Onshore wind turbines operate at full load for about 2200 hours annually (39),
which is equal to a load factor of about 25% (2200/24/365 = 25%), offshore wind turbines have
higher load factor (35%). For PV, on the other hand, NW-E is not a favorable region. In fact, the LF
for PV decreases with declining yearly average solar irradiation. A PV system in London, for
example, produces roughly half the amount of electricity compared to a PV system in Malta, for a
given capacity (30). Therefore, the LF for a PV system in NW-E is very low (12%). By contrast, the
load factor of fossil fuelled power plants is much higher. In this study, the 2012 LFs were estimated
to be 85%, 75% and 60% respectively for nuclear, coal and gas.
It
I 0 CC a
(5)
With
It
= investment cost at time t [EUR/MW]
I0
= cost of the first unit produced (theoretically) [EUR/MW]
CC
= cumulative capacity [MW]
a
= learning elastic ity
From this the learning rate can be calculated as follows:
PR
With
LR
PR
2 a and 1 LR
PR thus LR
1 2a
= learning rate
= progress ratio
388
(6)
This formula indicates how the cost will decrease with each doubling of production. For example, if
the LR for a given technology is 20% and the installed capacity is 1 GW, the costs are assumed to
be 20% lower when capacity has reached 2 GW. The LRs for the various technologies were found
in the literature. More information is available in Appendix A.
Although a broad literature on technology learning exists, there is no consensus on the typical and
prospective learning rates for the considered generation technologies. However, the learning rates
presented here are in line with the literature (49). To address the uncertainty in estimating learning
rates, a sensitivity analysis was carried out to illustrate the impact of changes in learning rates.
3.1.2.3. Discount rate
The discount rate reflects the return on capital for an investor in the absence of specific market or
technology risks. In the energy sector, relatively high discount rates can be expected due to the
uncertain market environment of today with the ongoing liberalization, multiple CO2 policies,
subsidies for new technologies (such as offshore wind and CCS) and the challenge of integrating a
growing share of intermittent generation. Furthermore, LCOE methodology assumes a single set of
future fuel prices but mostly neglects the impact of higher fuel prices and investment costs (e.g.
steel and concrete prices should follow fuel prices). In our calculations, a 10% discount rate was
used. This is complemented by a sensitivity analysis with discount rates of 7% and 13%.
3.1.2.4. Fuel and CO2 prices
The cost evolution of pellets (in bulk) between now and 2030 is probably the most difficult
parameter to project. In order to cope with this problem, three pellet price scenarios were evaluated.
Next to a standard scenario, w ith the price of biomass pellets ris ing at a rate of 1% each year, two
other scenarios were added: an optimistic and a pessimistic scenario. In the optimistic scenario, the
supply is assumed to be able to keep up with demand and due to improved logistics and better
technologies, this would result in a stable pellet price (73 /MWh), from now to 2030. In the
pessimistic case, the supply will have a hard time to keep up with a very sharp increase in pellet
demand, resulting in a doubling of the pellet price by 2030. Not surprisingly, the pellet price will
have a major impact on the LCOE of a biomass produced MWh of electricity.
Regarding fossil-based technologies, the price of gas and coal is likely to rise with time. Also the
CO2 price is likely to increase between now and 2030. The estimated gas and coal fuel costs were
calculated based on the averages of the recent price trends. Biomass pellets costs for 2012 were
estimated based on recent literature. However, all the fuels cost estimates have been adapted after
discussion with the energy industry. Information on these assumptions can be found in Appendix B.
389
with coal, gas and biomass technologies. Keep in mind that we have assumed low irradiation and
very conservative learning rates.
In order to provide a better picture of the LCOE for the other technologies, next to Figure 5, we
present Figure 6, which is the same figure with the exclusion of PV.
Figure 6: LCOE of new generation assets (PV excluded) standard pellet price scenario
Since nuclear does not emit CO2 and uranium prices were assumed to be constant, the cost of
nuclear does not evolve in the selected time span, making it one of the cheapest technologies from
2017 onwards. However, from 2022 onwards, the LCOE of w ind onshore is lower than the LCOE
of nuclear, thus becoming the cheapest technology. Despite having higher investment and
maintenance costs, offshore wind benefits from higher load factors than onshore wind (39), also, in
this analysis, offshore benefits from the relatively high growth in installed capacity between 2012
and 2030, therefore, due to the bigger learning effect its cost will decrease more than onshore wind.
390
On the other hand, the gap with nuclear and onshore wind remains substantial. The LCOE of cofiring lies above that of biomass from 2023 onwards, but below gas, PV and coal. The latter
becoming the most expensive technology from 2030 onwards in this scenario.
3.1.3.3. Discussion
The three scenarios with stable load factors 80% for both biomass technologies confirm that the
economic attractiveness of biomass technologies is highly sensitive to the expectations about the
pellet price evolution. In the pessimistic pellet pr ice scenario, the LCOE of biomass technologies is
significantly higher than the LCOE of PV, coal and gas. Investing in biomass is moderately
attractive in the standard pellet price scenario and even very attractive in the optimistic pellet price
scenario. If pellet prices were to evolve as assumed in the standard or the pessimistic scenario, then
biomass will roughly be as expens ive as fossil-based technologies in the long run. From an
investment perspective, policy measures to ensure a sufficient supply of biomass are of utmost
importance to trigger significant investments in new biomass generation capacity. However, it
should be pointed out that finding this cheap supply considering the sustainability issues
mentioned in the first chapter will be far from easy. Policymakers should be cautious in
promoting biomass overnight, without a clear sustainability framework.
Cofiring Cofiring
S
O
134
126
137
129
130
136
98
120
In general, the reduced load factors do not significantly impact the gap between biomass
technologies and coal and gas. Only nuc lear is subjected to a high impact under the conditions of
the reduced load factor scenario. In fact, the LCOE of nuc lear increases from 98 to 120 /MWh,
therefore the gap with the other non-intermittent technologies is reduced slightly. Nevertheless, it
remains the most attractive non-intermittent technology.
The lower load factors obviously increase the 2030 LCOE-gap between wind and all the nonintermittent technologies. With the reduced load factor, the high pellet price scenario leads to a
100% biomass LCOE that is about double the LCOE of wind in 2030. When we compare the latter
pellet price scenario to the LCOE of coal, we find a worst case cost-disadvantage of biomass of
some 32% ( 193 vs. 146). On the other hand, under the optimistic pellet pr ice scenario with the
reduced load factors, the LCOE of 100% biomass is only 20 above the LCOE of wind onshore.
With standard pellet prices, the 2030 cost gap of 100% biomass technologies with wind onshore
increases from 35%, under the full load scenario, to 39%, under the reduced load scenario. In short,
the reduced load factor scenario illustrates that the 2030 LCOE differences between biomass-based
technologies and the other non-intermittent assets remain roughly the same, while wind onshore
392
(the most competitive technologies) slightly increases its competitiveness against all the nonintermittent technologies.
3.1.4.2 Learning Rate
The presented results depend on many assumptions and have intrins ic limitations. As we assumed
that future investment cost reductions depend on learning efforts, higher or lower learning rates are
likely to influence our findings. Figure 9 illustrates the learning rate sensitivity of our results in the
standard pellet pr ice scenario with stable load factors. This figure projects the LCOE with the
assumed learning rate (see Table 1) together with the alternative LCOE when we increased and
reduced this learning rate by 2%.
Figure 9 shows that the variation in the learning rates mainly influences the LCOE of PV, offshore
and onshore wind and biomass co-firing. Not surprisingly, the technologies with the biggest growth
potential are more influenced by a change in the learning rate. For biomass co-firing, the difference
in LCOE with the extreme levels of the learning rates is limited to some 6/MWh. For PV
technologies, a 2% change in the learning rate can result in a LCOE variation of 15/MWh.
Nuclear is not depicted in Figure 9 since it was assumed that, due to increasing safety measures,
costs for nuclear plants would not decrease in the future. The learning effect is thus, according to us,
not applicable to nuclear.
Figure 9: Learning rate sensitivity of 2030 LCOE (standard pellet price scenario)
3.1.4.3. Discount Rate
Another important parameter for the calculation of the LCOE is the discount rate. We used a 10%
discount rate and present in Figure 10 alternative results with discount rates of 7 and 13%. As
expected, the variation in the discount rate significantly impacts the results.
Figure 10: Discount rate sensitivity of 2030 LCOE (standard pellet price scenario)
393
A 3% variation in the discount rate markedly changes the LCOE-ranking in the standard scenario.
With a 7% discount rate, nuclear technologies offer the lowest LCOE in 2030, followed by onshore
wind, offshore wind, PV, 100% biomass and gas, biomass co-firing and coal. Lower discount rates
increase the LCOE-gap between wind onshore and biomass technologies. A discount factor of 13%
leads to a scenario where the gap between nuclear and wind technologies and biomass technologies
is significantly reduced if compared both with the 7 and 10% discount rate scenarios. Higher
discount factors reduce the LCOE-gap between wind and nuclear and biomass technologies.
3.1.5 Conclusion
From a LCOE-perspective, including a pragmatic carbon cost and a discount rate of 10%, we can
conclude that today gas offers the least expens ive generation opportunity with full production costs
of around 86 per MWh. The LCOE of biomass technologies 100% biomass and 50% co-firing
is respectively some 48 and 38% higher than the LCOE of gas, which represents the 2012
benchmark. Without a CO2 cost, biomass technologies are some 68% (100%) and 56% (50% cofire) more expensive than gas. Onshore wind today (2012) has a LCOE slightly below our estimates
for biomass technologies.
To explore the opportunity of biomass technologies from a stand-alone perspective excluding all
external costs for the electricity system at large we compared three pellet price scenarios and two
load factor scenarios for the period 2012-2030. Table 3 summarizes our findings. We compare both
(100% and 50% co-fire) biomass technologies together by averaging their LCOE in 2030 to the
LCOE of coal, gas and wind technologies in 2030. From Table 3 we can conclude that biomass
technologies offer attractive investment opportunities from an LCOE-perspective in 2030. Only
with high pellet pr ices, the LCOE of biomass technologies is 27-35% higher than the LCOE of coal
and gas technologies. Compared to wind technologies, biomass faces a LCOE-disadvantage of 70 to
74% in 2030 under the pessimistic pellet pr ice scenario.
With the standard pellet price scenario, biomass technologies are less expensive in 2030 than coal
and as expens ive as gas, and the LCOE-disadvantage to wind technologies is between 29 and 32%.
With low pellet pr ices, biomass technologies have a significant LCOE-advantage over coal and gas
technologies between 9 and 15% while wind technologies still offer a better investment
opportunity.
Table 3: Competitiveness of biomass considering six scenarios in 2030
Pessimistic
pellet price
Standard
pellet price
Optimistic
pellet price
1.27
1.23
0.96
0.93
0.88
0.85
1.35
1.32
1.02
1.00
0.93
0.91
1.70
1.74
1.29
1.32
1.17
1.21
lower system costs and thus generate system benefits. Also, it needs to be acknowledged that the
impact of a given technology on system dynamics varies over time.
Gas/Coal
Nuclear
Wind
Solar
Biomass
Hydro
Yes
Yes
No
No
Yes
Yes
Yes
Yes
Yes
Yes
No,
not flexible
X
X
No
No
No,
subsidized
No
No
No,
subsidized
Yes
Yes
No,
subsidized
Yes
Variable
Yes
X
X
Yes
X
X
The above discussion indicates that biomass technologies can provide important system benefits.
Only biomass and hydro plants offer the potential for renewable base-load and mid-merit/
intermediate generation. As it is very unlikely to develop a 100% renewable electricity system
without renewable base-load and mid-merit production, biomass and hydro should play a pivotal
role in energy transition scenarios. In the next decades, flexible biomass and hydro plants can also
play a role in balanc ing the production from the intermittent renewable technologies.
The debate on the adaptation of the system to accommodate a growing intermittent production is far
from settled. A more flexible system requires significant investments in transmission, distribution
and smart grids to facilitate balancing and to accommodate changes in supply and demand. The
need to foresee back-up and the long-term impact on loading factors and shedding should equally
be considered given their potential impact on investment decis ions. As biomass offers the potential
395
for renewable generation without the typical system challenges of intermittent renewables,
estimates of the system implications of additional intermittent generation can provide indications of
the economic value of biomass in high intermittency scenarios. In the next sections, we elaborate
the system benefits of biomass starting from high intermittency perspectives on future generation.
Firm availability
90%
10%
70%
Back-up need
10 MW
90 MW
30 MW
In Table 6, we present an example of how a future generation mix could look like in a NWEuropean region without hydro. Such a region could have a generation portfolio of 15,000 MW, of
which 7000 MW is intermittent capacity (5000 MW wind + 2000 MW solar PV). In addition, there
is 3000 MW biomass together with 2000 MW nuclear. The fossil capac ity is limited to 3000 MW.
The required back-up capacity depends on all the assets in the generation mix. Starting from baseload assets, the back-up need for 2000 MW nuclear capacity is around 100 MW. Adding 2000 MW
gas will increase the total back-up needs to 250 MW, which is below 300 MW because the
correlation of the non-availabilities of nuclear and gas assets is rather low. Adding 1000 MW coal
further increases the back-up needs to 300 MW. The 3000 MW biomass capacity requires a back-up
of 900 MW but 300 MW back-up is already established (in response to investments in nuc lear, gas
and coal capacity). As the correlation between the non-availabilities of biomass and the other assets
is low, we do not need to increase the total back-up pool by 900 MW. We assume that an additional
back-up investment of 700 MW is sufficient. With respect to the 5000 MW wind capacity, a backup need of 4500 MW emerges from a stand-alone perspective. As already 1000 MW back-up assets
are provided, total back-up provision w ill not increase by 4500 MW but by e.g. 3700 MW. Finally,
the back-up needs for 2000 MW solar PV are 1900 MW from a stand-alone perspective but as
already 4700 MW back-up is foreseen, a modest increase of the total back-up can be sufficient. In
396
the example of Table 6 we end up with a total back-up need of 5000 MW or 33% of total installed
capacity. With only nuclear, gas and coal capacity the back-up requirement would be only 6% of
total installed capacity.
Table 6: Back-up needs for a portfolio of 15,000 MW with 7000 MW intermittent capacity
Frontline generation Firm
Back-up
Cumulative
% of
Increase of
portfolio
availability needed per
Required back-up installed back-up need
(15000 MW
(capacity
asset
capacity due to RES
(nuclear PV)
capacity)
credit)
2000 MW nuclear
95%
100 MW
100 MW
5%
2000 MW gas
90%
200 MW
250 MW
6.25%
1000 MW coal
90%
100 MW
300 MW
6%
3000 MW biomass
70%
900 MW
1000 MW
12.5%
+ 700 MW
5000 MW wind
10%
4500 MW
4700 MW
36%
+ 3700 MW
2000 MW solar PV
5%
1900 MW
5000 MW
33%
+ 300 MW
Total: 15000 MW
7700 MW
5000 MW
+ 4700 MW
Alternative portfolio with only changes in RES capacity (nuclear, gas and coal unchanged)
6000 MW biomass
70%
1800 MW
1900 MW
17%
+ 1600 MW
3000 MW wind
10%
2700 MW
2900 MW
20%
+ 1000 MW
1000 MW solar PV
5%
950 MW
3100 MW
20%
+ 200 MW
Total: 15000 MW
5850 MW
3100 MW
+ 2800 MW
In an alternative portfolio in Table 6 (Alternative portfolio with only changes in RES capacity) we
increase total biomass capacity up to 6000 MW (+ 3000 MW) and lower the intermittent assets by
3000 MW. In this second example, the maximal generation on a given moment is identical to the
maximal generation with the upper panel of Table 6 (under the assumption of strong wind and a
high solar irradiation). With the alternative portfolio, total back-up needs are lower and the increase
of back-up due to renewables is also much lower than in the upper panel. With 6000 MW of
biomass, total back-up needs are only 20% of total installed capacity. 3000 MW additional
biomass capacity lowers total back-up needs by 1900 MW (while replacing 3000 MW intermittent
capacity). In this example, trading 1 MW wind capacity for 1 MW biomass capacity lowers back-up
needs by 0.6 MW per additional MW biomass.
From an environmental perspective, the back-up pool in the example of Table 6 should consist of
very efficient assets. If not, the environmental benefits of a low-carbon generation system risk to be
lowered by frequent use of ineffic ient high-carbon back-up plants. In principle, part of the back-up
challenge could be met by integrating the regional market of Table 6 into a larger European market.
In case of sufficient transmission and interconnection capacity, some of the needed back-up
generation can be provided for by the excess production of wind electricity in neighboring
countries. The ability to import electricity is sometimes presented as an alternative to the local
provision of back-up capacity. The most comprehensive historical weather models however
conclude that total wind and solar output in NW-E will be highly correlated and will not average
out over regions (51). In fact, when there is no wind in Belgium, there is probably no massive
production of wind electricity in the Netherlands, Germany or France either.
As local back-up is essential to guarantee generation capacity, we can observe a massive expansion
of total generation capacity in all scenarios with a high penetration of intermittent renewables in the
next decades. As mentioned before, investments in additional wind capacity do not lead to
equivalent reductions of fossil or nuclear capacity. When the UK would like to increase the share of
intermittent renewables up to 50% by 2030, total installed capacity would have to increase from 80
GW today to 125 GW in 2030 (or increase by 56%). To further strongly increase the share of
intermittent generation between 2030 and 2050, total generation capacity in the UK has to increase
397
from 125 GW in 2030 to 230 GW in 2050. The latter increases by 85% between 2030 and 2050 will
lead to a rise in electricity production of some 33% (52).
%
RES
4%
51%
60%
64%
80%
+90%
LF
offshore
LF
biomass
LF
nuclear
LF
CCGT
LF
peakers
40%
40%
40%
40%
40%
36%
65%
51%
42%
50%
40%
30%
88%
88%
75%
81%
62%
0%
50%
19%
12%
18%
12%
19%
5%
0%
5%
0%
3%
9%
Shedding
(% of
demand)
1%
1%
7%
7%
20%
Shedding
(TWh)
6
6
38
41
120
Massive investments in intermittent renewables can thus produce high external costs in terms of
lower load factors, lower investment opportunities in other generation assets and significant
shedding of low-carbon generation. These costs should be interpreted as uncompensated external
costs: not the investors in intermittent generation but investors in other assests and society at large
will have to bear these costs. The investors in intermittent generation will not compensate the
owners of other generation assets for the reductions in load factors. Only in the extreme Max
scenario with very significant shedding of offshore wind capacity, investors in intermittent
generation will bear themselves part of load factor and shedding losses.
As a reduction of the load factor increases the LCOE of a technology, final consumers will face
higher prices because of the increasing intermittency of the electricity system. According to models
by Mott MacDonald, the levelized cost of nuclear and coal CCS would triple when the load factors
would fall from 70% to 30%. For flexible gas plants, the levelized cost would only increase strongly
once load factors are below 15% (53).
398
Investing in biomass plants can lead to lower shares of intermittent renewable and hence lower
external load factor and shedding costs from high penetration rates of intermittent renewables.
However, investors in biomass capacity will not be rewarded for their contribution to lower external
costs while investors in intermittent generation are not held responsible for these external costs.
Cons
(TWh)
Cap
(GW)
% ren
High 2030
High 2050
Very High 2030
Very High 2050
M ax
409
551
409
551
611
125
171
158
230
298
51%
60%
64%
80%
+95%
Wind
on+off
59
102
68
119
191
Solar
3
3
25
38
38
Peaker
0
6
1
10
21
AFC
(bn/a)
5.4
5.9
8.5
10.0
16.6
Pyry stresses that the high intermittency scenarios have not been selected from a cost-effectiveness
perspective, as the main goal of the analys is was to find out whether high intermittency can be
technically accommodated. It is thus possible to lower the high flexibility cost estimates in the left
column of Table 8 by replacing wind and solar capacity with hydro and biomass capacity. The
potential to increase hydro capacity is however limited in the UK. A strong increase of biomass is
3
To include the cost of all assets to move electricity demand would imply that the upfront investment cost of electric
vehicles is part of the flexibility cost.
399
not integrated in the Pyry scenarios as the goal was to assess high intermittency. Furthermore, the
UK currently adopts a holding position with respect to bio-energy at large. No significant increase
in bio-energy use is assumed in official documents such as Renewable Energy Review given
concerns over sustainability and questions over the best long-term use for this limited resource
(executive summary, p16, (54)).
In an effort to lower the flexibility cost of the system, we have to consider the replacement potential
of biomass. With a load factor for biomass of 60% in 2030, 8 GW of additional biomass capacity
can replace 16 GW wind capac ity (with an average load factor of 30% for onshore together with
offshore) or 9 GW wind and 21 GW solar (LF 10%). With a load factor of 75% for biomass plants
in 2030, much more intermittent capacity can be replaced. The difference between High 2030 and
Very High 2030 also includes more tidal energy capacity (+4 GW), more CCGT capacity (+5 GW)
and more peaking capacity (+1 GW). As especially the CCGT and peaking capacity is related to the
increasing intermittency between High 2030 and Very High 2030, 8 GW additional biomass
capacity can partly replace the additional CCGT and peaking needs in Very High 2030.
Although the comparison is simplified and not complete, we can conclude that the investment in 8
GW biomass capacities by 2030 can prevent most of the projected increase of annualized flexibility
cost in the shift from the High 2030 scenario to the Very High 2030 scenario. Based on the Pyry
assessment for the UK (52), investing in 8 GW of biomass capacity avoids an increase of system
flexibility costs close to 3 billion per year. By 2050, more biomass capacity can be developed but
the (much) lower load factors will make it difficult to replace much more wind and solar capacity.
Summarized, in this framework with an electricity system of 400 TWh we have to distinguish two
pathways to a high share of renewable generation (50 to 60% share of renewable in generation): the
massive deployment of intermittent generation will lead to high system flexibility costs while the
alternative with a lower but still very important deployment of intermittent renewables is
complemented by investments in additional biomass capacity. We estimate that the annual
flexibility cost from mainly intermittent renewables can be reduced by roughly one third ( 3.1 bn /
8.5 bn) when intermittent capacity is lowered by 15% in response to additional biomass
investments.
