energy for sustainable future - eminent...

i

ENERGY FOR SUSTAINABLE FUTURE

Proceedings of Workshop Early-Stage Energy Technologies for Sustainable Future:

Assessment, Development, Application - EMINENT 2 5 – 6 May 2008, University of Pannonia, Veszprém, Hungary

Workshop of the EMINENT-2 project TREN/05/FP6EN/S07.56209/019886 funded by the European Community under the 6-th Framework Programme for Research and

Technological Development.

The workshop is organised in synergy with the projects

TEMPUS-TACIS JEP_26045_2005, ECORSE "Ecological and Resource Saving Engineering"

MARIE CURIE CHAIR (EXC), MEXC-CT-2003-042618, INEMAGLOW "Integrated Waste to Energy Management to Prevent Global Warming"

and European Reintegration Grants (ERG) project MERG-CT-2007/46579, ESCHAINS "Energy Supply Chains Design and Management for Higher Efficiency

and Sustainable Future".

Editors:

Petar Varbanov and Jiří Klemeš

University of Pannonia, Veszprém, Hungary

Igor Bulatov

The University of Manchester Manchester, UK

ii

Copyright This volume is protected by the copyright of the Authors of the manuscripts and the following terms and conditions apply to its issue: Assignment of copyright is not required from Authors. In these proceedings the copyright is retained by the Authors of the included manuscripts. However, no part of this volume may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, electrostatic, magnetic, mechanical recording or otherwise without permission from the copyright holder, in this case the Authors.

Disclaimer Whilst every effort is made by the publisher and editorial board to see that no inaccurate or misleading data, the opinion appearing in the articles herein are the sole responsibility of the contributor concerned. Accordingly the publisher, the copyright holder, the editorial board and their respective employee, officer and agents accept no responsibility or liability whatsoever for the consequence of any inaccurate or misleading data, opinion or statement.

Official organisers: University of Pannonia, Egyetem utca 10, Veszprém, H – 8200, Hungary Phone: +36 88 421 664, [email protected] The University of Manchester, Sackville Str 88, Manchester, M60 1QD, UK Phone: +44 161 306 4389, [email protected]

© Cover Photograph László Sikos, 2008

© Cover Design Hon Loong Lam, 2008

Veszprém, May 2008

ISBN 978-963-9696-38-9

Publisher: Faculty of Information Technology Printed: UoP University Library Archives, University Publisher, UoP Press

Director-general: Egyházy Tiborné dr. Number of the work: 2008/50

iii

Dear Delegates and Guests On behalf of the Faculty of Information Technology of the University of Pannonia, Regional Committee of the Hungarian Academy of Sciences, Veszprém and The University of Manchester, UK we are very pleased to welcome you on the Workshop of the EMINENT-2 project being held May 5th-6th, 2008 in the Main Chamber of Veszprém Regional Centre of the Hungarian Academy of Sciences, located in the Castle of the city Veszprém. Castle of Veszprém The historic castle was originally built in the 10th-11th centuries on one of the seven hills of Veszprém. Because of wars in the middle ages, it was ruined and rebuilt several times. Finally, it gained a Baroque character as can be seen now. Many sights, including excellent landscapes and important cultural venues, are clustered here. One of the masterpieces of Baroque, the Archiepiscopal Palace holds a valuable library and archive open to the public. Next to the palace are the Romanesque St George's Chapel and the Baroque Grand Provost's House of medieval origin. Other Baroque treasures include the Piarist Priory (today an archive), the convent’s church and the Piarist grammar school, as well as the Franciscan church and monastery. Among the most beautiful buildings in the district are some that were named after the people who commissioned them, for example, the Dubniczay House, the Márton Bíró House, the Dravecz House or the Tejfalussy House, today home to the Queen Giselle Museum. The Baroque Fire Watch Tower calls from afar, tempting visitors with great sight to climb up the winding staircase to enjoy a fantastic view of the town and its surrounding hills. University of Pannonia Veszprém was among the first Hungarian cities to have a university - students studied law and arts here. The university was destroyed by fire in 1276, after which Veszprém became a university town again in the 20th century. In 1949 the Heavy Chemical Industrial Faculty opened its gates to its first students in Veszprém. In 2006 University of Veszprém changed its name from the late Roman province to the University of Pannonia to reflect its role at the centre of an important and historic region. Currently the University of Pannonia has 5 faculties - Engineering, Arts, Georgikon Faculty of Agriculture, Faculty of Information Technology and Faculty of Economics. Faculty of Information Technology The structure of the Faculty of Information Technology can be described as a three dimensional organisation: research, development and education, with all three