It is important to realize that these findings are mainly indicative and based on the rather radical
deployment scenarios for the UK (52). These scenarios should however not be interpreted as unique
island scenarios since increased interconnection with Ireland, NW-E and Norway is included, as
well as powerful active demand management systems that move up to 30% of total demand and
bulk storage possibilities.
3.3. Conclusions
In this electricity system overview we focused on back-up needs, load factors, shedding, costefficient RES targets and the general system flexibility costs. We can identify that biomass capacity
offers several important benefits. Most of them will however only become vis ible in the next
decades assuming that the share of intermittent renewables will indeed strongly increase
although there are also benefits to be experienced as of today;
1. In high intermittency scenarios, biomass capacity can significantly lower total back-up needs; in
our example, trading 1 MW w ind capacity for 1 MW biomass capacity lowers back-up needs by
0.6 MW per additional MW biomass
2. In high intermittency scenarios, biomass capacity can limit the projected reduction of load
factors
3. In high intermittency scenarios, biomass capacity can avoid massive shedding of low-carbon
generation (up to 30% of demand in extreme scenarios)
4. In high intermittency scenarios, massive deployment of biomass capacity can lower system
flexibility costs by 30%
400
From a societal perspective, additional biomass capacity lowers the system investment needs. Also,
biomass capacity limits the expected price increases from a system that becomes more capitalintens ive but has lower load factors and requires increasing shedding of effic ient low-carbon
generation. For the final consumer, the electricity bill will increase with every additional euro
invested in new assets. Although we can only estimate the system benefits of additional biomass
capacity, it is obvious that all these pos itive externalities from biomass use are currently not
considered in our policy frameworks. In the market configuration of today, there are no incentives
to consider the external cost and benefits of generation technologies. The debate on externalities is
much wider than the conventional focus on negative externalities such as pollution and CO2 costs.
Renewable energy frameworks are dominated by f lat production incentives for all generation
technologies irrespective of their system consequences. In optimal incentive frameworks targeting
high shares of renewable energy sources (RES), assets with the potential to significantly limit
system costs should be favored over assets that not only generate renewable electricity but high
system costs as well.
From a public policy perspective, the existence of positive system externalities typically leads to
underinvestments in the assets producing these externalities. To correct for negative externalities,
the underlying activity should be supported, leading to incentive schemes that internalize the
external benefits for investors. In the context of support for biomass assets, an optimal support
framework should internalize the system benefits of biomass to tr igger additional investments in
biomass capacity.
4. Conclusion
Biomass electricity production has the potential to become a very important piece in the energy
puzzle of tomorrow. It comes with many interesting benefits which are currently underestimated
and thus unrewarded. Some of these are already vis ible; others will become prominent in coming
decades. Today, biomass can be used as a low-carbon source for electricity, and can help member
states in the EU to reach their 20/20/20 targets efficiently. In the world of tomorrow, biomass power
plants can become a crucial part of the electric ity mix, as a grid stabilizer and a renewable source of
back-up power supply. This is especially true for NW-Europe, since this region does not have
sufficient hydro capacity to balance weather dependent, intermittent renewables, such as solar and
wind.
Society as a whole can strongly benefit from the use of biomass for electricity production; however,
in order to reach this goal, some conditions need to be met. The primary condition is that sufficient
supplies of sustainably produced biomass need to be available. Failing to meet this condition will
result in either high, non-competitive electricity production costs or in producing electricity with a
high carbon footpr int, or, in the worst case, both. This can be avoided by investing in both
sustainable forest management to ensure the sustainability and reliable biomass supply chains
to avoid shortages. Since the likelihood of reaching the 2C target has only decreased in recent
years, these investments should start as soon as possible.
Acknowledgments
The authors would like to thank GDF Suez, Electrabel and Laborelec for their cooperation and
support in the making of this paper.
401
Appendix A
In Table A. 1, we present an overview of the LCOE for new investments in different generation
technologies in 2011/2012. The cost figures in this table should be interpreted as averages for the
period 2011-2012. We opted for average cost figures to avoid that our results strongly depend on
temporary price movements. Table A. 1 is based on the international literature while the load
factors are set in line with the NW European context. We added a CO2 cost of 30 US$ per ton (or
23/tCO2 ) to reflect the cost of climate policy measures for electricity producers. This CO2 cost is
above the ETS price of January 2011, but is an adequate illustration of the resource cost of climate
measures. The total cost per technology in /MWh should be interpreted as the average cost (in
present value) per MWh for an investor who invests today in a particular technology and uses a
discount factor of 10%.
In Table A. 1, we included two biomass technologies. The first 100% Biomass is a dedicated
biomass plant of significant scale (>300 MW) which only burns wood pellets. As biomass is mostly
co-fired in coal plants with co-firing rates between 5 and 10%, we also wanted to include co-firing
in our overview. Today, an investor will however not build a new coal plant with the purpose of 5%
or 10% biomass co-firing. A typical co-firing plant therefore does not fit in the LCOE methodology
for new projects. To accommodate the co-firing technology to the LCOE philosophy, we assume
that co-firing technologies evolve in way that 50% co-firing of biomass becomes possible in the
next years. As this 50% co-firing does not yet exist, we refer to it as Cofire(sim) in our simulations.
We want to emphasize that the latter plant should not be confused to the co-firing practice of today
in most countries. In our approach, we estimate the investment cost of a 50% co-firing plant to be
around 2300 per kW and use this value as a starting pos ition in our analysis. From an investment
perspective, flexible coal plants with the ability of 50% biomass co-firing offer the benefit of
flexible generation in response to the relative prices of coal and pellets. The insurance against
feedstock price spikes can compensate the higher investment cost for this type of plant.
Once the CO2 cost is included, Table A. 1 shows that coal and gas offer the least expensive
generation opportunities with production cost of some 100 per MWh. The LCOE with biomass
technologies is some 35% above the LCOE w ith coal and gas technologies. Without a CO2 cost,
biomass technologies are some 60% more expens ive than coal and gas technologies. Onshore wind
technologies have a LCOE that is close to the estimates for biomass technologies. Offshore wind is
some 20% more expensive than onshore wind while the LCOE of PV is still prohibitive. The LCOE
of nuclear technologies is between the LCOE of biomass and the LCOA of gas and coal
technologies. We opted for a high investment cost for nuclear capacity in response to postFukushima concerns and cost overruns for new nuclear in France and Finland 4 .
http://www.guardian.co.uk/environment/damian-carrington-blog/2011/jul/22/nuclear-power-cost-delay-edf and
http://www.nytimes.com/2009/05/29/business/energy-environment/29nuke.html?pagewanted=al
402
World
Lifespan
Capacity
Load
Learning
Factor
Rate
Investment. Cost
O&M
Feedstock Cost
Carbon
Total Costs
Cost
GW
Years
/kW
/MWh
/MWh
/MWh
/MWh
PV
61a,b
25o
12f,e, k
15h,i
2600b
10f
264a
Wind onshore
286c,g
25e
25f,e, k
8h,i
1800j
18j
107a
Wind offshore
5c
20e
35f,e
8-10h,i
3300f,j
30j,f
152a
100% Biomass
64a,c
25f
80f
7a,i
2100f,m
15f
73l,a
122a
Co-firing (sim)
3a
30f
80f
8a,f
2300f
15f
54a
11.5e,a
115a
Coal
1513c,d
35f
80f, k
7h
1700d,e,f
7f,e,a
30a
23e
90a
Gas
1308c,d
30e
70f, k
10h,i
900 d,e,f
5f,e,a
50a
10e
86a
Nuclear
388c,d, n
50
85e, k,n
0i
5700e,p,q
13e,a
8e,a
98a
Table A. 1: LCOE of new generation technologies in 2011-2012 (first quarter) with starting assumptions for simulation up to 2030
a) our calculation; b) EPIA, 2010 (55); c) REN21, 2011 (1); d) IEA ETSAP, 2010 (56); e) IEA, 2010 (57); f) discussion with the industry; g) EWEA, 2011 (58); h)
Weiss et al., 2010 (49); i) Neij, 2008 (42); j) IEA, 2009 (59); k) Lenzen, 2010 (60); l) Sikkema et al., 2011 (61); m) Faaij, 2006 (62); n) World Nuclear Association
(63); o) Laleman et al., 2011. (30); p) NY times (64); q) Guardian (65)
403
Appendix B
In Table B. 1, the projected evolution of fuel and carbon costs is summarized. In our simulations,
we also considered the impact of lower load factors for non-intermittent generation technologies in
response to an increasing share of intermittent generation. The two left columns of Table B. 1
present the load factors of Table A. 1 used in the stable load factor scenarios and the reduced
load factors used in the reduced load factor scenarios (see further). The three fuel cost levels for
biomass technologies by 2030 in Table B. 1 refer to the three pellet pr ice scenarios; optimistic (O),
standard (S) and pessimistic (P).
2010
2030
2010
2030
2030
2030
Fuel Cost
Fuel Cost
CO2 Price
CO2 Price
Stable LF
Reduced
/MWh
/MWh
/MWh
/MWh
100% Biomass
73
73 - 89 - 145
(O S P)
80
70
Cofiring(Sim)
54
11.5
23
80
70
Coal
30
66 - 75 - 103
(O S P)
60
23
46
80
60
Gas
Nuclear
50
8
90
6
10
0
20
0
70
85
50
65
404
References
[1]
[2]
[3]
[4]
yield and soil carbon balance in life cycle analyses. Global Change Biology Bioenergy.
2009;1(4):267-81.
[21] Jaramillo P, Gr iffin W, Matthews H. Comparative life-cycle air emissions of coal,
domestic natural gas, LNG, and SNG for electricity generation. Environmental Science &
Technology. 2007;41(17):6290-6.
[22] IEA. Energy Technology Perspectives 2010: International Energy Agency; 2010.
[23] Junginger M, Bolkes jo T, Bradley D, Dolzan P, Faaij A, Heinimo J, et al. Developments in
international bioenergy trade. Biomass & Bioenergy. 2008;32(8):717-29.
[24] Greenpeace.
Fuelling
a
biomess2011.
Available
from:
http://www.greenpeace.org/canada/en/campaigns/boreal/Resources/Reports/Fuelling-aBiomess/.
[25] Zhang Y, Mckechnie J, Cormier D, Lyng R, Mabee W, Ogino A, et al. Life Cycle
Emissions and Cost of Producing Electricity from Coal, Natural Gas, and Wood Pellets in
Ontario, Canada. Environmental Science & Technology. 2010;44(1):538-44.
[26] Brando M, Clift R, Mil i Canals L, Basson L. A Life-Cycle Approach to Characteris ing
Environmental and Economic Impacts of Multifunctional Land-Use Systems: An Integrated
Assessment in the UK. Sustainability. 2010;10(2):3747-76.
[27] Lettens S, Muys B, Ceulemans R, Moons E, Garcia J, Coppin P. Energy budget and
greenhouse gas balance evaluation of sustainable coppice systems for electricity production.
Biomass & Bioenergy. 2003;24(3):179-97.
[28] Spielmann M, Bauer C, Dones R, Tuchschmid M. Transport services. ecoinvent report n
17 Swiss center for life cycle invertories,. Dbendorf: 2007.
[29] DIRECTIVE 2010/75/EU OF T HE EUROPEAN PARLIAMENT AND OF THE
COUNCILof 24 November 2010on industrial emissions (integrated pollution prevention and
control) [database
on
the
Internet].
2010.
Available
from:
http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2001:309:0001:0021:EN:PDF.
[30] Laleman R, Albrecht J, Dewulf J. Life Cyc le Analys is to estimate the environmental
impact of res idential photovoltaic systems in regions with a low solar irradiation. Renewable
& Sustainable Energy Reviews. 2011;15(1):267-81.
[31] Martinez E, Sanz F, Pellegrini S, Jimenez E, Blanco J. Life cycle assessment of a multimegawatt wind turbine. Renewable Energy. 2009;34(3):667-73.
[32] Azapagic A, editor. Sustaining Future Ecosystem Services. 2009; London.
[33] Pehnt M. Dynamic life cycle assessment (LCA) of renewable energy technologies.
Renewable Energy. 2006;31(1):55-71.
[34] Djomo S, El Kasmioui O, Ceulemans R. Energy and greenhouse gas balance of bioenergy
production from poplar and w illow : a review. Global Change Biology Bioenergy.
2011;3(3):181-97.
[35] Manish S, Pillai IR, Banerjee R. Sustainability analysis of renewables for climate change
mitigation. Energy for Sustainable Development. 2006;10(4):25-36.
[36] Odeh NA, Cockerill TT. Life cycle analysis of UK coal fired power plants. Energy
Conversion and Management. 2008;49(2):212-20.
[37] Lenzen M, Munksgaard J. Energy and CO2 life-cycle analyses of wind turbines - review
and applications. Renewable Energy. 2002;26(3):339-62.
[38] International Energy Agency. Projected costs of generating electricity. 2010.
[39] EEA. Europe's onshore and offshore wind energy potential. European Environment
Agency, 2009.
[40] Kobos PH, Erickson JD, Drennen TE. Technological learning and renewable energy costs:
implications for US renewable energy policy. Energy Policy. 2006;34:1645-58.
406
[41] McDonald A, Schrattenholzer L. Learning rates for energy technologies. Energy Policy.
2001;29(4):255-61.
[42] Neij L. Cost development of future technologies for power generation - A study based on
experience curves and complementary bottom-up assessments. Energy Policy.
2008;36(6):2200-11.
[43] Nemet GF. Beyond the learning curve: factors influenc ing cost reductions in photovoltaics.
Energy Policy. 2006;34(17):3218-32.
[44] Pan H, Kohler J, editors. Technological change in energy systems: Learning curves,
logistic curves and input-output coefficients. Forum on Sustainability Well Being and
Environmental Protection; 2005 Dec 02; Washington, DC.
[45] Rout UK, Bles l M, Fahl U, Remme U, Voss A. Uncertainty in the learning rates of energy
technologies: An experiment in a global multi-regional energy system model. Energy Policy.
2009;37(11):4927-42.
[46] Soderholm P, Sundqvist T. Empirical challenges in the use of learning curves for assessing
the economic prospects of renewable energy technologies. Renewable Energy.
2007;32(15):2559-78.
[47] van der Zwaan B, Rabl A. The learning potential of photovoltacs: implications for energy
policy. Energy Policy. 2004:1545-54.
[48] Albrecht J. The future role of photovoltaics: A learning curve versus portfolio perspective.
Energy Policy. 2007;35(4):2296-304.
[49] Weiss M, Junginger M, Patel MK, Blok K. A review of experience curve analyses for
energy demand technologies. Technological Forecasting and Social Change. 2010;77(3):41128.
[50] Eurelectric. Integrating intermittent renewables sources into theEU electricity system by
2020:
challenges
and
solutions2010.
Available
from :
http://www.eurelectric.org/PublicDoc.asp?ID=63539.
[51] Pyry. The challenges of intermittency inNorth West European power markets2011.
Available from: http://www.poyry.com/linked/en/press/NEWSIS.pdf.
[52] Pyry. Analys is technical constraints on renewable generation to 2050: A report to the
Committee
on
Climate
Change2011.
Available
from:
http://hmccc.s3.amazonaws.com/Renewables%20Review/232_Report_Analys ing%20the%20t
echnical%20constraints%20on%20renewable%20generation_v8_0.pdf.
[53] Mott MacDonald. Costs of low-carbon technologies2011. Available from:
http://hmccc.s3.amazonaws.com/Renewables%20Review/MML%20final%20report%20for%2
0CCC%209%20may%202011.pdf.
[54] Committee on Climate Change. Renewable energy review2011. Available from:
http://www.theccc.org.uk/reports/renewable-energy-review.
[55] EPIA. Global Market outlook for photovoltaics until 2014. 2010.
[56] International Energy Agency. Energy Technology System Analysis Program. 2010;
Available from: http://www.iea-etsap.org/web/Supply.asp.
[57] International Energy Agency. Energy Technology Perspectives 2010. Paris: IEA; 2010.
[58] EWEA. Pure Power, wind energy targets up to 2020 and 2030. 2011.
[59] International Energy Agency. Technology Roadmap: Wind energy. International Energy
Agency
[Internet].
2009.
Available
from :
http://www.ieawind.org/wnd_info/IEA_Paris/Wind_Roadmap.pdf.
[60] Lenzen M. Current State of Development of Electric ity Generating Technologies : A
Literature Review. 2010;2010(3):462-591.
407
[61] Sikkema R, Steiner M, Junginger M, Hiegl W, Hansen M, Faaij A. The European wood
pellet markets: current status and prospects for 2020. Biofuels Bioproducts & BiorefiningBiofpr. 2011;5(3):250-78.
[62] Faaij A. Modern Biomass Conversion Technologies. Mitigation and Adaptation Strategies
for Global Change. 2006;11(2):335-67.
[63] World Nuclear Association. Available from: http://www.world-nuclear.org/.
[64] Kanter J. In Finland, Nuclear Renaissance Runs Into Trouble. New York Times; 2009;
Available
from:
http://www.nytimes.com/2009/05/29/business/energyenvironment/29nuke.html?pagewanted=all.
[65] Carrington D. Nuclear power's real chain reaction: spiralling costs. The Guardian; 2011;
Available
from:
http://www.guardian.co.uk/environment/damian-carringtonblog/2011/jul/22/nuc lear-power-cost-delay-edf.
408
Abstract:
This paper deals with the use of straight vegetable oil as fuel for compression ignition engine applied to
distributed electric generation. It was studied three typical oils from southern of the Brazil, soybean oil,
sunflower oil and tung oil. For this purpose it was designed and assembled two conversion kits that allow the
use of the selected oils directly in the engine. The first kit preheats the fuel to a temperature where their
physical properties, mainly viscosity, reaches the diesel oil levels. The other kit is an electronic injection
mobile bench that allows controlling the timing, the pressure and the mass of injected fuel. Both kits are
electronic controlled and it was used standardized measurements of some physical properties of the fuels,
for the design of their control software. The straight vegetables oils, its blends 50/50 v/v with petrodiesel fuel
and neat petrodiesel fuel were tested in dynamometer bench. It was obtained results of brake power, torque,
specific fuel consumption and emissions as function of the rotation of the engine. The discussions about the
results in terms of efficiency first law showed the technical feasibility of the using of straight vegetable oils
and the effecti veness of the developed conversion kits.
Keywords:
Renewable energy, Vegetable Oils, Biofuels, Compression Ignition Engine.
1. Introduction
The use of complementary sources for the Brazilian energetic matrix has been sought by
government and corporative stakeholders. Biofuels are the choice for transportation, either as a neat
fuel or as an additive to petroleum derived fuels (Martines-Filho et al. [47], Pousa et al. [57]).
Ethanol has been used in automobiles mixed with gasoline in volumetric fraction from 20% to
100%. Biodiesel has also been mixed to diesel oil in a 5% volumetric basis. This has led to
reduction of sulphur content in fuels for CI engines, higher octane rating fuel for SI engines and to a
more sustainable transportation energy matrix. There are also applications in which the presence of
biofuels could increase. For example, isolated communities in the Brazil northern areas shall use
diesel oil for electrical energy generation. This diesel consumption could be displaced by a more
environmentally friendly and economical fuel mix (Andrade and Miccolis [9]). Rural areas can also
benefit from the use of locally produced fuels. In this scenario, oil from vegetable and animal
sources could provide alternative solutions.
One of the routes for the use of vegetable oil and fats is the transesterification that leads to
monoalkyl esters known as biodiesel (Shahid and Jamal [72]). Several studies have shown that
biodiesel can displace diesel oil without the need for any modifications in CI engines, both for
transportation or stationary applications (see references in Agarwal [2], Dwivedi et al. [25]).
However, this requires an industrial transesterification process that carries its own intrinsic
complexities and costs. A more direct approach would be the use of straight vegetable oil in CI
409
engines, obtained after filtration only, with a minimum of modification of the engines hardware
(Sidib et al. [68]).
Table 1. List of the literature reporting tests in CI engines fuelled with raw vegetable oils and their
mixtures with diesel oil.
Character
Ed ible
Nonedible
References
Engelman et al. [27], Pryor et al. [61], A ltin et al. [6]
Peterson et al. [56], Nwafor [51], Bialkovski et al.
[18], Hazar and Aydin [33], Kleinova et al. [42],
Yilmaz and Morton [79]
Sapaun et al. [70], Bari et al. [12], Almeida et al. [5],
Belchior and Pimentel [15], Antwi [10]
Thaddeus [74], Kalam et al. [41], Antwi [10]
Altin et al. [6], A mba and Rama [8], He and Bao [34],
Fontaras et al. [29], Siva et al. [73], Sarada et al. [67],
Martin and Prithviraj [46]
Altin et al. [6]
Rakopoulos et al. [66]
Karaos manoglu et al. [39], Alt in et al. [6], Maziero et
al. [48], Yilmaz and Morton [79]
Barsic and Hu mke (1981), Yilmaz and Morton (2011)
Bettis et al. [14], Isigigur et al. [38]
Altun et al. [7]
Bari and Roy [12], Agarwal et al. [2], Raghu et al.
[64]
Mbarawa [47]
It is well accepted that Rudolph Diesel used peanut oil in a public demonstration of his engine
(Knothe [40]). Times of war and the petroleum crises in the seventies have spurred modern
developments in the use of raw vegetable oils in diesel engines. Chan and Wan [22] report studies
carried out in continental China during the Second World War using tung oil as fuel for
compression ignition engine. In that work they presented the performance of a compression ignition
engine fuelled with diesel oil, cotton seed oil, rapeseed oil and tung oil. Recent reviews may be
410
found in Bhattacharya and Reddy [17], Sinha and Misra [71], Babu and Devaradjane [11], Mondal
at al. [49] and No [50].
It is estimated that there are in Brazil more than 20 varieties of crops for the production of vegetable
oils. Among them, the varieties that have national or regional interest are soybean, castor beans,
sunflower, peanut, cottonseed, babassu, palm, jatropha and canola (CONAB [24]). Most of them
(and others) have already been tested as neat fuels for CI engines. Table 1 lists oil producing crops
that have been used as straight vegetable oil (SVO) fuel in CI engines in recent years. The
references for their use as biodiesel far outnumber those shown and are omitted.
Vegetable oils present, comparatively to diesel oil, lower LHV (from 10% to 17% lower, leading to
lower energy release per mass burned), higher viscosity (leading to poor atomization), higher
boiling temperatures (delaying evaporation and formation of a combustible mixture), higher bulk
modulus (causing injector to open earlier), higher flash point (delaying mixture ignition), higher
oxidation instability (leading to higher tendency to degradation during storage) and a tendency for
thickening with time (Babu and Devaradjane[11], Franco and Nguyen [28]). When using the same
injectors and settings adjusted for diesel fuel, the higher viscosity, surface tension and density of the
vegetable oils result in changes in injected oil volumes, injection delay after injector opening, spray
patterns (cone and penetration) and atomization (droplet size distribution) (Bialkowski et al. [18]).
As a result of poor atomization, mixing and ignition there are:
(a) longer ignition delay, smaller pressure rise, lower cylinder peak pressure and a longer
combustion duration (Venkanna et al. [75]), resulting in 5% to 25% (Chalatlon et al. [20] )
reduction in thermal efficiency at maximum power when compared to pure diesel oil,
(b) the formation of carbon deposits on fuel injectors, ring landings, exhaust valves, and
contamination of the lubricating oil by vegetable oil. These effects can severely limit the
engine lifespan.
(c) a reduction in the formation of NOx, but a possible increase in smoke, CO and HC
(Pimentel et al. [58], Belchior and Pimentel [15]).
Tests with long term operation (> 60 hours) with neat sunflower oil (Maziero et al. [48]) have
reported obstruction of the lubricating oil flow galleries, damage of piston rings and gaskets,
presence of tar in the exhaust manifolds and the presence of elevated concentration of copper,
chromium, iron and lead in the lubricating oil. Bialkowski et al. [18] (working with Elsbett,
Germany) reported coking, spray obstruction by lacquer formation at spray tip and engine
deterioration. They also reported problems with slow flow along oil lines, filter clogging and engine
oil deterioration.
Since the pioneering work, a few solutions have been proposed and implemented:
(a) The viscosity of the vegetable oil can be reduced by heating (Forson et al. [30], Venkanna et
al. [75], Kleinova et al. [49], Aksoi [4]). The oil temperature before injection has been
regulated from 20oC (Bialkowski et al. [18]) to 160oC (Raghu and Ramadoss [64]). Most
authors preheat the SVO such as to bring the physical properties closer to those of diesel oil.
Most commonly, the oil temperature does not exceed 130oC (Pugazhvadivu and
Sankaranarayanan [62]).
(b) Blending vegetable oil with diesel also decreases viscosity and improves volatility. These
improved properties result in better mixture formation and spray penetration. A number of
investigators tried the vegetable oils in varying proportions with diesel. Most remarkably,
few studies (Forson et al. [30]) show engine performance even above that of operation with
neat diesel oil.
411
(c) Advanced injection timing compensates the effects of the longer delay period and slower
burning rate that is exhibited by vegetable oils (Nwafor and Rice [52]). Staged injection may
not lead to improvement in fuel/air mixing when it occurs too late along the expansion cycle
(Bialkowski et al. [18]).
Most of the recent work was developed using low power single cylinder CI engines fuelled by
mechanical pumps (Altin et al. [6], Forson et al. [30], Venkanna et al. [75], Pugazhvadivu and
Sankaranarayanan [62], Sarada et al. [67], Aksoi [4], Martin and Prithviraja [46], Raghu and
Ramadoss [64], Sivalakshmi and Balusamy [69]) in the context of the application of small engines
for rural and remote areas. Fewer works were developed with larger engines for general use
(Bialkowski et al. [18]), for use in agriculture (Maziero et al. [48]), and for transportation (Kleinova
et al. [42], Chalatlo et al.[20]). Usually, large IDI (indirect injection compression ignition) operate
better during long duration tests while small IDI and DI engines present problems (He and Bao
[34], Bialkowski et al. [18]). Only Kleinova et al. [42] and Bialkowski et al. [18]) have developed
their studies using common rail injection systems. Most modern CI engines nowadays employ
common rail injection. The use of a central electronic unit and the common rail has enabled great
advances in performance and in-cylinder emissions control using diesel oil and these improvements
could also be expected when using straight vegetable oils. For example, Venkanna et al. [75]
investigated the effect of the injection pressure in a mechanical system, varying the injection
pressure from 200 bar to 280 bar. Even in this small range they measured differences in thermal
efficiency that point out to an optimum operation pressure for a given combustion chamber, injector
and oil temperature. They argue that, for their engine, a further increase in pressure would cause an
excessive diminution of droplet sizes and insufficient spray penetration. They also noticed that
smoke reduces continually with the increase in pressure. This indicates the need to explore further
the effects of injection pressure, injection timing, and split injection.
This work is part of a larger project including engine test in a dynamometric bench, measurement of
spray parameters in an interferometric laser bench and studies of maintainability of the engine when
applied to stationary electricity generation. The main project is supported by the Southern Brazilian
Power Plants Company through its Research and Development Fund. Here, the initial tests of a
mechanical injected engine in a dynamometric bench operating with straight vegetable oils of
Soybean, Sunflower and Tung and their mixtures with diesel oil are reported. This work relies on
the assumption that oil heating and higher injection pressure contribute to a better spray
development and atomization, leading to better performance, efficiency and smaller emission of
smoke. The basic strategy for pre-heating consists in bringing the straight vegetable oil before the
injector to a temperature in which the viscosity of the oil approaches that of the diesel oil at ambient
temperature. To allow for the control of the fuel heating an electronically controlled heating unit
was developed and adapted to the engine. This set up is described next.
controlled using a student version of the LabVIEW software. Table 2 lists the main characteristics
of the engine.
Table 2. Engine basic characteristics.
Manufacturer
Yanmar
Model
YT22
Bore [mm]
115
Stroke [mm]
115
Compression Ratio
17.3:1
Volume [cm]
1194
Nominal conditions:
Speed [rpm]
2000
Power [kW]
14,7
BSFC [g (kWh)-1]
238
Injection pressure [bar]
200
Injection timing [BTDC]
18
The engine had a mechanical speed control system that uses a mechanism called governor. To
adjust the engine to full load, the handle load is set to its maximum. This load handle is connected
to an arm that directly acts in the engine injection pump, controlling the fuel flowing to the
combustion chamber. Between the handle and the arm there is a spring. Also connected to the arm,
there is a centrifugal regulator that acts towards diminishing the fuel flowing if the engine speed
increases. The engine subsystem formed by the arm, the spring and the centrifugal regulator is
called governor. The governor has the function of keeping the engine in the same speed set by the
operator, when the engine is subject to variable loads. Figure 1 pictures the governor.
to the engine, bringing back the speed to the initially value set. In this regime, the speed remains
approximately constant, as presented in the high speed range (> 2000 rpm) in figure 2. This process
continues until the spring is totally stretched. When this occurs, any subsequent increase in the
dynamometer load result in the decrease of the engine speed, completing the remaining of the
characteristic curves, as showed for the lower speeds (< 2000 rpm) in figure 2.
3. Fuels
The three vegetable oils and their blends with commercial Brazilian agricultural diesel oil were
used. Brazilian diesel oil has a volumetric addition of 5% of biodiesel, as required by law and
regulated by ANP, the National Petroleum Agency. The fuels used are labelled as presented in
Table 4.
The straight vegetable oils and their respective mixtures were preheated before injection in the
engine. The strategy was to bring the fuel properties close to those of diesel oil. For that, the
viscosities of the fuel mixtures were measured as a function of temperature by standardized
414
viscosity experiments carried out in the National Institute of Technology (INT [37]) at Rio de
Janeiro. Table 5 summarizes the injection temperature, the kinematic viscosity and the specific
mass of the fuel blends. The specific mass was measured using Archimedes principle with a Kern
electronic balance model EW 220 3 NW. The fuel temperature was kept sufficiently low to avoid
thermal degradation.
Table 4. Nomenclature for the fuel mixtures used.
100% SW
Straight Sunflower Oil
100% SY
Straight Soybean Oil
100% TG
Straight Tung Oil
50%SW-50%D
Volumetric Mixture between Sunflower Oil and diesel oil
50%SY-50%D
Volumetric Mixture between Soybean Oil and diesel oil
50%TG-50%D
Volumetric Mixture between Tung Oil and diesel oil
100% D
Brazilian Commercial (countrifield) diesel oil
Table 5. Injection temperatures and kinematic viscosity measured (INT [37]).
Injection
Kinematic
Specific mass
Fuel
Temperature
Viscosity [mm/s]
[kg/m3]
[C]
100% SW
85
10.52
863.2
100% SY
85
8,48
859.8
100% TG
95
16.33
869.1
50%SW-50%D
65
5.45
857.4
50%SY-50%D
65
5.53
856.1
50%TG-50%D
85
8,26
854.9
100% D
25
4.52
846.4
Table 6 summarizes the energy content of the fuel blends. It is observed that the diesel oil has the
higher lower heating value (LHV), which is about 15% greater than that of the straight vegetable
oils. For completeness, Table 7 shows the results for an elemental analysis of the vegetable oils. For
comparison, for an alkane with sixteen carbon atoms, the percentage mass of carbon is near 85%
and of hydrogen is 15%.
Fuel
100% SW
100% SY
100% TG
100% D
415
The main technical development in this project is the conversion kit to control the preheating of the
straight vegetable oils and blends to the temperatures showed in table 3. It is important that besides
preheating the fuels, the temperature must be controlled within a narrow range. The electronic
control unit ECU was developed and assembled in the lab. The main component of the ECU is a
microcontroller model dsPIC. The conversion kit preheats the vegetable oil using the rejected heat
of the engine. This heat is recovered from the exhaust gases by a heat exchanger that uses a solution
of water and mono-ethylene glycol as thermal fluid. The solution is then pumped to another heat
exchanger, where the vegetable oils are heated to the selected temperature. The temperature of the
vegetable oil is measured at the outlet of the heat exchanger. This value is sent to the ECU that
controls the temperature of the vegetable oils by controlling the flow of the thermal fluid. A second
conversion kit that will allow the use of a common rail system is under development. The results
reported here were obtained with a mechanical control of injection, as it will be explained below.
4. Results
In this section, the results obtained in the dynamometric bench are shown. Figures 2 and 3 present
power. Figure 4 presents the thermal efficiency. Figures 5 and 6 present fuel consumption. Figures
7 to 9 present emissions and exhaust gas temperature.
In figure 2, the region for speed higher than 2000 rpm corresponds to the spring-controlled regime.
In this region, the curves for all fuels are similar. This is supposed to occur, since the governor will
allow for the fuel needed to sustain a constant speed. The fact that this is observed in the results
indicates that no accidental event occurred. In the following, each curve is analyzed separately.
416
13,0
100% SW
100% SY
100% TG
100% D
50% SW - 50% D
50% SY - 50% D
50% TG - 50% D
12,5
Power (kW)
12,0
11,5
11,0
10,5
10,0
9,5
9,0
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
1 300
0,0
1400
1 500
1600
1 700
1800
19 00
2000
21 00
2200
-2,0
-4,0
-6,0
-8,0
-10,0
-12,0
1 00 % SW
-14,0
1 00 % SY
-16,0
1 00 % TG
-18,0
5 0% SW - 5 0% D
5 0% SY - 50 % D
-20,0
5 0% TG - 50% D
-22,0
Fig. 3. Percentage delivered power difference versus engine speed. The lines represent 2 nd degree
polynomial curve fits.
Figure 4 presents the mean value of fuel mass injected per cycle as a function of engine speed. The
amount of fuel injected increases as the speed decreases for all fuels. This occurs because when the
speed decreases, the roller that drives the injection pump actually stays longer pressing the pump,
thus injecting a larger amount of fuel. Notice that this increase in fuel injected per cycle may not
mean an increase in fuel mass flow rate.
417
Since the engine speed is decreasing and the load handle and the spring are both promoting the
maximum pump injector aperture, i.e. the set volumetric flow rate to the injection pump is the same
for the range since 2000 rpm to lower speeds, the volumetric fuel flow increases with the decrease
of the engine speed. It is interesting to note that, again, the vegetable oils delivered lower torque
than diesel oil, as can be obtained through the analysis of figure 2.
64
62
60
58
mg/cycl e
56
54
52
100% SW
50
100% SY
100% TG
48
100% D
46
50% SW - 50% D
44
42
1300
50% SY - 50% D
50% TG - 50% D
1400
1500
1600
1700
1800
1900
2000
2100
2200
Fig. 4. Mean fuel mass flow per engine cycle versus engine speed. The lines represent 2nd degree
polynomial curve fits.
310
1 00% SW
300
290
SFC (g/kWh)
280
1 00% SY
1 00% TG
Dies el
5 0%SW - 50 %D
5 0%SY - 50%D
5 0%TG - 5 0%D
270
260
250
240
230
220
1300
1400
1500
1600
1 700
1800
1900
2000
2100
2 200
Fig. 5. Specific fuel consumption - BSFC versus engine speed. The lines represent 2 nd degree
polynomial curve fits.
418
Figure 5 presents the brake specific fuel consumption (BSFC) as a function of engine speed for the
different oils and their blends. The BSFC increases with speed. All vegetable oils and blends result
in higher BSFC when compared to diesel oil. The BSFC for tung oil at 2000 rpm is approximately
18% higher than that for diesel oil. The blend with 50% soybean oil approximates more the BSFC
of the diesel oil than the BSFC of the pure soybean oil, resulting in a 2% increase only at 2000 rpm.
We note that the BSFC for diesel oil at the nominal speed matches the value listed in Table 2.
0,39
0,38
0,37
0,36
0,35
0,34
0,33
100% SW
100% SY
100% TG
0,32
100% D
50% SW - 50% D
0,31
50% SY - 50% D
50% TG - 50% D
0,30
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
Fig.6. First law efficiency versus engine speed. The lines represent 2 nd degree polynomial curve fits.
Figure 8 presents the volumetric NOx fraction in the exhaust gases as a function of engine speed.
The operation with soybean and tung resulted in smaller emission of NOx. Sunflower resulted in a
mirrored higher emission of NOx. It was noted that sunflower and soybean resulted both in
approximately the same values of power and BSFC, but sunflower emitted less CO than soybean.
Figure 9 presents the temperature of the exhaust gases as a function of engine speed. All vegetable
oils and their blends present lower exhaust gas temperature than diesel oil. Sunflower and tung
present the same exhaust temperature while soybean present the higher temperature.
900
100% SW
100% SY
800
100% TG
700
100% D
50% SW - 50% D
600
CO (ppm)
50% SY - 50 % D
50% TG - 50% D
500
400
300
200
100
0
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
1800
100% SW
1700
100% SY
1600
100% TG
NOx (ppm)
1500
100% D
1400
50% SW - 50% D
1300
50% SY - 50% D
1200
50% TG - 50% D
1100
1000
900
800
700
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
Fig. 8. Fraction of NOx in exhausted gases versus engine speed. The lines represent 2 nd degree
polynomial curve fits
420
550
Temperature (C)
525
500
475
100% SW
100% SY
450
100% TG
100% D
425
50% SW - 50% D
50% SY - 50% D
50% TG - 50% D
400
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
Fig. 9. Temperature of the exhaust gases versus engine speed. The lines represent 2 nd degree
polynomial curve fits.
The lower exhaust temperature is consistent with the higher efficiency, since more of the fuel
energy is being transformed into useful work.
5. Conclusions
The results show that the engine and heating system worked reasonably well for the straight
vegetable oils and their blends used.
The curves of power and specific fuel consumption as a function of engine speed behaved in
accordance with the values of LHV and specific mass of vegetable oils and their blends. The lower
LHV and higher specific mass of the SVO fuel tested led to a decrease in power as well as to an
increase in specific fuel consumption when compared to the operation with neat diesel oil.
The results for thermal efficiency were particularly interesting. They showed that the thermal
efficiency for sunflower and soybean oils are higher than that for the neat diesel oil. Besides, the
best efficiencies (around 37.5%) were achieved by the 50% blend of soybean, which are about 6%
higher than that for neat diesel at the same power output of 11 kW, at 1500 rpm. Also, the blends of
soybean and sunflower presented thermal efficiencies that were higher than the diesel oil and the
neat fuels themselves. Somehow, there is some improvement when the fuels are mixed. The reason
for this behaviour is not clear at the moment. This is encouraging from the point of view of
increasing the efficiency of the electrical energy generation systems.
The CO emitted by the engine increased as the engine speed decreased. This occurred because the
mechanical control of injection results in higher amount of fuel injected per cycle as the speed is
reduced. Since the volumetric efficiency remains constant or decreases (it was not measured), the
mixture might become richer for lower speeds and burn over a longer combustion period. The
emission of NOx is also higher at lower speeds, as a result probably of a longer residence time at
lower speeds. The exhaust gas temperature was lower for the vegetable oils and their blends when
compared to neat diesel, indicating that there is room to improve atomization and combustion.
In order to improve atomization, a common rail system will be adapted to the engine to provide
injection pressures up to 800 bar. Also, the use of an electronically controlled injection system will
421
allow for a more systematic study of the effect of the advance in the injection timing and the use of
a pre-injection. This will be reported later, in a future paper.
Acknowledgments
To Eletrobrs Eletrosul Centrais Eltricas S. A. for the financial support.
References
[1]
[11] Babu, A. K. and Devaradjane, G., Vegetable Oils And Their Derivatives As Fuels For CI
Engines: An Overview, SAE Technical Paper 2003-01-0767, 2003.
[12] Bari, S., Roy, M. M., Prospect of rice bran oil as an alternative to diesel, Proc. of the Fifth
Int. Conf. on small engines, their fuels and the environment. The Universty of Reading, UK,
pp. 31-36, 1995.
[13] Bari, S., Lim T. H. and Yu, C. W. , Effect of preheating crude palm oil on injection system
performance and emissions of a diesel engine. Renew. Energy. 77, 339-351, 2002.
[14] Barsic, N. J and Humke, A. L., Performance and emissions characteristics of a naturally
aspirated diesel engine with vegetable oil fuels. SAE. 810262, 1173-1187, 1981.
[15] Belchior, C. R. P., Pimentel, V. S. B., The use of palm oil in diesel engine, 18th
International Congress of Mechanical Engineering, Ouro Preto MG Brasil, 2005.
[16] Bettis, B. L., Peterson, C. L., Auld, D. L. , Driscoll, D. J., and Peterson, E. D. ,Fuel
characteristics of vegetable oil from oilseed crops in the Pacific Northwest. Agronomy
Journal, 74(March/April) pp 335-39,1982.
[17] Bhattacharya, .S and Reddy, C.S, Vegetable oils as Fuel for Internal combustion Engines: A
Review, Journal of Agricultural, Engineering Research. 57, pp. 157-166, 1994.
[18] Bialkowski, M.T.; Pekdemir, T.; Reuben, R.; Brautsch, M.; Towers, D. P.; Elsbett, G.,
Preliminary Approach Towards a CDI System Modification Operating on Neat Rapeseed
Oil Journal of KONES, vol 12 Issue 1, 2005.
[19] Bialkowski, M. T., Theoretical And Experimental Investigation Of A Cdi Injection System
Operating On Neat Rapeseed Oil - Feasibility And Operational Studies, Ph.D. Thesis,
Heriot-Watt University, 2009.
[20] Chalatlon, V., Roy, M. M., Dutta, A., Kumar, S., Jatropha oil production and an
experimental investigation of its use as an alternative fuel in a DI diesel engine, JPTAF,
2(issue 5), pp 76-85, 2011.
[21] Chandra, V. M. R., Experimental investigations on a four stroke compression ignition
engine using Eco friendly Non edible vegetable oils, PhD Thesis, Jawaharlal Nehru
Technological University, 2011.
[22] Chang, C. C., Wan, S. W., Chinas Motor Fuels from Tung Oil, Ind. Eng. Chem. 39, pp
1543-1548, 1947.
[23] Chauhan, B. S., Kumar N., Jun Y. D. , Lee K. B., Performance and emission study of
preheated Jatropha oil on medium capacity diesel engine. Energy, 35, pp 2484-2492, 2010.
[24] CONAB, Brazilian Crop Assessment: Grain, National Company of Food Supply, Third
Assessment, Brasilia, December, 2011.
[25] Dwivedi, D., Agarwal, A. K. and Sharma, M., Particulate Emission Characterization of a
Biodiesel vs. Diesel-Fuelled Compression Ignition Transport Engine: A Comparative Study,
Atmo[26] spheric Environment, 40, 5586-5595, 2006.
[27] Engelman, H. W., Guenther, D. A., and Silvis, T. W., Vegetable oil as a diesel fuel. Diesel
and Gas Engine Power Division of ASME Paper Number 78-DGP-19. New York, NY:
ASME, 1978.
[28] Franco, Z., Q.D. Nguyen, Flow properties of vegetable oildiesel fuel blends, Fuel, Volume
90, Issue 2, Pages 838-843, 2011.
[29] Fontaras, G., Samaras, Z., Miltsios, G., Experimental evaluation of cottonseed oil- diesel
blends as automotive fuels via vehicle and engine measurements. SAE paper, No. 2007-240126, 2007.
[30] Forson, F. K., Oduro, E. K., Hammond-Donkoh, E., Performance of Jatropha oil blends in a
diesel engine. Renew. Energ. 29(7): 1135-45, 2004.
[31] Giannelos, P. N., Zannikos, F., Stournas, S., Anastopoulos, G., Tobacco seed oil as an
alternative diesel fuel: physical and chemical properties, Industrial Crops and Products, 16,
pp. 1 9, 2002.
423
[32] Haldar, S. K., Ghosh BB, Nag A. Studies on the comparison of performance and emission
characteristics of a diesel engine using three degummed non-edible vegetable oils. Biomass
Bioenerg. 33: 1013-1018, 2009.