iv

functions fulfilled in the same proportion. The three areas enhance each other; the synergy brings benefits for the whole Faculty. The basic and applied research generates the theoretical foundation for the R&D activities, whereas the feedback of the industrial partners and end users provide guidance to the research orientation. The educational activity builds up on this mastery. The Faculty runs a country level elite program for the best secondary school students in mathematics for a tertiary level enrolment. Our undergraduate students have the opportunity to select either research or R&D oriented courses. This process generates highly qualified professionals both for the PhD program and for the industry. The Faculty has 5 departments: Computer Science – with a centre for Process Synthesis and Optimisation; Automation; Information Systems; Image Processing and Neurocomputing; Nanotechnology. The Research Institute of Chemical Engineering also functions as a part of the Faculty. About a sustainable future The future development of both the national and world economies is closely connected to exploitation of raw materials, energy saving and pollution reduction. The topic has received a lot of attention and has been widely discussed both in the media and scientific journals over the last several years. It is logical that an idea of a specialised conference closely related to environmental issues such as energy savings and pollution reduction in various areas of human activities has risen on various occasions. All nations are sharing just one planet and most environmental production and pollution issues have global impact in the long term. We have the pleasure of hosting the Workshop of the EMINENT-2 project (TREN/05/FP6EN/S07.56209/019886) funded by the European Community under the 6th Framework Programme for Research and Technological Development. We also appreciate contribution from the other EC supported projects, especially from TEMPUS – TACIS and Marie Curie projects. With over 50 contributors from Africa, America, Asia and Europe, representing 25 countries we look forward to a stimulating event. Prof Ferenc Friedler Dean of Faculty of Information Technology University of Pannonia

v

Contents The Integration of Waste and Renewable Energy Sources for Heating and Cooling Demands in Locally Integrated Energy Sectors 1 Simon Perry, Jiří Klemeš, Igor Bulatov Waste to Energy Technologies and Biomass Utilisation 11 Petr Stehlík Biofuels: Thermodynamic Sense - and Nonsense 19 Jack W Ponton Automated Targeting for Segregated Energy Planning 29 Denny Kok Sum Ng, Dominic Chwan Yee Foo, Raymond R. Tan Beyond Production Integrated Environmental Protection (PIEP) in Food Industry 39 David Napper, Igor Bulatov Design and Operation Optimization for a Stand-Alone Power System Using Renewable Energy Sources and Hydrogen Storage 47 Dimitris Ipsakis, Spyros Voutetakis, Panos Seferlis, Fotis Stergiopoulos, Simira Papadopoulou, Costas Elmasides Flow Distribution Effects on the Thermal Efficiency of a Brazed Plate Heat Exchanger 57 David J. Kukulka, Anthony DeStefano P-graph Methodology for Cost-Effective Reduction of Carbon Emissions Involving Fuel Cell Combined Cycles 67 Ferenc Friedler, Petar Varbanov, L. T. Fan Reactive and non-reactive distillation sequences: Energy saving by process integration 77 Ivo Müller, Eugeny Y. Kenig Evaluation of Heat-integrated Distillation Schemes 87 Mansour Emtir Using Captured Carbon Dioxide for Enhancing Growth of Algae Ponds 95 Attilio Converti, Patrizia Perego, Luigi Maga, Vincenzo Dovì Energy and resource saving at operating plants based on the analysis of historical process data 105 László Dobos, Balázs Balasko, Sándor Németh and János Abonyi