[33] Hazar, H., Aydin, H., Performance and emission evaluation of a CI engine fuelled with
preheated raw rapeseed oil (RRO)-diesel blends. Appl. Energy, 87(3): 786-790, 2010.
[34] He, Y., and Bao, Y. D., Study on cottonseed oil as a partial substitute for diesel oil in fuel
for single-cylinder diesel engine. Renewable Energy, 30: 805-813, 2005.
[35] Herchel, T., Machacon, C., Matsumoto, Y., Ohkawara, C., Shiga, S.,Nakamura, H. The
effect of coconut oil and diesel fuel blends on diesel engine performance and exhaust
emissions, JSAE Review, Volume: 22, Issue: 3, Pages: 349-355, 2001.
[36] Heywood, J. B., Internal Combustion Engine Fundamentals. New York, USA: McGraw-Hill
Inc, 1988.
[37] INT, Experiment Report 00300/2011, National Institute of Technology, Rio de Janeiro-RJ,
2011.
[38] Isigigur, A. , F. Karaosmanoglu, H. A. Aksoy, Hamdullahpur, F. and Gulder, O. L. ,
Safflower seed oil of Turkish origin as a diesel fuel alternative, Applied Biochemistry and
Biotechnology, Volume 39-40, Number 1, 89-105, 1993.
[39] Karaosmanoglu, F., Kurt, G., zakta , T., Long term CI engine test of sunflower oil,
Renewable EnergyVolume 19, Issues 12, Pages 219221, 2000.
[40] Knothe, G., Historical Perspectives on Vegetable Oil-Based Diesel Fuels, Inform, Volume
12, pp. 103-1107, 2001.
[41] Kalam, M. A. , Husnawan, M., Masjuki, M. H., Exhaust emission and combustion
evaluation of coconut oil-powered indirect injection diesel engine. Renewable Energy, 28:
24052415, 2003.
[42] Kleinova, A., Vailing I., Franta, R., Mikulec, J., Cvengro, J., Vegetable oils as diesel fuels
for rebuilt vehicles, 44th International Petroleum Conference, Bratislava, Slovak Republic,
September 21-22, 2009.
[43] Knothe, G. H. Hictorical Perspectives on Vegetable Oil-Based Diesel Fuels. 2001. Available
at: http://www.ars.usda.gov. [Acessed 24.8.2011].
[44] Kumar, M. S., Ramesh, A., Nagalingam, B., An experimental comparison of methods to use
methanol and Jatropha oil in a compression ignition engine, Biomass Bioenerg., 25: 309318, 2003.
[45] Lefebvre A. H., Atomization and Sprays. Boca Raton/FL, USA: Routledge Taylor and
Francis Group, 1989.
[46] Martin, M. and Prithviraj, D., Performance of Pre-heated Cottonseed Oil and Diesel Fuel
Blends in a Compression Ignition Engine, Jordan Journal of Mechanical and Industrial
Engineering, Volume 5, Number 3, pps 235 240, 2011.
[47] Martines-Filho, J., Heloisa L. Burnquist, and Carlos E. F. Vian, Bioenergy and the Rise of
Sugarcane-Based Ethanol in Brazil, CHOICES, 2nd Quarter, 21(2), 2006.
[48] Maziero, J. V. G., Corra, M., ngaro, M. R., Bernardi, J. A., Storino, M., Diesel Engine
Performance With Raw Sunflower Oil Fuel, R. Bras. Agrocincia, Pelotas, v. 13, n. 2, p.
249-255, abr-jun, 2007.
[49] Naga, P. Ch. S, Vijaya, K.R.K, Kumar, B.S.P., Ramjee., E., Hebbel, O.D. and Nivendg,i
M.C., Performance and emission characteristics of a diesel engine with castor oil, Indian
Journal of Science and Technology, 2 (No.10), pp.25-31, 2009.
[50] No, S. Y., Inedible vegetable oils and their derivatives for alternative diesel fuels in CI
engines: A review. Renewable Sustain. Energy Rev., 15: 131-14, 2011.
[51] Nwafor, O. M. I., The effect of elevated fuel inlet temperature on performance of diesel
engine running on neat vegetable oil at constant sspeed conditions. Renewable Energy 28,
pp 171-181, 2003.
[52] Nwafor, O. M. I., Emission Characteristics of Diesel Engine Running on Vegetable Oil With
Elevated Fuel Inlet Temperature. Biomass and Bioenergy 27, PP 507-511, 2004.
424
[53] Nwafor, O. M. I. and Rice, G., Performance of rapeseed oil blends in a diesel engine,
International Journal of Applied Energy, Vol.54, No.4, 1996, 345-354.
[54] Park, H. S., Kim, H. J., Suh, H. K., Lee, C. S., Experimental and numerical analysis of
spray-atomization characteristicas of biodiesel fuel in various fuel and ambient temperatures
conditions, International Journal of Heat and Fluid Flow 30, pp960-970, 2009.
[55] P. K. Devan and N. V. Mahalakshmi, Performance, emission and combustion
characteristics of poon oil and its diesel blends in a DI diesel engine, Fuel, Vol. 88, 2009,
861-867.
[56] Peterson, C. L., Auld, D. L., Korus, R. A., Winter rape oil fuel for diesel engines: Recovery
and utilization, Journal of the American Oil Chemists' Society, Volume 60, Number 8, 15791587, 1983.
[57] Pousa, G. P.A.G., Andr L.F. Santos, Paulo A.Z. Suarez, History and policy of biodiesel in
Brazil, Energy Policy, Volume 35, Issue 11, Pages 53935398, 2007.
[58] Pimentel, V. S. De B., Belchior, C. R. P. , Pereira, P. P., Avaliao Experimental do
Desempenho e Emisses de Motores Diesel Usando Misturas de leo de Mamona e leo
Diesel. AGRENER 2004: Proceedings of the 16Th Conference on Rural Energy. BrasiliaDF, Brasil. PP 504-513, 2004.
[59] Mondal, P., Basu, M.,Balasubramanian, N., Direct use of vegetable oil and animal fat as
alternative fuel in internal combustion engine, Biofuels, Bioproducts and Biorefining,
Volume 2, Issue 2, pp 155174, 2008.
[60] Pramanik K. (2003). Properties and use of Jatropha oil and diesel blends in compression
ignition engine. Renewable Energy, 28: 239-248.
[61] Pryor, R. W., Hanna, M. A., Schinstock, J. L. and Bashford, L. L., Soybean oil fuel in a
small diesel engine. Transactions of the ASAE 26(2): 333-337, 1983.
[62] Pugazhvadivu, M., and Sankaranarayanan, G., Experimental studies on a diesel engine using
mahua oil as fuel, Indian Journal of Science and Technology Vol. 3 No. 7 ISSN: 09746846, 2010.
[63] Quick G. R. (1980). Developments in use of vegetable oils as fuel for diesel engines. ASAE
Paper, No. 80-1525.
[64] Raghu, R., Ramadoss, G., Sairam, K., Arulkumar, A., Experimental Investigation on the
Performance and Emission Characteristics of a DI Diesel Engine Fueled with Preheated
Rice Bran Oil European Journal of Scientific Research, Vol.64 No.3, pp. 400-414, 2011.
[65] Rakopoulos, C. D., Olive oil as a fuel supplement in DI and IDI diesel engines, Energy,
Volume 17, Issue 8, August 1992, Pages 787790.
[66] Rakopoulos, D. C., Rakopoulos, C.D., E.G. Giakoumis, A.M. Dimaratos, M.A. Founti,
Comparative environmental behaviour of bus engine operating on blends of diesel fuel with
four straight vegetable oils of Greek origin: Sunflower, cottonseed, corn and olive, Fuel,
Volume 90, Issue 11, November 2011, Pages 3439-3446.
[67] Sarada, S. N., Shailaja, M., K. Kalyani Radha, Optimization of injection pressure for a
compression ignition engine with cotton seed oil as an alternate fuel, International Journal of
Engineering, Science and Technology, Vol. 2, No. 6, pp. 142-149, 2010.
[68] Sidib, S.S., Blin, J. , Vaitilingom, G., Azoumah, Y., Use of crude filtered vegetable oil as a
fuel in diesel engines state of the art: Literature review, Renewable and Sustainable Energy
Reviews, Volume 14, Issue 9, Pp 2748-2759, 2010.
[69] Sivalakshmi, S., Balusamy, T.,Experimental investigation on a diesel engine using neem oil
and its methyl ester, Thermal Science Year, Volume 15, Issue 4, pp 1193 - 1204, 2011.
[70] Sapaun, S. M., Masjuki, H. H., Azlan, A., The use of palm oil as diesel substitute. Power
Energy (Part A) 210: 47-53, 1996.
[71] Sinha, S., Misra, N. C., Diesel fuel alternative from vegetable oils. Chem. Eng. World,
32(10), pp 77-80, 1997.
[72] Shahid, E. M., Jamal, Y., Production of Biodiesel: A Technical Review, Renewable and
Sustainble Energy Reviwes (15): 4732-4745, 2011.
425
[73] Siva, K.A, Maheswar, D., V. Kumar, R. K., Comparison of diesel engine performance and
emissions from neat, IJAEST (11), pp 121-128, 2009.
[74] Thaddeus, H. C. M., Asuyuki, S., Okobatae, T., Operation and Combustion Characteristics
ofa a DI Diesel Engine Fueled with Biomass Oil-Diesel, SAE Technical Paper 2001-280030, 2001.
[75] Venkanna, B. K., Swati, B., and Reddy, C. V., Effect of Injection Pressure on Performance,
Emission and Combustion Characteristics of Direct Injection Diesel Engine Running on
Blends of Pongamia Pinnata Linn Oil (Honge oil) and Diesel Fuel. Agricultural Engineering
International: The CIGR Ejournal. Manuscript number 1316. Vol. XI. May, 2009.
[76] Wang, Y. D., An experimental investigation of the performance and gaseous exhaust
emissions of a diesel engine using blends of a vegetable oil, Applied Thermal Engineering,
Volume: 26, 2006, Pages: 1684-1691.
[77] Yoshimoto, Y., Onodera, M., Tamaki, H. Performance and emission characteristics of
diesel engines fueled by vegetable oils, SAE Technical Paper Series , Paper No. 2001-011807/4227.
[78] Yaodong, W., Huang, Y., Roskilly, A. P., Ding, Y., Hewitt, N., Trigeneration running with
raw jatropha oil, Fuel Processing Technology, Volume 91, Issue 3, pp 348353, 2010.
[79] Yilmaz, N, Morton, B., Effects of preheating vegetable oils on performance and emission
characteristics of two diesel engines. Biomass Bioenerg., 35: 2028-2033.
[80] Ziejewski M, Goettler H and Pratt GL (1986) Influence of vegetable oil based alternative
fuels on residue deposits and components wear in a diesel engine. SAE. 860302, 297 307.
[81] Zimmerman, J., Brazil Oilseeds and Products Annual Oilseeds Report, USDA Foreign
Agricultural Service, 2011.
426
Abstract:
Wind energy conversion technology is by now fully developed on an industrial scale; commercial wind
turbines capacity factors have reached very good values and now the scientific community is engaged in
understanding in details all the phenomena which can affect the power performances of an aerogenerator.
Among them atmospheric stability is still not well investigat ed comparing to others environmental conditions;
this is partially due to the low incidence of non-neutral conditions on the actual productive periods.
Usually the wind energy assessment studies were generally performed referring to neutral stability; this was
considered acceptable because neutral conditions prevail on the high wind situations.
Anyway, especially on coastal and offshore sites, stability can induce meaningful effect on power production
both directly on the net power output and on the wakes.
The wind energy industry is now producing wind turbines with a high ratio of the rotor surface by the nominal
power; in this way producing energy even with low wind regimes and non-neutral conditions can involve
significant periods. In such situations the variations of the vertical wind shear can affect the energy
production and it could be fundamental to investigate how atmospheric stability can affect the overall power
conversion efficiency.
In present work the effect of atmospheric stability was investigated analysing the production data of a small
wind farm operating in flat terrain in southern Italy; in the site only two turbines with a very high ratio of rotor
surface by nominal power are operating under a low-medium wind regime.
Results demonstrate that atmospheric stability can have a meaningful impact on power production especially
for unstable conditions
Furthermore a good overall agreement was discovered between the results from the experimental dataset
and from numerical simulations of different thermal conditions through a CFD (Computational Fluid
Dynamics) code and the actuator disc model.
Keywords:
Wind Energy conversion, Atmospheric stability, CFD.
1. Introduction
This work deals with the production and wind data of a little wind farm located in the south of Italy;
power performances were analyzed in order to investigate the influence of thermal stratification on
the overall production. On the site, located in a quite flat land, there are two turbines with a specific
rated power of 1000 kW, a hub height of 65 m and a rotor diameter of 77 m. The site altitude is
between 40 and 55 m over the sea level. A 50 m anemometer is located on the site 315 m (4.06
diameters) south-east from turbine nr. 2 and 670 m (8.07 diameters) south-west from turbine nr. 1.
The distance between the turbines is 720 m (9.35 diameters). The relative locations are shown in
Fig. 1:
427
Fig. 1. Layout for the two wind turbines and the anemometer mast.
Experimental data from the SCADA (supervisory control and data acquisition) system was
available and analyzed for the first yearly period of operation for the two aerogenerators from
01/01/2010 to 31/12/2010; at the same time also measurements from the anemometer mast were
used. All the meaningful signals were aggregated on a synchronized dataset with 10 minutes
averaged values. Atmospheric stratification analysis was performed using the insolation, the
radiation and other meteorological data calculated by the numerical model WRF (Weather Research
and
Forecasting);
such
data
were
provided
by
LAMMA
laboratory
(http://www.lamma.rete.toscana.it). The WRF data are hourly averaged. Production, insolation and
anemometer data were synchronized to split them into stability classes.
The influence of thermal stratification was investigated using the following databases:
the production time history of the two aerogenerators (10 minutes averaged);
the wind database (10 minutes averaged) of the met mast placed in the wind farm area;
the hourly meteorological time history calculated by the WRF (Weather Research &
Forecasting) model.
All the sources of data were synchronized in order to select the production periods according to
different direction sectors and different thermal conditions; for the atmospheric stability analysis the
Pasquill classification was used.
In each different situation the upstream turbine production time history was used to reproduce the
wind turbine's power curve normalizing the reference wind speed according to the IEC standard.
Better results were obtained through the nacelle anemometry included in the production dataset.
Fig. 2. Wind rose measured at the anemometer position 50 m a.g.l. for the yearly period.
The most populated sectors are 300 and 330, but in these sectors the anemometer is downstream
the turbine nr. 2 and data from these sectors cannot be used for the power curve evaluation due to
wake effects. The third most populated sector is 90.In this sector turbine nr. 2 is in the wake of
turbine nr. 1; looking at production data an energy loss for turbine nr. 2 can be estimated for this
sector. As a consequence only the production time history of turbine nr. 1 in the 90 sector was used
for the power performance analysis.
Production data available from the SCADA were used to assess the power performance of the
turbines in different atmospheric conditions. From the overall dataset only the following parameters
were extensively analyzed for the present work:
the turbine power output (kW);
the nacelle anemometer wind speed (m/s);
the nacelle orientation ().
A lot of work was done to filter the dataset in order to exclude all invalid data; especially in the
early period of operation of the turbine there were some problems with the power control so that
production data were useless for the present study.
Another important issue was the choice of the wind speed reference value for the power curve
evaluation. IEC standards [1] recommend measuring the power curve using an anemometer mast
placed at a distance of 2.5 rotor diameters (D) from the turbine in the upstream direction.
The nearest turbine to the anemometer mast on the site under investigation is placed 4.06 D (a bit
more of the maximum allowable distance of 4 D) and looking at the wind rose of figure 2 the
anemometer is frequently affected by the wake of such turbine.
The only meaningful direction sector for which the anemometer is not affected by turbines wakes
is 90; for this direction the upstream undisturbed turbine is placed 8.07 D from the mast.
Plotting the power output of turbine nr. 1 versus the anemometer wind speed for the 90 sector
(figure 3) reveals that, in this way, the power curve of the aerogenerator cannot be assessed.
429
Fig. 3. Power output for turbine nr. 1 versus the wind speed at the anemometer position (50 m
a.g.l.).
The power curve of a wind turbine can also be estimated using the nacelle anemometry; this
procedure cannot be used when measuring the power curve for certification purpose but can be very
useful when assessing the power performance of the machine in different weather conditions or in
different time periods.
Fig. 4. Power output for turbine nr. 1 versus the nacelle wind speed.
Comparing figure nr. 3 and nr. 4 it is possible to assess the improvement of the power curve
representation when using nacelle anemometry; for this reason in present work such technique was
applied and data from anemometer mast were retained to assess the boundary layer conditions and
to define the wind direction. The wind direction could have been also estimated using the nacelle
position from the SCADA dataset; anyway direction data from the anemometer vane were
discovered to be more reliable and representative.
430
Following the IEC standard [1] the production data need to be normalized to the standard air density
conditions. For this purpose the 10 minutes averaged air density was calculated in each time steps
using the following formulation:
(1)
where:
For modern wind turbines with a full-pitch control system the power curve referred to the standard
air density can be evaluated using the normalized wind speed:
(2)
where
The production dataset was also filtered to exclude malfunction periods; to do this also the
manufacturer theoretical power curve was used to reproduce the expected power history calculated
from the nacelle wind speed. If the ratio between the expected power and the actual power was
outside the range between 0.8 to 1.2 for the upstream turbine the data of the specific timestep were
disregarded. In figure nr. 5 it is possible to verify how the experimental power curve is very close to
the expected one.
Fig. 5. Comparison of the manufacturers power curve and the experimental one for the direction
sectors from east (0-180) for turbine nr. 1.
431
Analyzing all the hourly 2010 WRF dataset calculated for the nearest node it was possible, for each
timestep, to give an estimation of the Pasquill class.
432
The distribution of such calculus is shown in table 2 from which it is possible to assess how the
neutral conditions prevails on stable and unstable conditions.
Table 2. Results for the stability classification of the site.
Atmospheric conditions frequency %
NEUTRAL
41
STABLE
32
UNSTABLE
27
These results were used to characterize the power performances of turbine nr. 1 in the three
different stability conditions; the Pasquill classes were aggregated in only three groups in order to
reach a good power curve data population.
Atmospheric stability was also considered, in an easier way, in the numerical model of the wind
farm.
The initialization of thermal conditions within the numerical code can be done in different way: in
present work a rough comparison of the numerical and the experimental power performances of the
wind farm was done using the initialization with the Monin-Obukhov length L. In order to obtain
good results with a full convergence of the code only slightly stable and unstable conditions were
simulated. The Monin-Obukhov length can be estimated from the surface roughness length Z 0 [3]
and the Pasquill classes using the following formula:
1
b
aZ 0
(3)
L
where a and b are coefficients that can be estimated from the following table:
Table 3. Different values for a and b coefficients.
Pasquill Class
a
-0.0875
A
-0.03849
B
-0.00807
C
0
D
0.00807
E
0.03849
F
b
-0.1029
-0.1714
-0.3049
0
-0.3049
-0.1714
433
that is
(5)
where A is the rotor area.
The 1D momentum theory also gives the following equation relating the thrust coefficient with the
axial induction factor:
(6)
The numerical model was used to simulate the wind field and finally calculate the power output P
multiplying the pressure drop
across the rotor and the local axial component of the wind speed
on the disc. Such calculation was done in the post processing approximating on the discrete
domain the following integral formulation:
(7)
Fig. 6. Calculated wind field (Speed 2D) at the turbines hub height (65 m a.g.l.) for the 90 sector
in neutral condition.
In figure 6 the simulated wind field is shown for the neutral condition case.
The CFD model domain is approximately 5x3 kilometres with a grid horizontal resolution varying
from 6 up to 120 m; the overall number of cells is 331958. Simulations for neutral and stable
conditions were performed using the standard k- turbulence model while the for unstable
conditions the RNG k- model was used in order to achieve a better convergence.
The boundary conditions were tuned to reach quite similar conditions of wind speed on the rotor for
the three thermal conditions (figure 7).
434
(a)
(b)
Fig. 7. Calculated wind field (Speed 2D) at the turbines hub height (65 m a.g.l.) for the 90 sector
in unstable (a) and stable (b) condition.
3. Results
Using the stability classes time history obtained through the WRF model dataset it was possible to
split, for turbine nr. 1 in the eastern sectors (0-180), the power curve obtained in neutral, stable and
unstable conditions. Unfortunately the filtered database was poor especially for stable conditions
events and an extensive analysis of the power curve was not possible. Moreover, as can be deduced
from table 1, non neutral conditions occurs especially in low-wind regimes so that a reliable
comparison between the power performances in the three different stability conditions was possible
only in the early part of the power curve (from 3 to 6.5 m/s) and a comparison for all the analyzed
dataset is quite useless (figure 8).
More interesting information can be found looking at the power curve obtained calculating the
mean values within bin intervals of 0.5 m/s width for wind speeds from 3 to 6.5 m/s (figure 9).
435
Fig. 9. Comparison of the binned power curves for turbine nr. 1 in the eastern sectors (0-180)
for different atmospheric stability conditions.
Unfortunately the poor quality of the available experimental dataset allowed only a weak
comparison of the conversion efficiency in the different thermal conditions.
In figure 10 it is represented the standard deviation of the power output (expressed in percentage of
the mean value) and the maximum difference observed between the productions in the three
different thermal conditions (expressed in percentage of the mean value as well).
The maximum difference is higher and comparable with the standard deviation only for a wind
regime of 9 m/s; in such conditions the possible intervention of the turbine control make the
comparison not significant.
Fig. 10. Comparison of the standard deviation of power output with the maximum difference
observed for the analysed thermal conditions.
436
Anyway the mean values of figure 9 (even if affected by high uncertainties) give a first estimation
of how conversion efficiency can be affected by thermal stratification: in unstable conditions the
turbine seems to work with a lower power even with the same normalized nacelle wind speed. This
is consistent with some previous numerical investigations [8] and can be addressed to the poorest
energy content of the wind profile.
Using the anemometer dataset a different coefficient of the power law profile was estimated in the
three thermal conditions: 0.167 in the neutral case, 0.191 in the stable case and 0.0827 in the
unstable case.
Results from the power output estimation in the numerical model for the same turbine nr. 1 are in
good agreement with the experimental results. The numerical model solves the wind field using the
Reynolds Average method (RANS solution) and a comparison with the experimental data can be
done mainly in a qualitative way. The numerical values of the calculated power output in the three
stability conditions were used to estimate the power coefficient defined as:
(8)
where the undisturbed wind speed was calculated using the axial induction factor that can be
estimated through the wind speed on the disc and the thrust coefficient curve.
At the same time the power coefficient can also be estimated from the axial induction factor using
the formula from the 1D momentum theory:
(9)
This is the theoretical value of the power coefficient that derives from 1D momentum theory and is
generally higher than the numerical power coefficients due to three-dimensional effects and to the
boundary layer effects.
Table 4. Theoretical and numerical power coefficients..
Neutral
Stable
Unstable
Power Coefficients
1 D Momentum Theory
0.5786079
0.5786079
0.5815354
Numerical
0.4273396
0.4602921
0.4254921
In the numerical wind field both the upstream wind speed as well as the rotor wind speed can be
estimated so that the value of the axial induction factor can be calculated. Using that value and
equation 9 an estimation of the power coefficient according to the 1D momentum theory is possible.
At the same time the numerical pressure drop on the rotor can be calculated and, using equation 7, a
second estimation (numerical) of the power output can be extrapolated; in this second estimations
also the three-dimensional effect are numerically considered and the obtained values are
comparable with the actual power coefficient of a real operating turbine,
As can be deduced from table 4 in the three different atmospheric stability conditions the gap
between the numerical and theoretical power coefficients can vary due to the influence of the
boundary layer conditions.
In unstable conditions the gap is higher and the power performances are poor while the power
efficiency is higher in neutral and stable conditions.
437
3. Conclusions
In present work an analysis of the impact of atmospheric stability on the wind energy conversion
efficiency was done through numerical and experimental investigations. Results demonstrate that
poor efficiency is to be addressed especially to unstable thermal stratification.
An improvement of the numerical model can give very useful information in order to optimize the
overall production and the turbine lifetime in wind sites where non neutral conditions can involve
relevant operational periods.
Acknowledgments
The authors acknowledge the Italian company Calypso Engineering spa who provided all data
treated and analyzed in present work.
Nomenclature
calculated air density averaged over 10,
air absolute temperature averaged over 10,
air pressure averaged over 10 min, Pa
R
gas constant,
normalized wind speed,
10 averaged speed,
standard air density,
RT
global radiation,
RN
net radiation,
L
Z0
Monin-Obukhov length,
surface roughness length,
wind speed on the rotor,
local axial component of the wind speed on the rotor,
undisturbed speed,
axial induction factor
power coefficient
References
[1] IEC 61400-12 WIND TURBINES Part 12-1: Power performance of electricity producing
wind turbines, 2006
[2] A. Venkatram, 1996, An examination of the Pasquill- Gifford- Turner dispersion scheme.
Atmospheric. Environment, 30:1283-1290
[3] A. Capanni, G. Gualtieri, 1999: Sodar application for estimating boundary layer parameters.
In: Proc. EUROPTO series: Optics in atmospheric propagation and adaptive systems III.
Florence, Italy, Sept., 23-24.
[4] Van Kuik GAM., 1991, "On the limitations of Froudes actuator disc concept." PhD Thesis,
Technical University of Eindhoven Doctoral Thesis, 1991.
[5] Vermeer L.J.,Srensen J.A., Crespo A., 2003, Wind turbine wake aerodynamics. Progress
in Aerospace Sciences 39 (2003) 467510
[6] B. Sanderse - Aerodynamics of wind turbine wakes Literature review, Energy researc h
centre of the Netherlands http://www.ecn.nl/docs/library/report/2009/e09016.pdf
[7] Rthor P.E.,Srensen N. N., Zahle F. "Validation of an Actuator Disc Model." - Europea n
Wind Energy Conference & Exhibition Proceedings - Tuesday 20 - Friday 23 April 2010,
Warsaw, Poland
[8] Castellani F., Vignaroli A., Piccioni E. - "Wind shear investigation for site-specific wind
turbine performance assessment" - Journal of Energy and Power Engineering ISSN 1934-8975,
USA
439
School of Energy, Power and Mechanical Engineering, Beijing Key lab of Energy Safety and Clean Utilization,
Key Laboratory of Condition Monitoring and Control for Power Plant Equipment of Ministry of Education,
North China Electric Power University, Beijing, 102206, China,
E-mail: dlq@ncepu.edu.cn
Abstract:
Based on a traditional SOFC (Solid Oxide Fuel Cell) hybrid power system, three different SOFC hybrid
power systems with zero-CO2 emission are proposed in this paper and their performances are analyzed and
compared. The exhausted gas from the anode of SOFC is burned with pure oxygen and the concentration of
CO2 gas is greatly increased. Because the combustion produce gas is only composed of CO2 and H2O , the
separation of CO2 hardly consume any energy. At the same time, three different methods (steam injection,
CO2 gas injection, heat exchange layout) are taken to maintain the proper turbine inlet temperature. With the
exergy analysis method, this paper studied the exergy loss distribution of every unit of SOFC hybrid systems
with CO2 capture and revealed the rules of exergy loss distribution. The effects of main operating
parameters on the overall SOFC hybrid system with CO2 capture are also investigated.
The research results show that the zero-CO2 SOFC hybrid systems proposed in this paper still have higher
efficiencies when compared with the SOFC hybrid system without CO2 capture. Their efficiencies only
decreases about 3-4 percentage points compared with the basic SOFC hybrid system without CO2 capture.
The research achievements obtained from this paper will provide the new idea and method for further study
on zero emission CO2 system with high efficiency.
Keywords:
SOFC, Hybrid power system, Zero CO2 emission, Steam injection, Exergy analysis.
1. Introduction
Now, climate change due to the emission of greenhouse gases, especially the emission of CO2, is
becoming more and more serious. Though many countries have taken some measures to control and
reduce the emission of CO2, in the short term, CO2 emission is still maintaining the rapid growth
trend. Power industry is the biggest CO2 emission sector. So, there exists the greatest CO2 emission
reduction potential in the power industry [1]. Now, many kinds of fossil fuel power generation
systems with CO2 recovery are usually based on the chemical absorption method or the oxygen
combustion method. The former demands a chemical absorption and separation unit to recover CO2
from the exhaust combustion gas. The latter demands the special oxygen combustion technology,
the special equipment and larger ASU (air separation unit). And these technologies all consume
great energy and result in the huge equipment investment and higher operating costs [2]. Now,
people are eager to develop the high-efficiency power generation technology with the less energy
consumption for CO2 capture. Fuel cell can satisfy the above requirements, with the higher energy
conversion efficiency and less CO2 capture energy consumption, so it has attracted considerable
interest in recent years.
Solid Oxide Fuel Cell (SOFC) is an attractive power-generation technology that can convert the
chemical energy of fuel directly into electricity while causing little pollution. Because the anode
fuel gas is naturally separated from the cathode air by the solid electrolyte, the CO2 gas with the
higher concentration can be obtained in the anode exhaust gas. In addition, SOFC can employ all
kinds of fuels, including various hydrocarbon fuels. Compared with the traditional power generation
system, the SOFC hybrid system power plant has the higher overall system efficiency (LHV). Even
440
after CO2 is captured, the efficiency of SOFC hybrid system still can be higher or equal to that of
the traditional power system without CO2 capture. In order to further improve the CO2
concentration of anode exhaust gas, SOFC can employ the O2/CO2 combustion mode in the
afterburner. Because the required mass flow of pure O2 is less, the energy consumption is lower.
After capturing the CO2, the SOFC hybrid system does not result in the bigger efficiency reduce.
So the SOFC hybrid power system with zero CO2 emission becomes a new way which can
simultaneously solve the problem of efficient energy utilization and lower pollution emission.
In the last decades, many researchers were involved in study of SOFC stack and the hybrid power
system with CO2 capture. Y.Inui proposed and investigated two types of carbon dioxide recovering
SOFC/GT combined power generation systems in which a gas turbine with carbon dioxide recycle
or water vapor injection is adopted at the bottoming cycle system [2]. The overall efficiency of the
system with carbon dioxide recycle reaches 63.87% (HHV) or 70.88% (LHV), and that of the
system with water vapor injection reaches 65% (HHV) or 72.13% (LHV). A. Franzoni considered
two different technologies for the same base system to obtain a low CO2 emission plant [3]. The
first technology employed a fuel decarbonization and CO2 separation process placed before the
system feed, while the second integrated the CO2 separation and the energy cycle. The result
showed that the thermodynamic and economic impact of the adoption of zero emission cycle
layouts based on hybrid systems was relevant. Philippe Mathieu presented the integration of a solid
oxide fuel cell operating at a high temperature (9001000 , 5560% efficiency) in a near-zero
emission O2/ CO2 cycle [4]. Takeshi Kuramochi compared and evaluated the techno-economic
performance of CO2 capture from industrial SOFC-combined heat and power plant (CHP) [5]. CO2
is captured by using an oxyfuel afterburner and conventional air separation technology. The results
were compared to both SOFC-CHP plants without CO2 capture and conventional gas engines CHP
without CO2 capture. B.Fredriksson Moller examined the SOFC/GT configuration with and without
a tail-end CO2 separation plant and based on a genetic algorithm, selected the key parameters of the
hybirid system[6]. The result of the optimization procedure shows that the SOFC/GT system with
part capture of the CO2 exhibits an electrical efficiency above 60%. Some researchers also studied
the performance parameters of the different SOFC hybrid power systems from the thermoeconomic
or exergy efficiency point [7-9]. For example, Ali Volkan Akkaya proposed a new criterionexergetic performance coefficient (EPC), then applied it in the SOFC stack and SOFC/GT CHP
system[10-11]. F. Calisa discussed the simulation and exergy analysis of a hybrid SOFC-GT power
system. The result showed that the SOFC stack was the most important sources of exergy
destruction [12].
In this paper, three kinds of zero-CO2 emission SOFC hybrid power systems are proposed. Using
exergy analysis method, the exergy loss distributions of every unit of zero-CO2 emission SOFC
hybrid system are revealed. The effects of different operating parameters on exergy loss of every
unit, as well as the overall system performance, are also investigated. The results obtained in this
paper provide useful reference for further study on high-efficient zero emission CO2 power system.
76 log(
p
)
pref
0.008 T Tref
Vcat (mV )
(1)
ic
(2)
p H 2 p H2 O
(3)
( pH 2 pH 2O )ref
92 log(
p O2
( pO 2 ) ref
442
(4)
Where ic is the current density; p H 2 / p H 2 O is the ratio of H2 and steam partial pressure;
Po 2
is the
average oxygen partial pressure at the cathode for the actual case; Vc is the actual cell voltage; the
subscribe ref is the reference case.
Pref = 1 bar ; T ref = 1000 ; ( pO2 )ref = 0.164; ( pH 2 / pH 2 O )ref = 0.15.
The current density (ic) is calculated by the following equation:
ic
zne F
NA
(5)
Where A is the active area of single cell, cm2 , F is Faraday constant, 96487 C/mol , ne is electrons
transferred per reaction , N is Number of single cells , z is the mole flow of reacted H 2;
By summing the above five equations (1)-(5) , the actual voltage Vc can be calculated as:
Vc Vref
Vp
VT
Vcat
Van
The fuel cell power output is the product of the cell voltage and current.
Power
ic ( A) Vc (V )
(6)
(7)
(8)
Where G f is the flow rate of fuel, mol/s, LHVf is the low heating value of fuel, kJ/mol;
In the process of modelling the SOFC hybrid power system, the accuracy of the fuel cell stack
model is critical for the overall system. So this paper firstly checked the accuracy of SOFC stack
model. With the same input parameters of the literature [17-20], the simulation result of this paper is
compared with those of the literature [17-20] (as shown in Table 1). The results show that this papers
simulation results are very close to the literature results which are close to the experimental data [1820].
So the model of SOFC stack is feasible and reliable, it can be applied to simulate the overall
SOFC hybrid system.
Table 1 The Simulation Results of SOFC Stack 120kW DC output
Literature
Simulation data of this
Parameter (unit)
Literature Data [17]
Data [18-20]
paper
Voltage V
0.70
0.70
Current Density
178
180
179.1
(mA/cm2
Air utilization factor
19%
18.2%
Pre-reformer Outlet
536
550
537.2
Temperature
77%N2
77.3%N2 15.9%O2
Stack exhaust composition
77.5%N2 16.4%O2
16%O2
4.5%H2O
4 %H2O 2.1%CO2
5%H2O
EXHAUST
2.3%CO2
2%CO2
48% H2O
50.9% H2O
50.9% H2O
28%CO2
Anode outlet composition
24.9% CO2
24.9%CO2
14%H2
11.6%H2 7.4%CO
11.6%H2 7.4%CO
stream5
5%CO
5.1%N2
5.1%N2
5%N2
Stack exhaust
834
847
832
temperature
Stack efficiency AC
52%
50%
51.99%
443
2.3. Description of the basic SOFC hybrid system without CO2 capture
The basic SOFC hybrid system is a tubular SOFC stack with higher operating temperature. Figure 2
shows the flowsheet of the system. The fuel is compressed and preheated, then is put into prereformer to generate the required H2, CO and CO2. Air is supplied by a blower and preheated prior
to entering the SOFC stack. Then air participates an electrochemical reaction with fuel in the fuel
cell stack. Because the pre-reformer needs a larger amount of water vapor, a hot recycled stream
from the anode outlet is directed to the pre-reformer. The outlet stream of SOFC anode is mainly
composed of H2O, CO2 and unconverted fuel (H2 and CO). Part of this stream is injected into the
after-burner, then the hot outlet gas with high pressure expands in gas turbine, the rest is recycled
and mixed with the compressed and preheated fuel. The recirculation fraction is calculated to meet a
given steam/carbon (S/C) ratio. In this way, the system can prevent from the carbon deposition
phenomenon, enhance the pre-reformer temperature and get more H 2. Finally, the product gas of
SOFC is sent into the after-burner, where the unreacted fuel is burnt with part of the excess air from
the cathode. After expansion of the hot gas in gas turbine, it is sent into heat exchangers to heat the
inputted fuel and air. Then, it enters into the heat recovery steam generator and steam turbine
system to produce addition power.
Although SOFC hybrid power system has realized energy conversion with higher efficiency, the
hybrid system still emits greenhouse gas when the fossil fuel is used. In addition, the exhausted gas
from the afterburner contains a large number of N2 that will greatly impact the energy consumption
of capturing CO2. The paper proposed three kinds of zero CO2 emission SOFC hybrid power
systems and made the detailed exergy analysis on them.
444
3.1. Zero CO2 emissions SOFC hybrid power systems with steam
injection (case a)
The system flowchart is shown in Figure 3. The compressed and preheated nature gas and air are
sent into the SOFC anode and cathode, respectively. Then the electrochemical reaction takes place
at the three-phase boundaries of both electrodes to generate electricity. The depleted air out from
the cathode expands in the turbine and finally is channelled to HRSG to produce steam. The steam
produced from the heater4 is injected into the oxy-fuel burner to control temperature not exceed the
maximum allowed turbine inlet temperature. The exhausted flue gas from the afterburner firstly
expand in turbine1, then enters into a serial of heaters to exchange heat with flue and air, finally, the
exhausted flue gas is introduced into heater4 to produce steam for the afterburner. The flue gas from
the heater4 enters into the CO2 compression and liquefaction unit. The cathode outlet gas firstly
expands in turbine2, and then enters into HRSG and ST system to produce additional electricity.
3.2. Zero CO2 emissions SOFC hybrid power system with CO2 gas
injection (case b)
The system flowchart of case b is shown in Figure 4. Compared with case a, the main difference is
that CO2 gas, not steam, is injected into the oxy-fuel burner, which can further increase the
concentration of CO2. After the heat transfer with the fuel and air through a series of heat
exchangers, the combustion product gas is channelled into the HRSG to produce steam to generate
more power. Then, the flue gas (mainly CO 2 and steam) exhausted from the HRSG is split into two
parts. One part is compressed and injected into the O2 /CO2 afterburner, while the other part is
exhausted from the system and liquefied.
445
E X D ,tot
100%
9
It means the ratio of local exergy destruction to total exergy destruction. This coefficient can be
used to show the distribution of each units exergy destruction, which will help to assess the energysaving potentials of the systems. To highlight exergy variation rule of the system with the change of
key parameters, exergy destruction coefficient Ra as another evaluation index is defined as follow:
Ra
E X D . tot
E X tot
100%
10
It means the ratio of total exergy destruction ( E X D ,tot ) to the output exergy of the system ( E X tot ). For
a given system, the smaller Ra is , the smaller the exergy destruction is , the better the system with
more output exergy is .
2.5
85
15
910
5
of
80
99
448
Case c
6.06
4.97
3.55
0.635
177.31
141.96
44.78
31.21
21.44
3.296
5.217
187.99
47.65
63.10
Three different zero CO2 emission SOFC hybrid power systems are about 3-4 percentage points
lower than the basic SOFC hybrid power system without CO2 capture. The differences of output
power and power consumption in other units lead to disparity of system efficiencies among three
systems. The efficiency of case b is the highest as shown in table 4.
The comparison results show that turbine power of Case c has the lowest power output than those of
other systems because of the mass rate of turbines input gas (cathode outlet gas and combustion
product gas), other reason for it is that the temperature of turbine 1 inlet gas is reduced after
transferring heat in heater 4 and heater 2.
After exchanging heat with the fuel and air through a series of heat exchanger, the combustion
product gas has higher temperature to produce steam to generate more power in the HRSG in case
c, so case c has the biggest power output from HSRG among all three systems.
449
Before injecting the H 2O or CO2 gas into the oxy-fuel burner, the gas must be compressed firstly
with additional equipment and power consumption. For case a, pump is used to enhance the
pressure of H2O. For case b, gas compressor is employed to compress the injected CO 2 gas. It is
clear that the power consumption of case b for CO2 capture is higher than that of case a and case c.
In a word, the different performances of the system configurations bring about different system
efficiency among three systems. The system efficiency of case b is 2.7 percentage points lower than
that of the traditional SOFC hybrid power system and 1.25% and 0.22% higher than that of case a
and case c, respectively.
According to the simulation result and exergy balance equation, it is possible to obtain the exergy
destruction distribution in each unit of systems as shown in Fig.6. Results show that SOFC stack,
afterburner and HSRG are responsible for the major exergy destruction in the system. In order to
improve the system exergy efficiency, the exergy destructions of these units should be firstly
considered and further reduced.
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Case a
Case b
Case c
450
451
Figure 9 Effect of the operating temperature on system total power and exergy destruction (case a)
452
Figure 11 Effect of the fuel utilization factor (Uf) on system exergy destruction
6 Conclusions
This paper proposed three kinds of SOFC hybrid systems with CO 2 capture. The research results
show that the new SOFC hybrid system still has a higher efficiency even with the CO2 capture.
Their efficiencies only decrease 3-4 percentage points compared with the basic SOFC hybrid
system without CO2 capture. The system efficiency of case b is 1.25 and 0.22 percentage points
higher than that of case a and case c respectively.
The O2/CO 2 combustion mode can fully burn the anodes fuel gas, and increase the concentration of
CO2 gas; at the same time with the steam injection and the combustion products are channelled into
turbine, the efficiency of system greatly increases. The liquefaction of CO 2 by the mode of multistage compression and intermediate cooling can also greatly reduce the energy consumption.
Exergy analysis of SOFC hybrid power systems provides better understanding of losses and draws
ones attention to the most important sections. It is clear that SOFC stack, afterburner and the HSRG
are responsible for the major exergy destruction in the system. Consequently, some appropriate
measures should be adopted to lower exergy destruction to enhance the overall system efficiency.
The operating temperature, the operating pressure and the fuel utilization factor as key variables
influence the overall system performance greatly, which should be taken into account for the
systems to achieve the optimal condition. Above research achievements will provide the new idea
and method for further study on zero emission CO2 system with high efficiency.
453
Acknowledgments
This study has been supported by the Fundamental Research Funds for the Central Universities
(No.10ZG03), National Nature Science Foundation Project (No. 50606010 and No. 51025624), the
National Basic Research Program of China (No. 2009CB219801), the Scientific Research
Foundation for the Returned Overseas Chinese Scholars, State Education Ministry and by the 111
Project B12034 .
References
[1] Kartha S, Grimes P. Fuel cells: energy conversion for the next century. Physics Today, 1994
(11) : 5461
[2] Y.Inui, T.Matsumae, H.Koga, K.Nishiura. High performance SOFC/GT combined power
generation system with CO2 recovery by oxygen combustion method. Energy Conversion and
Management, 2005, 46, pp: 18371847
[3] A. Franzoni, L. Magistri, A. Traverso, A.F. Massardo. Thermoeconomic analysis of pressurized
hybrid SOFC systems with CO2 separation. Energy,2008, 33, pp: 311320
[4] Philippe Mathieu. Towards the hydrogen era using near-zero CO2 emissions energy systems.
Energy, 2004, 29 , pp: 19932002
[5] Takeshi Kuramochi, Hao Wu, Andrea Ramirez, Andre Faaij and Wim Turkenburg. Technoeconomic prospects for CO2 capture from a Solid Oxide Fuel Cell Combined Heat and Power
plant. Preliminary results. Energy Procedia, 2009,1, pp: 38433850
[6] B. Fredriksson Mller,, J. Arriagada, M. Assadi, I. Potts. Optimisation of an SOFC/GT system
with CO2 -capture. Journal of Power Sources, 2004, 131, pp: 320326
[7] Bozzolo M, Brandani M, Traverso A, Massardo AF. Thermoeconomic analysis of gas turbine
plants with fuel decarbonization and carbon dioxide sequestration, ASME transactions. J Eng
Gas Turbines Power, 2003, 125 (4) 94753.
[8] Asle Lygre Matteo Ce. Solid Oxide Fuel Cell Power with Integrated CO2 capture. Second
Nordic Minisymposium on Carbon Dioxide Capture and Storage, Goteborg, October 26, 2001
[9] Takuto Araki, Takuya Taniuchi, Daisuke Daisuke Sunakawa. Cycle analysis of low and high H2
utilization SOFC/gas turbine combined cycle for CO2 recovery, Journal of Power Sources. 2007,
171, pp: 464470
[10]Ali Volkan Akkaya, Bahri Sahin , Hasan Huseyin Erdem. Thermodynamic model for exergetic
performance of a tubular SOFC module. Renewable Energy, Volume 34, Issue 7, July 2009, pp:
18631870
[11]Ali Volkan Akkaya, Bahri Sahin, Hasan Huseyin Erdem. Exergetic performance coefficient
analysis of a simple fuel cell system. International Journal of Hydrogen Energy, 2007, 32,
pp:4600 4609
[12] F. Calisea, M. Dentice dAccadiaa, A. Palomboa, L. Vanoli. Simulation and exergy analysis of
a hybrid Solid Oxide Fuel Cell (SOFC)Gas Turbine System. Energy, 2006, 31, pp: 32783299
[13] Akkaya AV, et al. Exergetic performance coeffcient analysis of a simple fuel cell system.
International Journal of Hydrogen energy, 2007: 110
[14] J.H. Hirschenhofer, D.B. Stauffer, R.R. Engleman, Fuel Cells, A Handbook (Rev.3),
Commonwealth Inc. for US Department of Energy (DOE) , Gilbert , 1994.
[15] H. Ide, et al., Natural gas reformed fuel cell power generation systemsa comparison of three
system efficiencies, Proceedings of the 24th IECEC , Washington , 1989.
[16] N.F. Bessette, R.A. George , Electrical performance of Westinghouse's AES solid oxide fuel
cell, Proceedings of the Second International Fuel Cell Conference , Japan , 1996.
[17]W. Zhang, E. Croiset, P.L.Douglas, M.W.Fowler, E.Entchev. Simulation of a tubular solid oxide
fuel cell stack using AspenPlusTM unit operation models. Energy Conversion and Management,
2005,(46): 181-196.
454
[18]Veyo SE, Forbes CA. Demonstractions based on Westinghouses prototype commercial AES
design. In: Proceedings of the Third European Solid Oxide Fuel Cell Forum, 1998, pp: 79-86.
[19]Veyo SE. The Westinghouse solid oxide fuel cell programea status report. In: Proceedings of
the 31st IECEC, No. 96570, 1996, p. 1138-43.
[20]Veyo S, Lundberg W. Solid oxide fuel cell power system cycles. ASME Paper 99-GT-356,
International Gas Turbine and Aeroengine Congress and Exhibition, Indianapolis, June 1999
[21]Wang Jiaxuan, Zhang Shufang. Exergy method and its application in power plants, China
Water Power Press, 1993.87121
[22]Lu Lining. Calculation and optimization study on the performance of a Combined Power
Generation System Incorporating a Solid-oxide Fuel Cell and a Gas Turbine. Masters thesis,
Dalian University of Technology, 2004
[23]JIA Junxi. Study of Model Heat and Mass Transfer in Solid Oxide Fuel Cell, Doctors thesis,
Dalian University of Technology, 2006
[24]Calise F, d,Accadia MD, Palombo A, Vanoli L. Simulation and exergy analysis of a hybrid
solid oxide fuel cell (SOFC) gas turbine system. Energy, 2006, 31, pp:7899
[25]Meng Ni, Michael K.H. Leung, Dennis Y.C. Leung. Parametric study of solid oxide fuel cell
performance. Energy Conversion and Management, 2007, 48, pp:15251535
[26]Li Xiaolin. Study on heat/power ratio of SOFC-GT hybrid system, Masters thesis, Central
South University. 2008
455
Abstract:
In this paper we present the exergetic analysis and optimisation of a steam methane pre-reformer system for
marine fuel cells. Hybrid marine energy systems combining conventional power generation with fuel cells can
offer significant reductions in emissions and overall efficiency gains, promote the shift to more
environmentally-friendly fuels, such as LNG, and can increase the operational flexibility of a vessel. High
temperature molten carbonate fuel cells (MCFC) is a promising option for marine use due to their high
efficiency and high energy content exhaust gasses. A key component of this technology is the steam
methane pre-reformer which reforms completely higher-chain hydrocarbons contained in LNG and partly
reforms methane to hydrogen in order to assist the internal reforming process within the fuel cell stack. We
consider a pre-reformer system concept consisting of a water evaporator, a steam/methane heater and a
catalytic tubular packed-bed reformer. The water evaporation and steam/methane heat exchangers utilise
the high temperature exhaust gases of the fuel cell stack, while the pre-reformer uses the heat input of the
process stream. The pre-reformer system uses about 39% of the exhaust gas exergy, which amounts to
about 10% of the chemical exergy of the overall systems fuel. Therefore, there is considerable potential for
improving the design in terms of energy efficiency that will achieve higher performance and applicability for
the overall fuel cell unit.
In this work we present a three-step approach. First, we develop a dynamic mathematical model that
describes the physical/chemical behaviour of the pre-reformer system. Second, we perform exergy analysis,
and third, we optimise the reformer with respect to its exergetic performance. Our methodology is based on
the thermofluid and chemical reactions modelling of the system components, using a generic and
reconfigurable process modelling framework. The developed models are spatially distributed in order to
account for, and give insight on, the internal process characteristics of each individual component. In
addition, spatially distributed exergy balances are considered within the component models, to capture the
interrelation of the local exergy destruction (irreversibility) with component design geometry, configuration
and process phenomena. A design optimisation problem is then formulated that minimises the total
irreversibility of the steam methane reforming system, subject to design, technical, operational and space
constraints.
Our approach was applied to a standard marine MCFC pre-reformer system and the exergy analysis and
optimisation results have successfully yielded the sources of irreversibility, and provided a low irreversibility
optimal design with more than 50% less exergy destruction. The exergy analysis and optimisation proved to
be particularly suited for this chemically reacting and heat exchanging system since traditional energy (firstlaw) and pinch-point calculations have certain shortcomings in their usage, which often fail to reveal the full
picture of losses within the system. The results will be subsequently used to improve an existing marine
system design and serve as the basis for the integrated optimisation of the entire fuel cell-based unit.
Keywords:
Marine fuel cells, Steam methane reformer, Exergy analysis, Modelling, Optimisation.
1. Introduction
The rising fuel costs, increased environmental concerns and forthcoming emissions regulations
impose a pressure on ship energy conversion systems to become more efficient, cost-effective and
environmentally friendly. In that context, hybrid marine energy systems combining conventional
power generation with fuel cells emerge as promising future solutions offering reductions in
456
emissions and overall efficiency gains, promoting the shift to more environmentally-friendly fuels,
such as LNG, and increasing the operational flexibility of a vessel. High temperature molten
carbonate fuel cells (MCFC) is a promising option for marine use due to their high efficiency and
high energy content exhaust gasses. This technology has already been installed onboard the offshore
supply vessel Viking Lady (Eidesvik Offshore ASA) as an auxiliary power unit during the
research demonstration project FellowSHIP [1, 2]. A key component of the MCFC plant is the
steam methane pre-reformer (SMR) which is used to completely reform higher-chain hydrocarbons
contained in the liquefied natural gas (LNG) fuel of the MCFC and partly reform methane to
hydrogen in order to assist the internal reforming process within the fuel cell stack.
Steam methane reformers (SMR) are widely used for hydrogen production from hydrocarbon and
alcohol fuels, like natural gas and methanol [3, 4]. In this work we present the modelling, exergetic
analysis and optimisation of a SMR fuel pre-processing system for marine MCFC modules. The
purpose of this study is to gain insight on the internal process characteristics of the SMR system
components, identify the sources of exergy losses and to optimise the system with respect to its
exergetic performance. Exergy analysis and optimisation is particularly suited for this chemically
reacting and heat exchanging system since traditional energy (first-law) and pinch-point
calculations have certain shortcomings in their usage, often failing to reveal the full picture of losses
within such systems [5-8]. In previous studies, exergy analysis approaches have been successfully
used to assess and improve fuel cell systems [7-16] and steam methane reformers [6, 17, 18]. This
work combines exergy analysis and optimisation at a system level [6, 17] with spatially distributed
exergy balances within components [7, 12, 18] in order to investigate the interrelation of the local
exergy destruction (irreversibility) with the component design geometry, configuration and process
phenomena. Subsequently, a design optimisation problem is formulated that minimises the total
irreversibility of the SMR system, subject to design, technical, operational and space constraints.
The following sections present the SMR system description, the mathematical formulation of the
individual component models and the optimisation problem formulation. Finally, the results of the
exergy analysis and optimisation are given and discussed.
2. System description
We consider a marine MCFC unit with a MCFC stack and a SMR sub-system [2], Fig. 1. The
MCFC stack is a low-pressure module with direct internal reforming (DIR). The SMR system
mainly consists of a cross flow water evaporator, a cross flow steam/methane heater and an
adiabatic catalytic tubular packed-bed reformer. These components constitute a fuel pre-processing
/ reforming module with main mission the complete reform of higher-chain hydrocarbons contained
in the natural gas and the partial reforming of methane to hydrogen in order to produce the suitable
operating conditions for the internal reformers of the MCFC-DIR stack. This is the main difference
of the SMR of Fig. 1 with SMR concepts examined in other studies [6, 17], where the reformers are
used for the bulk production of hydrogen. Literature results indicate that this MCFC-DIR concept
has an improved exergetic performance compared to the complete external reforming solutions [4,
19].
The heat exchange process takes place on the water evaporator and the steam/methane heater,
utilising the high temperature exhaust gases of the fuel cell stack. These were selected to be of a
cross flow arrangement in order to minimise their footprint and simplify the layout. The adiabatic
steam methane reformer utilises the heat input of the feed process stream. This heat input is
required to initiate and drive the endothermic reforming reactions within the reactor. The main
reactions taking place in the reforming process are the water gas shift and the methane reforming.
Apart from the heat input, a catalyst material is also required. Usually the catalyst materials are in
the form of a Nikel-based active metal packing with a porous MgAl2O4 support surface [20, 21].
The structure of the packing material can be either structured or random. In this concept a random
457
packing material of pellets is considered. Finally, the adiabatic operation of the pre-reformer and the
heat exchangers is achieved via external layers of heat insulation.
In terms of operational boundaries, the fuel pre-processing system must supply the MCFC stack a
fuel mixture with a steam-to-carbon (S/C) ratio of 2.0 to 2.1 and a hydrogen molar fraction between
10 and 15%, so as the MCFC stack to operate within the operational limits given by the
manufacturer [2]. In addition, the complete fuel pre-processing system must be able to be enclosed
in a standard 20 feet container. The aforementioned boundaries and space constraints have been
used in the optimisation problem formulation of Section 4.
Fig. 1. Marine MCFC module: Fuel pre-processing system and Fuel cell stack.
d
u tube , 2i
dt
1
Ltube
H tube , 2i
Ah ,tube
Ltube
k tube , 2i
x Ttube , 2i
Tmetal , 2i
(1)
0, 1 , i 1,2...N row
d
u tube , 2i x
dt
1
Ltube
Ah,tube
H tube, 2i x
Ltube
k tube , 2i x Ttube , 2i x
Tmetal , 2i x
(2)
0, 1 , i 1,2...N row
It is noted that the equations are formulated for a non-dimensional domain along the tube length.
This facilitates significantly the numerical treatment of design studies, in which the tube length may
vary. In addition, in the gas flow direction, Nrow+1 nodes are considered (index i), representing the
tube metal conditions exposed in the shell-side temperature profile in the in-between locations of
tube rows (Fig. 3).
The shell-side energy conservation equation is:
Ax , shell
x
d
u shell ,i x
dt
0, 1 , i
H shell ,i x
H shell ,i
Ltube
Ah , shell
Ltube
k shell ,i x
Tshell ,i x
Tmetal ,i
(3)
2,..., N row 1
The tube side heat transfer and pressure drop correlations depend on the phase of the fluid. In single
phase flow (i.e. cross flow heat exchanger) the heat transfer and pressure drop correlations of
Gnielinski are used [31]. In the case of the cross flow evaporator model, where boiling (two-phase
flow) inside the tubes takes place, the heat transfer correlation of Kandlikar [31, 32] is used.
459
For the shell side, two different correlations are used: one for bare tubes and another for finned
tubes [30]. The first heat transfer correlation (for no fins) was developed by Zhukauskas and is
modelled as presented in [30]. The second heat transfer correlation of Briggs and Young for finned
tube bundles in cross flow is employed, as presented and further adapted in [30].
The energy balance for the metal mass interface is given as:
cp
metal
Ah ,tube
Ltube
x
dTmetal,i x
Ametal
dt
k tube,i x
Ttube,i x
Tmetal,i x
Ah , shell
Ltube
k shell ,i x
Tshell ,i x
Tmetal ,i x
(4)
0, 1 , i 1,2.... N row
Especially for the cross flow evaporator model the two-phase flow regime is characterised by the
vapour fraction:
htube,i x hsat ,liq
(5)
vp ,tube ,i x
hsat ,vap hsat ,liq
It is noted that all partial differential equations (1) to (4) are completed by appropriate boundary and
initial conditions. In addition, thermodynamic and physical properties are calculated using
appropriate equations of state, already implemented in DNV COSSMOS [23].
CH 4
H 2O
CH4 2H 2O
CO H 2 O
CO 3H 2 (I)
(6)
CO2 4H 2 (II)
CO2
(7)
H 2 (III)
(8)
The reforming reactions (I) and (II) are strongly endothermic, and thus, high temperatures are
required to obtain the forward reaction. The water-gas shift reaction (III) is moderately exothermic,
so is favoured by low temperatures. This trade-off limits the temperature range for efficient
reforming of CH4. In addition, the reforming reactions are also favoured by low pressures. On the
other hand, the water-gas shift reaction is not significantly affected by pressure.
The SMR is modelled as a tubular reactor consisting of a Ntubes tubes. Since the reformer is
considered to be adiabatic without external heat input or extraction, it can be represented by an
equivalent tube having the same thermochemical/physical regime of the entire tube bundle.
The conservation of mass for each individual species, considering an axial distribution domain
along the tube length is [35, 36, 39]:
MWi
x
dci x
dt
mi x
x
1
L Atube
MWi
L Atube
I , II , III
III
vij r j x
j I
460
(9)
The reaction kinetic model of Xu and Froment [35, 36, 38] is used to obtain the molar reaction
rates:
Ea ,CH 4
rI x
L Atube
RT x
ACH 4 e
pack
2.5
H2
pCO p H3 2
pCH 4 p H 2O
26830
30.11
T x
10266.76 10 e
Ea , H 2 O
rII x
L Atube
RT x
AH 2O e
pack
pH12
pCO pH 2O
pCO2 pH 2
e
4400
4.063
T x
Ea ,CO
rIII x
LAtube
RT x
pack ACO e
1
DEN 2 x
pCO2 p H4 2
p H32.5 pCH 4 p H2 2O
26830
30.11
T x
10266.76 106 e
1
DEN 2 x
4400
4.063
T x
(10)
(11)
1
DEN 2 x
(12)
with:
DEN x
6.65 10
p CH 4 x
4604 . 28
T x
6.12 10
pH2 x
14
9971 .13
T x
(13)
8.23 10
p CO x
8497 . 71
T x
p H 2O x
1.77 10 5 e
p H2 x
10666 . 35
T x
H x
L x
1
Atube
qx
Atube
(14)
0,1
N species
ux /
hx
yi x
hf - p x /
(15)
i 1
N species
H x
mx
hx
yi x
hf
(16)
i 1
qx
k x
Din Tmetal
T x
0,1
(17)
The heat balance across the adiabatic metal surface of the tube is:
mmetal C p, metal
dTmetal
dt
q x dx
0
Qext
N tubes
(18)
0,1
where
is an external heat addition or heat rejection term which in the perfect, insulated, case is
equal to zero.
The heat transfer coefficient k, appearing in Eq. (17), is calculated using the semi-empirical heat
transfer correlations of [34, 37] for packed bed columns with spherical packing material, which are
widely applied in the literature [35, 37, 39, 40]:
k x
2.03
Din
Re
0.8
x e
D pack
Din
for Re
461
20
(19a)
k x
6.15
1 evoid
evoid
Din
0.139 evoid
(19b)
for Re 20
2
0.0339
3
(20)
The momentum conservation is expressed by the steady-state Ergun spatial pressure drop
correlation, which is a standard approach for packed tubular columns [33, 35, 36, 39, 40]:
1 px
L x
C flow,1 v x
u x
1 evoid
evoid D
C flow, 2
2
pack
evoid D pack
1 evoid
0,1
(21)
Finally, the SMR model is complemented by the necessary boundary and initial conditions.
N species
xi
ch
ch, 0
RT0
xi ln xi / MW
i 1,..., N species
(23)
i 1
i 1
The specific chemical exergy is calculated under the assumption of perfect gas mixtures [5]. The
reference state is set at a pressure of p0 = 1.013 bar and temperature T0 = 298.15 K.
The exergy rate at each node and /or spatial position is then calculated:
E
ph
(24)
ch
Within the major system components, namely XHX, XEV and SMR (c.f. Fig. 2) spatially
distributed exergy balances are derived for infinitesimal control volumes. The equations are
formulated under the assumption of steady-state conditions, which are well-suited for normal
MCFC operation [2].
For the SMR component the local exergy balance at each location along the tube length is:
SMR
1
Ltube
ESMR x
x
q
SMR
x ,
0, 1
(25)
q SMR (x) 1 -
T0
TSMR ( x)
(26)
0, 1
The irreversibility rate of SMR results from the integration of the unit irreversibility rate:
462
I SMR
N tubes Ltube
SMR
(27)
x dx
I SMR
E10
(28)
For the XHX and XEV the following expression for the local exergy balance is derived:
1
Y ,i
E shell ,i ( x ) E shell ,i 1 ( x )
Ltube
Etube ,i ( x )
Y XHX , XEV
(29)
The irreversibility rate and the exergetic efficiency of XHX and XEV are:
IY
XHX , XEV
XHX
N row
i 1
N col Ltube
I XHX
and
E1 E2
XEV
(30)
x dx
Y ,i
I XEV
E 2 E3
(31)
For the steam drum (SDR) the irreversibility rate and exergetic efficiency are:
I SDR
E6
E5
E7
E4
(32)
SDR
E7
E4 / E6
E5
(33)
For the steam/ methane mixing junction (MIX) the irreversibility rate and exergetic efficiency are:
I MIX
MIX
E8
E9
(34)
E9 / E7
E8
(35)
E7
The irreversibility and the exergetic efficiency of the SMR system are:
I TOT
TOT
I XHX
I HEV
I SDR
I MIX
(36)
I SMR
I TOT
E1 E3
(37)
Finally, the component contributions to the overall SMR system irreversibility and to the whole
MCFC system, as parts of the fuel input, are:
Y
Y, f
I Y / I TOT , Y
I Y / E8 , Y
(38)
(39)
the pre-defined ranges set by the manufacturer. The total irreversibility of the SMR system is an
explicit performance criterion that is particularly suited for chemical reacting and heat exchanging
systems, since energetic (first-law) approaches do not reveal adequately the full picture of losses
within such systems [5-8]. The problem formulation is:
(40)
min I TOT
X
with independent decision variables the geometry characteristics of the key system components:
X
N col , XHX , Din , XHX , Ltube , XHX , Ptube , XHX , N col , XEV , Din , XEV , Ltube , XEV , Ptube , XEV ,
N tubes ,SMR , Din , SMR , Ltube , SMR
(41)
The optimisation problem is subject to space, design, and operability constraints. The space
constraints mandate that the SMR system must be able to fit in a standard 20 feet container,
resulting in:
Wshell , XHX ,Wshell , XEV 2.5m
(42)
HeightTOT
Lshell , XHX
Lshell , XEV
(43)
2.4m
The design constraints involve the allowable tube pitch to diameter ratios for the heat exchangers:
1.25
Ptube
Dout
(44)
3.0
XHX , XEV
In order to ensure safe and uninterrupted operation and nominal feed conditions to the MCFC
system, a set of operability constraints is imposed. The velocities in the heat exchangers must be
within allowable limits:
5m / s
60m / s
(45)
in , out XHX , XEV
The inlet and outlet velocities of the SMR reactor must be sufficiently low to avoid catalyst
flooding conditions:
0.03m / s
SMR
( x) x
0
x 1
0.38m / s
(46)
The (S/C) at the SMR reactor inlet is set to a near constant value to ensure optimum operation of the
MCFC stack. This entails that for a given fuel flow a respective water flow is derived.
x H 2O
(47)
2.0 S / C
2.12
xCH 4 xCO x XO 2
There is a minimum inlet temperature to the SMR reactor to ensure that the reactions will occur:
T10 640 K
(48)
A minimum pressure is required in the MCFC stack inlet to ensure flow within the stack:
p11 1.032bar
(49)
The hydrogen content at the MCFC stack inlet must be within a certain range:
0.10
xH 2
SMR , x 1
(50)
0.15
Finally, the overall pressure drop in the fuel cell exhaust gas path must be under a limit to ensure
sufficient exhaust back-pressure to the MCFC stack:
464
pTOT
p1
p3 1400Pa
(51)
The problem formulation is completed with the specification of the system design parameters, the
assignment of the independent optimisation variables bounds and the introduction of the boundary
conditions to integrate the SMR system with the MCFC stack. These are presented in Table 1.
Table 1. SMR system design parameters and boundary values.
Optimisation variables
Variable
LB
UB
Ncol, XHX , XEV
1
50
p1 1.035bar
0.01 m
0.09 m
0.m
T1
800K
0.5 m
2.0 m
0.m
m1
0.9203kg/s
0.05 m
0.15 m
300
0.01 m
0.10 m
Ntubes,SMR
Din,SMR
Ltube,SMR
0.5 m
2.4 m
Geometry / design
parameters
0.m
evoid,SMR
p4
0.003m
0.003m
139kg / m
0.5
Operational boundaries
p H 2O
1.4bar
T4
TH 2O
298K
p5
pSDR
1.3bar
p8
pCH 4
1.3bar
T8
m8
TCH 4
mCH 4
298K
0.0146kg/s
5. Results
The optimisation problem of Eqs. (40) (51) consists of 11 independent variables, 3 of which are
integer, and 13 inequality constraints. The PDAE system of the SMR system process model
flowsheet, depicted in Fig. 2, consists of 19,040 equations, implemented in the gPROMS process
modelling environment [41]. The optimisation problem was solved in approximately 111 CPU sec
on an Intel Core Duo @ 2.53GHz with 3.5 GB of RAM.
The optimisation algorithm started from an initial feasible design point of minimum footprint and
progressed towards the minimum total irreversibility optimum. The initial and optimal SMR system
design results are shown in Table 2. The optimal SMR system design offers a significant reduction
of the total system irreversibility of about 50%. It is noted that this efficiency improvement can only
be produced via second-law exergy considerations. The SMR system exergy losses amount to 10%
in the initial and 5% in the optimal design of the chemical exergy of the overall systems fuel. This
formally justifies the need for exergetic optimisation of the SMR system. The improved heat
recovery of the optimal system allows for a higher hydrogen production of 13.7% compared to 10%
of the initial design, due to the increased temperature/ heat input to the reformer. In addition, the
exhaust gas exergy rate and temperature leaving the SMR system are higher than the initial design,
which can be subsequently exploited via heat recovery e.g. for additional power production. The
optimal SMR system design has a higher than the initial height and footprint, however, the 20 feet
container enclosure constraints are still satisfied.
At a component level, the contributions of XHX, XEV, SDR, MIX and SMR to the overall exergy
losses are depicted in Fig. 4. The biggest sources of exergy losses are the water evaporation and
steam/ methane mixing processes. Through the exergetic optimisation the evaporator losses can be
reduced by 67%. On the other hand, the mixing process irreversibility is not affected by the
optimisation and can be considered as an unavoidable source of losses. In the optimal solution, the
mixing process contributes 21.5% to the overall system irreversibility. The steam/ methane heater
contributes 9.8 and 6.6% to the initial and optimal design, respectively. This is another component
that its irreversibility is significantly reduced, by 66%, via optimisation. The SMR contributes
roughly 6% to the overall system irreversibility. The SMR irreversibility in the optimal solution is
465
higher than the initial design due to the higher temperature profile within the reformer. Finally, the
steam drum has a nearly constant irreversibility in both designs, with a contribution from 1.5 to 3%.
Table 2. SMR system optimisation results.
Figure
N col,XHX (-)
Initial
5
Optimal
Figure
Initial
Height (m)
35
Optimal
2.340
1.202
0.0349
0.0340
Footprint (m )
0.854
1.969
Ltube,XHX (m)
2.000
1.125
TSMR,in
717.4
763.3
Ptube,XHX (m)
0.0854
0.0500
616.9
651.4
N col,XEV (-)
PSMR,out
1.0511
1.2214
0.0900
0.0213
xH 2,SMR,out (%)
10.092
13.744
2.000
0.500
S/C (-)
2.12
2.11
0.1485
0.0702
p exh
134
1400
N tubes,SMR (-)
141
300
Texh,out
T3 (K)
653.0
709.2
Din,SMR (m)
0.1000
0.1000
E exh,out
E3 (W)
142334
175301
Ltube,SMR (m)
0.500
2.400
TOT, f
10.01
5.05
I TOT (W)
75445
38076
TOT
17.39
35.33
T10 (K)
P11 (bar)
p1 - p 3 (Pa)
(%)
(%)
Fig. 4. Component irreversibilities (a) and contributions to the system irreversibility (b).
Zooming further to the spatially distributed properties within the system major components, the
exergy losses mechanisms are revealed. In Fig. 5 the unit irreversibility rate and water vapour
fraction profiles of XEV are depicted. Due to the evaporation process, the unit irreversibility rate
along the XEV is high, resulting in the high overall irreversibility rate of the component. The
variance of the unit irreversibility with tube row and distance are moderate resulting also in a
smooth profile in the vapour fraction (latent heat conversion).
In Fig. 6 the unit irreversibility rate and tube temperature profiles along the XHX tubes are shown.
Due to the high temperature regime of the exhaust gasses the unit irreversibility profiles are steep.
There is a significant reduction of the unit irreversibility in the first three tubes. From this row and
on there are only marginal changes in the unit irreversibility. Therefore, the bulk of the heat
addition and temperature increase occurs at the first tube rows of the XHX.
Finally, in Fig. 7 the unit irreversibility, temperature and molar fraction profiles of the SMR are
given. The unit irreversibility exhibits a very steep decrease within the first 20% of the total SMR
length. Therefore, the main heat addition and reactions take place within that distance. This is due
466
to the fact that the SMR operates with the heat input of its incoming stream and there is no other
heat addition across its length. Since the reactions consume heat, the temperature drops quickly in
the first 10% of the total tube length and the hydrogen conversion slows down significantly. For the
rest of the tube distance the hydrogen conversion, temperature decrease and unit irreversibility
contributions progress very slow, having only incremental changes over the length. This is mainly
due to the adiabatic operation of the SMR.
(b)
(a)
Fig. 5. Unit irreversibility rate (a) and vapour fraction (b) profiles within XEV.
(a)
(b)
Fig. 6. Unit irreversibility rate (a) and tube temperature (b) profiles within XHX.
6. Conclusions
In this paper the exergetic analysis and design optimisation of a steam methane pre-reformer system
for marine fuel cells was presented. With the aid of exergy analysis the internal sources of
irreversibility were identified and minimised through exergetic optimisation. The exergetic
optimisation resulted in a system design with 50% less exergy losses compared to the initial low
footprint design. The exergy analysis identified the water evaporation and mixing processes as the
main sources of exergy losses. A step further to the exergy analysis was the formulation of spatially
distributed exergy balances within components. Through this analysis, the propagation of the local
irreversibility within components was identified and the system-level contributions were revealed.
467
The exergy analysis and optimisation proved to be particularly successful for assessing and
improving this chemically reacting and heat exchanging system. It is noted that the energy (firstlaw) and pinch-point calculations have certain shortcomings in their usage, that are not able to
adequately identify the true losses within such systems. The results of this study will serve as the
starting point for the integrated optimisation of the entire fuel cell-based unit.
Temperature (left)
H2 (right)
CH4 (right)
H2O (right)
780
0.7
1200
760
0.6
740
0.5
720
0.4
700
0.3
680
0.2
660
0.1
1400
1000
800
600
400
200
0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
640
0.0
1.0
0.1
0.2
(b)
(a)
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1600
0.0
1.0
Fig. 7. Unit irreversibility rate (a) and temperature / molar fractions (b) profiles within SMR.
Nomenclature
Ai
Ah
Ax
ci
Cflow
cp
Din
Dout
Dpack
E
Ea,i
evoid
h
H
hfin
I
k
L
m
MW
Ncol
Nreaction
Nrow
Nspecies
Nu
p
P
R
rj
Re
s
spfin
T
tfin
ttube
u
Wshell
X
xi
yi
Greek letters
avp
hf
468
f
liq
mean
metal
out
pack
ph
row
shell
tot
tube
vap
i,j
Subscripts
0
ch
col
Reference state
Chemical
Column
Fuel
Liquid
Mean
Metal interface
Outlet
Packing material
Physical
Row
Shell-side
Total
Tube-side
vapour
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
FellowSHIP project web page, Available at: < http://www.VikingLady.no >, [accessed
15/01/2012].
E. Ovrum and G. Dimopoulos, 2012, A validated dynamic model of the first marine molten
carbonate fuel cell, Applied thermal engineering, Vol. 35, No. 1, pp. 15-28.
M. A. Rosen, 1991, Thermodynamic investigation of hydrogen production by steammethane reforming, International Journal of Hydrogen Energy, Vol. 16, No. 3, pp. 207-217.
J. R. Rostrup-Nielsen, 2001, Conversion of hydrocarbons and alcohols for fuel cells,
Physical Chemistry Chemical Physics, Vol. 3, No. 3, pp. 283-288.
A. Bejan, G. Tsatsaronis and M. J. Moran, 1996, Thermal design and optimization, John
Wiley, New York ; Chichester.
A. Boyano, A. M. Blanco-Marigorta, T. Morosuk and G. Tsatsaronis, 2010, Advanced
exergetic analysis of a steam methane reforming system for hydrogen production,
Proceedings of The 23rd International Conference on Efficiency, Cost, Optimization,
Simulation and Environmental Impact of Energy Systems (ECOS), Lausanne
N. Hotz, S. M. Senn and D. Poulikakos, 2006, Exergy analysis of a solid oxide fuel cell
micropowerplant, Journal of power sources, Vol. 158, No. 1, pp. 333-347.
M. A. Rosen, 1990, Comparison based on energy and exergy analyses of the potential
cogeneration efficiencies for fuel cells and other electricity generation devices, International
Journal of Hydrogen Energy, Vol. 15, No. 4, pp. 267-274.
K. Ayoub, 2004, Exergy analysis of a PEM fuel cell at variable operating conditions, Energy
conversion and management, Vol. 45, No. 1112, pp. 1949-1961.
F. Calise, M. Dentice dAccadia, A. Palombo and L. Vanoli, 2006, Simulation and exergy
analysis of a hybrid Solid Oxide Fuel Cell (SOFC)Gas Turbine System, Energy, Vol. 31,
No. 15, pp. 3278-3299.
E. R. Delsman, C. U. Uju, M. H. J. M. de Croon, J. C. Schouten and K. J. Ptasinski, 2006,
Exergy analysis of an integrated fuel processor and fuel cell (FPFC) system, Energy, Vol.
31, No. 15, pp. 3300-3309.
N. Hotz, M.-T. Lee, C. P. Grigoropoulos, S. M. Senn and D. Poulikakos, 2006, Exergetic
analysis of fuel cell micropowerplants fed by methanol, International Journal of Heat and
Mass Transfer, Vol. 49, No. 1516, pp. 2397-2411.
A. Ishihara, S. Mitsushima, N. Kamiya and K.-i. Ota, 2004, Exergy analysis of polymer
electrolyte fuel cell systems using methanol, Journal of power sources, Vol. 126, No. 12,
pp. 34-40.
469
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
B. Pegah Ghanbari, 2007, Energy and exergy analysis of internal reforming solid oxide fuel
cellgas turbine hybrid system, International Journal of Hydrogen Energy, Vol. 32, No. 17,
pp. 4591-4599.
A. Sciacovelli and V. Verda, 2010, Entropy Generation Minimization for the Optimal
Design of the Fluid Distribution System in a Circular MCFC, Proceedings of The 23rd
International Conference on Efficiency, Cost, Optimization, Simulation and Environmental
Impact of Energy Systems (ECOS), Lausanne
C. Ciano, M. Cal and V. Verda, 2009, Analysis of Entropy Generation for the Performance
Improvement of a Tubular Solid Oxide Fuel Cell Stack, International Journal of
Thermodynamics, Vol. 12, No. 1, pp. 1-8.
A. Boyano, A. M. Blanco-Marigorta, T. Morosuk and G. Tsatsaronis, 2011,
Exergoenvironmental analysis of a steam methane reforming process for hydrogen
production, Energy, Vol. 36, No. 4, pp. 2202-2214.
M. J. Stutz, N. Hotz and D. Poulikakos, 2006, Optimization of methane reforming in a
microreactoreffects of catalyst loading and geometry, Chemical engineering science, Vol.
61, No. 12, pp. 4027-4040.
J. R. Rostrup-Nielsen, K. Aasberg-Petersen and P. S. Schoubye, 1997, The role of catalysis
in the conversion of natural gas for power generation, Studies in Surface Science and
Catalysis, Vol. 107, pp. 473-488.
K. Johnsen, 2006, Sorption-Enhanced Steam Methane Reforming in Fluidized Bed
Reactors, PhD Thesis, Norwegian University of Science and Technology, Trondheim,
Norway.
Topse, 2010, AR-401 Pre-reforming catalyst brochure, Haldor Topse A.S., (available at:
www.topsoe.com).
G. G. Dimopoulos, C. A. Georgopoulou and N. M. P. Kakalis, 2011, Modelling and
optimisation of an integrated marine combined cycle system, Proceedings of 24th
International Conference on Energy, Cost, Optimization, Simulation and Environmental
Impact of Energy Systems (ECOS), Novi-Sad, Serbia.
G. G. Dimopoulos and N. M. P. Kakalis, 2010, An integrated modelling framework for the
design, operation and control of marine energy systems, Proceedings of 26th CIMAC World
Congress, Bergen, Norway.
N. M. P. Kakalis and G. Dimopoulos, 2012, Managing the complexity of marine energy
systems, Position Paper 11/2012, Det Norske Veritas, Research & Innovation, (available at:
www.dnv.com).
A. Alebrahim and A. Bejan, 1999, Entropy Generation Minimization in a Ram-Air CrossFlow Heat Exchanger, International Journal of Thermodynamics, Vol. 2, No. 4, pp. 145158.
W. A. Khan, J. R. Culham and M. M. Yovanovich, 2007, Optimal design of tube banks in
crossflow using entropy generation minimization method, Journal of thermophysics and
heat transfer, Vol. 21, No. 2, pp. 372-378.
M. Mishra, P. K. Das and S. Sarangi, 2004, Optimum Design of Crossflow Plate-Fin Heat
Exchangers Through Genetic Algorithm, International Journal of Heat Exchangers, Vol. 5,
No. 2, pp. 379-402.
M. Mishra, P. K. Das and S. Sarangi, 2009, Second law based optimisation of crossflow
plate-fin heat exchanger design using genetic algorithm, Applied thermal engineering, Vol.
29, No. 14-15, pp. 2983-2989.
J. V. C. Vargas and A. Bejan, 2001, Thermodynamic optimization of finned crossflow heat
exchangers for aircraft environmental control systems, International Journal of Heat and
Fluid Flow, Vol. 22, No. 6, pp. 657-665.
R. K. Shah and D. P. Sekuli , 2007, Fundamentals of Heat Exchanger Design, John Wiley
& Sons, Inc., Hoboken, New Jersey.
470
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
A. Bejan and A. D. Kraus, 2003, Heat transfer handbook, John Wiley & Sons, Inc.,
Hoboken, New Jersey.
S. G. Kandlikar, 1990, A general correlation for saturated two-phase flow boiling heat
transfer inside horizontal and vertical tubes, Transactions of the ASME, Journal of Heat
Transfer, Vol. 112, No. 1, pp. 219-228.
M. E. E. Abashar, 2004, Coupling of steam and dry reforming of methane in catalytic
fluidized bed membrane reactors, International Journal of Hydrogen Energy, Vol. 29, No. 8,
pp. 799-808.
A. P. de Wasch and G. F. Froment, 1972, Heat transfer in packed beds, Chemical
engineering science, Vol. 27, No. 3, pp. 567-576.
Y. Ding and E. Alpay, 2000, Adsorption-enhanced steammethane reforming, Chemical
engineering science, Vol. 55, No. 18, pp. 3929-3940.
K. Kim, M. Wang, M. R. v. Spakovsky and D. J. Nelson, 2008, Stochastic Modeling and
Uncertainty Analysis With Multi-Objective Optimization Strategies for the
Synthesis/Design and Operation/Control of a PEMFC Fuel Processing Subsystem, ASME
Conference Proceedings, Vol. 2008, No. 48692, pp. 667-678.
C.-H. Li and B. A. Finlayson, 1977, Heat transfer in packed bedsa reevaluation, Chemical
engineering science, Vol. 32, No. 9, pp. 1055-1066.
J. Xu and G. F. Froment, 1989, Methane steam reforming, methanation and water-gas shift:
I. Intrinsic kinetics, AIChE journal, Vol. 35, No. 1, pp. 88-96.
H. Zhang, L. Wang, S. Weng and M. Su, 2008, Performance research on the compact heat
exchange reformer used for high temperature fuel cell systems, Journal of power sources,
Vol. 183, No. 1, pp. 282-294.
D. F. Rancruel, 2005, Dynamic Synthesis/Design and Operation/Control Optimization
Approach applied to a Solid Oxide Fuel Cell based Auxiliary Power Unit under Transient
Conditions, PhD Thesis, Virginia Institute of Technology, Blacksburg, VA, USA.
PSE, 2009, Process Systems Enterprise, gPROMS, www.psenterprise.com/gproms, 19972009.
471
Abstract
The paradigm of the sustainable energy community is recognized as the future energy approach due to its
economical, technical and environmental benefits. Future systems should integrate renewable energy
systems applying a community-scale approach to maximize energy performances, while minimizing
environmental impacts. Efforts have to be directed toward the promotion of integrated technical systems
needed to expand the use of renewable energy resources, to build sustainable local and national energy
networks, to guarantee distribution systems for urban facilities and to reduce pollution. In this framework
poly-generation is a promising design perspective, for building and district scale applications, in particular
where different types of energy demand are simultaneously present and when sufficient energy intensity
justifies investments in smart grids and district heating networks.
In situ anaerobic digestion of biomass and organic waste has the potential to provide sustainable distributed
generation of electric power together with a viable solution for the disposal of municipal solid wastes. A
thermal recovery system can provide the heat required for district-heating.
The system analysed is a waste-to-energy combined heat and power (CHP) generation plant that perfectly
fits in the sustainable energy community paradigm. The power system is divided in the following sections: a)
a mesophilic - single phase anaerobic digestion of Organic Fraction of Municipal Solid Waste for biogas
production; b) a fuel treatment section with desulphurizer and pre-reformer units; c) a Solid Oxide Fuel Cell
(SOFC) for CHP production; d) a solar collector integrated system(integrated storage system - ISS).
An integrated TRNSYS/ASPEN Plus model for simulating the power system behaviour during a typical
reference period (day or year) was developed and presented. The proposed ISS consists of a solar collector
integrated with storage systems system designed to continuously provide the thermal power required by the
anaerobic digester.
The net thermal energy production as a function of reactor volume, thermal insulation and additional preheating is evaluated.
Keywords:
Distributed Generation, Anaerobic Digestion, Fuel Cells, Transient Model, Waste-to-energy, Integrated
TRNSYS/ASPEN Plus model
1. Introduction
In the last years, the new trend in power generation drives toward distributed power generation [1].
The term distributed generation (DG) [2] indicates that the energy conversion units are situated
close to energy consumers using small power unit when compared to the usual plants used for large
scale production. Several definitions of distributed generation power unit rating exist: from 1 W to
300 MW of capacity, depending of the number of units connected in a modular form. Based on the
classification of [3], the system rating here analyzed is a small capacity (5 kW5 MW).DG
approach should be recognized as the new future power paradigm due to the economic, technical
and environmental benefits it achieves [4]. Application of DG means that single urban districts
could be, in the future, self-supported in terms of electricity, heat and cooling energy.
472
This is in contrast with the facilities present to date, which concentrate the production of goods,
thermal and electrical energy in large plants operating in a central location and connected with the
final users via transmission and distribution networks. The centralized generation paradigm shows
strong limitations due to the vulnerability of complex systems and the scarcity of the fossil fuels
commonly used in large scale plants. On the other hand, small scale plant can advantageously use
fuels locally available, often produced by Renewable Energy Sources (RES), e.g. biogas from
wastes and wood biomass, improving the sustainability of the power generation.
In the European Union, the path toward future energy systems has been clearly underlined with
several directives [5-7], research initiatives [8-10], and short terms energy initiatives (cutting
greenhouse gases emissions by 20%, 20% share of RES in UE energy consumption, cutting energy
consumption through improved energy efficiency by 20%). The success of DG systems is strongly
subjected to their ability to use the waste heat from electricity generation as a heat source, obtaining
total system efficiencies up to 90% [11]. These applications, commonly called Combined Heat and
Power (CHP) can lead to significant reductions of CO2 emissions.
Here a local community scale approach will be applied. It means that, following the definition of
[12], settlements of various types (residential, commercial, industrial) will be considered, in which,
through a mix of technological solutions, the interaction between fuel production and local electric
and thermal energy generation will be optimized, reducing fossil fuel consumption and applying
renewable sources. Poly-generation is then a possible solution, especially when different types of
energy are requested at the same time and different sources such as solar power and waste in this
case - are available.
The system presented in this paper is based on the use of Organic Fraction of Municipal Solid
Waste (OFMSW, [13]) for the generation of biogas, through an anaerobic digestion process
integrated with the heat produced by an array of solar collectors as cycle pre-heating, in order to
feed Solid Oxide Fuel Cell stacks for the electric and thermal power generation. This enhances the
system sustainability in terms of use of renewable sources and of efficient waste management.
Production and collection of OFMSW takes place at district level where energy demands occur,
making it a potential non-seasonal energy feedstock [14]. Moreover, using OFMSW for energy
reduces land demands for waste disposal sites near urban areas where land pressures are high. The
biogas can then be used as fuel for energy production, through several energy systems. Furthermore
biogas is an alternative to natural gas and reduces the amount of greenhouse gas (GHG), since the
carbon dioxide produced by the combustion of biogas is considered biogenic and does not to the
global warming [15].
The biogas can be efficiently used to feed an internal combustion engine for CHP. However, the
need to increase the electrical efficiency of the system leads to the adoption of equipment with
higher conversion efficiency.
Fuel cells are very promising energy conversion devices: they show a very high electrical
efficiency, no moving parts and important for the sustainability zero emissions at the point of
use. Among the different types of fuel cells commercially available, low temperature Polymeric
Electrolyte Membrane Fuel Cell (PEMFC) and high temperature Solid Oxide Fuel Cells (SOFC) are
considered reliable to be applicable for DG [16]. Despite a very low capital cost, PEMFC requires
pure hydrogen to be fed. Therefore biogas must be converted to H2 -rich gas before feeding the
PEMFC, using reformer and CO removal, dramatically increasing system complexity and
decreasing the overall electrical efficiency. On the other hand, SOFC are very flexible in burning
several type of fuels due to the high working temperatures (700-1000C) that allow them to convert
hydrocarbons into hydrogen internally, with an overall electrical efficiency of about 55-60 % even
when fed by methane [17, 18], the absence of moving component (reducing mechanical stresses,
noises and vibrations) and small efficiency reduction when working in off-design conditions.
Moreover this efficiency can be even higher- up to 70% [19] - when SOFCs are coupled with gas
turbine or organic Rankin cycle turbine in hybrid cycles. Due to their modularity, SOFC can be
easily integrated in combine heat and power plants (CHP), reaching very high global efficiency
(around 80 %) [20].
473
On the other hand, technology of fuel cells (and of SOFC in particular), is far to be mature and
many developments have to be done to increase reliability of such equipment before it will be
possible that SOFC can become economically competitive with ICE alternatives.
This paper aims at analysing a community scale poly-generation energy system, through an
integrated TRNSYS/ASPEN Plus model. The system is composed by four main units: a) a
mesophile - single phase anaerobic digestion of Organic Fraction of Municipal Solid Waste for
biogas production; b) a fuel treatment section with fuel steam reforming; c) a Solid Oxide Fuel Cell
(SOFC) for CHP production; d) a solar collector systems to supply heat to the anaerobic digester.
The anaerobic digester process needs a constant temperature in order to maximize the biogas
production. The heat demand can be supplied in different ways, generally with a furnace that uses
the biogas as fuel source. The low price of solar collectors and the sun availability as energy source
in the Mediterranean area supports the idea to use solar thermal collectors to maintain the desired
constant temperature inside the digester.
3. Methodology
In order to evaluate the time-dependent behaviour and the performance of the proposed system a
transient model is developed in the TRNSYS framework [23].To the best of the author knowledge
few studies have appeared to date in the open literature, that propose the use of TRNSYS-framed
models for complex energy systems. Some of the exceptions deal with studies on renewable energy
system integration in grid-connected or stand-alone power systems [24, 25]. Some studies deal with
the integration of fuel cells in buildings energy systems [26] and poly-generation systems with solar
collectors [27].
The transient system layout is schematically shown in Figure 1. Besides the standard TRNSYS
library components, the transient model also includes in-house made types for the digester, the
biogas storage, the SOFC system (derived from Aspen Plus model) and the output temperature and
thermal power controls. The simulation is performed with a hourly time step over a reference year
period.
474
the ambient temperature. It calculates the biogas production, the thermal losses and the thermal
energy need to maintain the inner design operating temperature.
The biogas storage subroutine takes the SOFC duty information and the input biogas flow-rate and
manages the biogas supply to the SOFC system in order to return a constant flow-rate.
The SOFC system subroutine takes into account all the pre-treatment elements described above and
the SOFC itself.
4. Components description
The energy system is composed by a solar collector array equipped with a thermal energy storage,
supplying thermal energy to maintain the digester at a constant inner temperature of 35C. The
digester needs to be warmed-up feeding thermal power from the TES to compensate the heat losses
and to warm-up the substrate .The fuel obtained from anaerobic digestion is sent to a biogas storage
and then fed the SOFC system in order to deliver heat and power to the final users.
A brief description of the main components follows. The main components sizes are summarized in
Table 1.
Table 1. Main components description and nominal size.
Solar collectors array
280 collectors, 560 m2
Thermal energy storage
15 m3 max energy supply 345 kWh
Digester
317.93m 3 volume
Biogas storage
42,800 kg biogas
Auxiliary TES heater (biomass furnace)
72,000 kJ/h (20 kW)
Input data
The input data [29] used for the simulation of the solar collectors are based on a hourly time step
over a year period, referring to Romes latitude, i.e. 4154'39"24 N, as indicative of a central Italian
location. The direct normal insulation (DNI) data show an annual cumulative irradiation of 5760
MJ/m2, with a maximum value of 733.68 MJ/m2 in July and a minimum value of 253.04 MJ/m2 in
December. The weather data are read in TRNSYS by type 89 of the standard TRNSYS library [23].
The monthly average environmental temperature, required to determine the heat losses of the
digester is reported in Figure 2.
476
The waste considered in this study for the digestion process is the organic fraction of municipal
solid waste. The organic substrate is defined by its Total Solid (TS) and Volatile Solid (VS)
concentrations. In the present study a 15%TS and 80%VS substrate of a food industry-sorted
organic waste is considered, following the study of Sans [28]. A continuous feeding schedule is set,
with a loading rate of 7.27 t/d of OFMSW.
Solar Field
The solar field is composed by 280 collectors distributed over 7 strings of 40 collectors each. The
total collector area is of 560 m2. The fluid flowing through the solar circuit is water, with an overall
flow rate of 700 kg/h. The thermal storage is a tank 4 m high with a volume of 15 m3. It is
equipped with an auxiliary biomass furnace of 20 kW to face the solar source fluctuations and
maintain a minimum fluid temperature of 60 C inside the tank.
The solar field is modelled by using the elements available in the standard TRNSYS library [23], in
particular type 73 for the parabolic troughs, type 60 for the thermal energy storage.
Anaerobic Digester
The Anaerobic Digester reactor (AD) is designed for a mesophilic temperature of 35C. The
process considered is a single stage continuous-flow stirred tank reactor (CSTR), operating in wet
regime, with total solids concentration less than 10%, with an organic load equal to 5 kgVS/m3 d
and a hydraulic retention time (HRT) of 15 days [30]. The digester is a concrete reactor, with a
polyurethane external insulation and steel internal layers, partially placed underground (4.5 m). The
top of the reactor consists of an elastic polyurethane surface. Table 2 shows the values used to
design the reactor. Due to some uncertainty in the design process a safety factor equal to 1.2 is
assumed. Then, the final volume of the digester is set equal to about 320 m3.
Table 2. Design parameters of anaerobic digester.
Diameter
Overall height
Organic Load Rate (OLR)
HRT
Safety factor
9m
7m
5kgVS/ m3 d
15days
1.2
The AD heat demand is evaluated considering the heat for warming-up the inlet organic substrate
assuming a specific heat of 1000 kcal/m3 C, and the heat losses through the side walls. The top
surface is considered adiabatic, while heat losses are present through the lateral and lower walls.
Table 3 shows the parameters used and the heat losses evaluated in the digester through the bottom
and lateral surfaces.
The digester is modelled in TRNSYS with an in-house made type which asks for the ambient and
soil temperature and the OFMSW flow rate, returning the produced biogas flow rate and the needed
heat to maintain the digester inner temperature of 35C. The soil temperature is obtained by type
501, which is the soil temperature profile component of the TESS Trnsys library [31].
Table 3. Parameters and power loss in the digester.
Ground surfaces heat transfer coefficient
Air surfaces heat transfer coefficient
Overall power losses
477
0.22 W/m2 C
0.5W/m2 C
2,684W
Biogas storage
The biogas storage is inserted into the system with the aim to decouple the biogas production from
the fuel consumption inside the SOFC. In fact, the digester process is continuously working
producing biogas (except for small maintenance periods) due to the need to treat the OFMSW.
On the other hand, the SOFC requires more frequent stops for maintenance and then the storage has
to be designed in order to compensate the difference in biogas production and consumption. To
guarantee safe operating conditions the storage is designed to allow biogas accumulation for 1000h
during the SOFC inactivity period and to gradually return it during the SOFC activity period, in
order to feed the SOFC with a constant biogas flow-rate. The biogas storage allows accumulating
42,800 kg of biogas.
The biogas storage is modeled in TRNSYS with an in-house made type, which just computes the
amount of biogas stored in the tank in relation with the SOFC duty.
SOFC
The energy system used for electricity and thermal power source consists of a SOFC of 150 kWel
size. The mass and energy flows into the SOFC are analysed by using the AspenPlus software,
successfully used for simulating fuel cells based energy systems [32, 33]. Referring to SOFC
systems, in Aspen Plus the most are simulated by user models [34 - 36] and only a few are
modelled using standard library elements of Aspen Plus [24, 37].
The system is fed by a biogas with a mole based composition of 8 % H 2, 60 % CH4 and 32 %
CO2and a low heating value of 21,600 kJ/m3 .
The SOFC model flow sheet, Figure 3, includes all the components and functions contained in the
system. It is worth noting that the software does not have the fuel cell component and then all the
processes and reactions occurring in a SOFC are modelled separately.
The fuel coming from the biogas storage is mixed with recirculating fluid coming from the fuel cell
(see blue dashed box) for obtaining a steam to carbon ratio equal to 2.5 that is considered optimal
for SOFC [32].
The red box shows the reforming process. This process occurs at a temperature of 600 C (steam
reforming). A constant temperature is imposed in the reformer to control the process and then two
heat exchangers are introduced to give the correct temperature to the fuel before and after the
reformer and to feed the digester with the heat needed for performing the endothermic reforming
reaction. Most of the needed heat comes from the SOFC cooling (thermal stream Q3). Out from the
reformer the fuel is mainly composed by H2, CO and CO2, with traces of methane, butane and
propane.
The blocks inside the black box are used to model the SOFC. The reaction occurs at 1000 C. A
flux of oxygen flows through stream 6 from cathode to anode and it reacts here with the CO and H2.
Energy stream Q2 contains both the electricity produced and the waste heat. The flow exiting the
anode contains some fuel not burned in order to guarantee safe operating working conditions to the
cell (if all the fuel is consumed near the cell exits the reaction does not occur anymore and then the
voltage becomes rapidly zero).
In the split component (green circle) the fluid leaving the SOFC is split in two streams: the first one
mixes with the inlet fuel and it is calibrated to guarantee a S/C ratio of 2.5; the second one, having
some fuel remaining is sent to an afterburner (purple circle) devoted to increase the exhaust outlet
temperature. The high energetic content of the exhausts is then used to pre-heat the air entering the
cathode and to feed the thermal user, i.e. district heating (turquoise circle).
478
5. Results
Overall hourly results
The global results of the transient simulation are shown in Table 4. As can be seen, the solar field
receives about 3550 GJ/y with a global solar energy availability of 4679 hours. The heat effectively
supplied to the TES from the solar field amounts to 134.62 GJ/y which is about 3.8% of the incident
solar energy. Such a low percentage of supplied energy is an effect of the fluid recirculation into the
collectors when the solar radiation is not high enough to determine a fluid temperature that is
adequate to feed the TES (i.e. greater than 70C). In a successive step the solar field will be leaved
without the recirculation and it will pass through the auxiliary heater to be eventually heated up to
the TES inlet temperature.
Assuming that the solar contribution is available only on daytime the TES requires an auxiliary heat
of 194.98 GJ/y during 3652 duty hours (including hours when both systems are working). Taking
into account the variable heat supply and withdrawal to the TES, the average TES temperature at
the tank top and bottom amount respectively to 82.58C and 63.28C.
The heat to be supplied to the digester to maintain the temperature of 35C required from the
mesophilic digestion is equal to269.03 GJ/y.
The produced biogas is about 240 ton/y, with a constant hourly output of 27.39 kg/h.
479
During the SOFC off-duty time the biogas output from the digester is sent to a biogas storage
which, during the SOFC on-duty time releases an hourly rate of 6 kg/h to the SOFC.
The SOFC is fed with 233.75 ton/year of biogas with a constant hourly delivery, during the on-duty
time, of 33.39 kg/h deriving partly from the digester and partly from the biogas storage.
The SOFC output is of about 317 GJ/y of thermal energy and 3,780 GJ/y of electric energy.
Table4. Overall results.
Solar field
3,552.51
4679
134.62
1513
194.98
3652
269.03
75.39
82.58
63.28
37.39
231.61
269.03
239.96
233.75
7,000
316.77
3,779.51
55%
Hourly results
To better understand the behaviour of the system, the focus is now on the hourly time-steps.
According to the system design hypothesis, the only sections which behave in a transient fashion
are those directly linked to the solar collectors field, thus the collectors field itself, the TES, the
biomass furnace and the digester.
Figure 4 shows the thermal power requested by the digester on an hourly-based time. The requested
thermal energy is obviously higher in the winter period, when the difference between the design
temperature and the ambient temperature is higher, reaching power needs of about 20 kW. In some
hours of the summer period the temperature difference leads to power request lower than 1 kW.
Nevertheless, there is a continuous variation of the request during the whole year, corresponding to
the alternation of day and night time.
480
20
kW
18
16
14
12
10
8
6
4
2
0
2000
4000
6000
Hours8000
C
100
90
80
70
60
50
2000
4000
6000
Hours
8000
Figure 5. TES top tank temperature (black line) and bottom temperature (grey line).
481
Figure 6 shows the thermal energy balance of the TES. As already stated, the digester needs a lower
amount of thermal energy during the summer period with respect to the winter one. The thermal
energy supplied by the TES (Qsup) follows the request of thermal energy from the digester.
Moreover, the seasonal behaviour of the solar energy supply (Q RES) influences the auxiliary thermal
supply Qaux (i.e. the biomass furnace) performance, which shows an inverted seasonal pattern. It is
worth noting that the sum of the thermal energy delivered to the TES (i.e. QRES and Qaux) is higher
than the supplied thermal energy to the digester from the TES (Qsup). This is due to the need to
maintain the minimum temperature of 60C in the TES itself in order to assure the thermal supply
when needed. During the summer period, in the months of July and August, the solar collectors
field thermal energy supply is able to almost fulfil the thermal request of the digester.
6. Conclusions
A transient model for the simulation of a biogas fuelled CHP 150 kWel plant was presented. The
integrated renewable sources power plant here described has the potential to achieve high
generation efficiencies as well as important environmental benefits through the mitigation of
greenhouse gas emissions.
The biogas is produced by an anaerobic digester processing OFMSW. The heat required to warm
the substrate and to compensate losses is obtained by a solar field integrated by an auxiliary
biomass furnace. The biogas produced is sent to a SOFC for co-generation of electricity and heat
for possible use in a district heating (not modelled here). A large fraction of the produced heat is
used for pre-heating the SOFC plant, so only 10% of the available heat is available for end users.
The transient model shows that the heat supplied to the digester by the solar field allows a saving of
7.63 ton/y of biogas or the 4% of the total biogas production, equal to 131 GJ/y of electricity
energy produced by the SOFC. Further improvements are in progress. A positioning of the solar
field on the top of the digester as cover for the gas holder will allow a more efficient heat supply
system for the digester. Moreover different type of solar collectors with a higher energy efficiency
could totally substitute the biogas furnace, with considerable economic saving. .
The whole system adopts a mix of renewable technologies to produce electricity and thermal power.
Furthermore this approach indicates a smart solution for the disposal of organic fraction of
municipal solid waste, allowing to define a fully sustainable chain from waste to energy.
482
Bibliography
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
Alanne, K. and Saari, A., Distributed energy generation and sustainable development,
Renewable and Sustainable Energy Reviews, 2006, 10: 539558.
Ackermann T., Andersson G., Sode, L., Distributed generation: a definition, Electric Power
Systems Research 2001, 57: 195204.
Elkhattam W. and Salama M. M. A., Distributed generation technologies, definitions and
benefits, Electric Power Systems Research, 2004, 71 (2): 119-128.
Manfren M., Caputo P. and Costa G., Paradigm shift in urban energy systems through
distributed generation: Methods and models, Applied Energy, 2011, 88:10321048
Directive 2002/91/EC, Directive 2002/91/EC of the European Parliament and of the Council
of 16 December 2002 on the energy performance of buildings.
Directive 2006/32/EC, Directive 2006/32/EC of the European Parliament and of the Council
of 5 April 2006 on energy end-use efficiency and energy services and repealing Council
Directive 93/76/EEC.
Directive 2009/28/EC, Directive 2009/28/EC of the European Parliament and of the Council
of 23 April 2009 on the promotion of the use of energy from renewable sources and amending
and subsequently repealing Directives 2001/77/EC and 2003/30/EC
European Commission C(2011)9493 of 20 December 2011, Work Programme, Theme 5
Energy
Commission of the European Community, Investing in the Development of Low Carbon
Technologies(SET-Plan), October 2009, COM (2009) 519
Commission of the European Community, A European Strategic Energy Technology Plan
(SET-PLAN): 'Towards a low carbon future', November 2007, COM (2007) 723
Strachan, N. and Farrell A., Emissions from distributed vs. centralized generation: The
importance of system performance, Energy Policy, 2006, 34: 26772689.
VV.AA., Dalleco-building al distretto energetico: la proposta ENEA per un modello di
sviluppo fondato su eco-edifici e generazione distribuita, ENEA Workshop,2007, Roma (in
Italian).
Eriksson, O., Bisaillon, M., Haraldsson, M., Sundberg, J., Integrated waste management as a
mean to promote renewable energy, World Renewable Energy Congress 2011, 8-11 May
2011, Sweden.
Gregg. J., National and regional generation of municipal residue biomass and the future
potential for waste-to-energy implementation, Biomass and Bioenergy, 2010, 34 (3): 379-388.
Morin, P., Marcos, B., Moresoli, C. and Laflamme C.B., Economic and environmental
assessment on the energetic valorisation of organic material for a municipality in Quebec,
Canada, Applied Energy, 2010, 87 (1): 275-283.
Calise, F., Design of a hybrid polygeneration system with solar collectors and a Solid Oxide
Fuel Cell: Dynamic simulation and economic assessment, International Journal of Hydrogen
Energy, 2011, 35: 6128- 6150.
Larminie, J. and Dicks, A., Fuel cell system explained. John Wiley & Sons Ltd; 2004.
Shiratori Y., Oshima T. and Sasaki K., Feasibility of direct-biogas SOFC, International
Journal of Hydrogen Energy, 2008, 33 (21): 6316-6321
Calise F., Dentice DAccadia M., Palombo A., and Vanoli L., Simulation and exergy analysis
of a hybrid Solid Oxide Fuel Cell (SOFC)Gas Turbine System, Energy, 2006, 31 (15): 32783299
Liso V., Zhao Y., Brandon N., Nielsen M. P., and KrS. K., Analysis of the impact of heatto-power ratio for a SOFC-based mCHP system for residential application under different
483
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
climate regions in Europe, International Journal of Hydrogen Energy, 2011, 36 (21): 1371513726
Axaopoulos, P., Panagakis, P., Tsavdaris, A., Georgakakis, D., Simulation and experimental
performance of a solarheated anaerobic digester, Solar Energy, 2001, 70 (2): 155164.
Alkhamis, T.M., El-khazali, R., Kablan, M.M. and Alhusein, M.A., Heating of a biogas
reactor using a solar energy system with temperature control unit, Solar Energy, 2001, 69 (3):
239-247.
Klein, S.A., Beckam, W.A., Mitchell, J.W., Braun, J.E., Evans B.L., Kummert J.P., et al.,
2000, TRNSYS a transient system simulation program. Version 15.1, Madison: Solar
Energy Laboratory, University of Wisconsin; 2000.
Quesada B., Snchez C., Caada J., Royo R., PayJ., Experimental results and simulation
with TRNSYS of a 7.2kWp grid-connected photovoltaic system, Applied Energy Volume 88,
Issue 5, May 2011, Pages 17721783.
Corsini, A., Rispoli, F., Gamberale, M., and Tortora, E., 2009, Assessment of H2- and H2Obased renewable energy-buffering systems in minor islands, Renewable Energy 34 (2009)
279288.
Dorer V., Weber R., Weber A., Performance assessment of fuel cell micro-cogeneration
systems for residential buildings, Energy and Buildings, Volume37, Issue 11, November
2005, Pages 11321146.
Calise F., Design of a hybrid poly-generation system with solar collectors and a Solid Oxide
Fuel Cell: Dynamic simulation and economic assessment, International Journal of Hydrogen
Energy Volume 36, Issue 10, May 2011, Pages 61286150.
Sans, C., Mata-Alvarez, J., Cecchi, F., Pavan, P. and Bassetti, A., Volatile fatty acids
production by mesophilic fermentation of mechanically sorted urban organic wastes in a plug
flow reactor, Bioresource Technology, 1995,51: 89-96.
Solar Energy Laboratory (NREL), University of Wisconsin-Madison, 2003, Generated hourly
weather data.
Laraia R., Il trattamento anaerobico dei rifiuti, Manuali e Linee Guida 13 ANPA Unita
Normativa Tecnica, 2002, (in Italian).
Thermal Energy System Specialists, LLC, TESS COMPONENT LIBRARIES General
Descriptions, www.trnsys.com.
Zhang, W., Croiset, E., Douglas, P.L., Fowler, M.W., Entchev, E., Simulation of a tubular
solid oxide fuel cell stack using ASPEN Plus unit operation models, Energy Conversion
and Management 46 (2005) 181-196.
Borello, D., Del Prete, Z., Marchegiani, A., Rispoli, F. and Tortora, E., Analysis of an
integrated PEMFC/ORC power system using ammonia for hydrogen storage, ASME
TurboExpo 2012, GT2012-68599, Copenhagen, 11-15 June, 2012.
Doherty W., Reynolds A., Kennedy D., Computer simulation of a biomass gasification-solid
oxide fuel cell power system using Aspen Plus, Energy 35 (2010) 4545e4555.
Kim Y., Hong S., Nam S., Seo S., Yoo Y., Lee S., Development of 1 kW SOFC power
package for dual-fuel operation, international journal of hydrogen energy 36 (2011) 1024710254.
Kuchonthara P., Bhattacharya S., Tsutsumi A., Energy recuperation in solid oxide fuel cell
(SOFC) and gas turbine (GT) combined system, Journal of Power Sources volume 117, Issues
1-2, 15 May 2003, Pages 7-13.
Barelli L., Bidini G., Gallorini F., Ottaviano A., An energetic-exergetic comparison between
PEMFC and SOFC-based micro-CHP systems, International Journal of Hydrogen Energy,
Volume 36, Issue 4, February 2011, pages 3206-3214.
484