vi

The Use of Plate Heat Exchangers for Energy Saving in Phosphoric Acid Production 115 Petro Kapustenko, Stanislav Boldyryev, Olga Arsenyeva, Gennadiy Khavin The role of Process Integration in CCS Systems in Oil Refineries and Pulp Mills 123 Erik Hektor, Thore Berntsson Energy Saving Processes in Biomass Refinery 133 Endre Nagy Energy Efficiency Improvement of Wastewater Treatment Processes Using Process Integration Techniques (Anaerobic Digestion) 143 J.Rojas-Hernandes, Toshko Zhelev Minimisation of Energy Use in Multipurpose Batch Plants Using Heat Storage: A Mathematical Programming Approach 153 Thokozani Majozi Early Stage Innovative Technologies for the Treatment of Wet Biomass and Organic Wastes: the HTUÒ Process and Supercritical Water Gasification 163 Jan Zeevalkink, Jaap Koppejan, Wilfrid Hesseling Comparison between EMINENT and other Energy Technology Assessment Tools 173 Raquel Segurado, Sandrina Pereira, Ana Pipio and Luís Alves Quick Progress in High-Efficiency Motors and Drives Opens New Markets and a Large Saving Potential 183 Norbert N. Vasen, Alessandra Boffa Environmentally Friendly Energy Supply in Wineries 191 Maximilian Lauer, Reinhard Padinger Raw materials for fermentative hydrogen production 201 Krzysztof Urbaniec, Robert Grabarczyk Development of a Graphical Analysis Method for Renewable Energy Supply Chain 209 Hon Loong Lam, Petar Varbanov, Jiri Klemeš Life Cycle Assessment as an Environmental Assessment Tool in Municipal Solid Waste Management 219 Luca De Benedetto, Jiří Klemeš Training Methods of Multiple-Skill Specialists in the Field of Energy-Saving in the Fuel and Energy Balance Complex 229 Sergey Mikhailov, Valery Meshalkin Technical Analysis of Desiccant Cooling System 233 François Boudéhenn, Françoise Burgun

vii

Fixed Bed Gasification of Biomass Fuels: Experimental Results 241 Nikolaos Koukouzas, Catalin Flueraru, Anastasios Katsiadakis, Evangelos Karlopoulos Mathematical simulation and optimization of forest management with consideration of carbon balance 251 Oleg Butusov, Valery Meshalkin Energy-Saving Issues in Reactive Distillation Schemes 257 Alexandra E. Pleşu, Jordi Bonet, Cristian I. Ciornei, Valentin Pleşu, Grigore Bozga, Maria-Isabel Galan Reliability and maintenance software in energy generation and saving 267 László Sikos, Jiří Klemeš CO2 Capture on Zeolites for Sustainable Energy Production 277 Gheorghe Bumbac, Aurelia Bolma, Anca Dumitrescu, Vasile Bologa New Energy-Saving Technologies in the Chemical Industry 287 Vladimir Panarin, Valery Meshalkin Innovative Master Course on Computer-Aided Synthesis of Energy-Saving Process Systems at Mendeleev University 291 Valery Meshalkin, Gleb Zakhodiakin New Efficient Technology for Stabilisation of Humid Biomass in View of Large-Scale Supply 295 Giuliano Grassi, Stephane Senechal Commercial Production of Bio-H2 from Agro-Forestry Residues 297 Giuliano Grassi, Stephane Senechal Innovative Low-Energy Distillation Technology for Bio-Ethanol 299 Giuliano Grassi, Stéphane Sénéchal Technological, Economic and Organisational Innovations in the Field of Energy-Saving in Industry Teaching Methods 301 Maksim Dli, Tatiana Kakatunova A successful heat integration at Lube oil Refinery Unit in Danube Refinery 305 Zoltán Varga; Tibor Karmacsi; Klára Kubovics Stocz; Alexandra Szűcs Mapping of Key Problems in Energy Saving Research 306 Andrzej Kraslawski, Yuri Avramenko Brent Field 3D Reservoir Simulation 307 Julius M. Tollas

viii

Energy Issues in the Restructuring of EU Sugar Industry 308 Krzysztof Urbaniec, Jacek Wernik On large point sources for CO2 and the emerging challenge relating to coal-based power generation in a global perspective 309 Jens Hetland Sugar beet based bio-refinery concepts - the results of sector wide process synthesis 310 Gernot Gwehenberger, Michael Narodoslawsky Local learning energy-efficiency networks – a mean of dissemination for knowledge about new energy technologies 311 Martin Jakob, Eberhard Jochem

ix

Author Index Abonyi J. 105

Alves L. 173

Arsenyeva O. 115

Avramenko Y. 306

Balasko B. 105

Berntsson T. 123

Boffa A. 183

Boldyryev S. 115

Bolma A. 277

Bologa V. 277

Bonet J. 257

Boudéhenn F. 233

Bozga G. 257

Bulatov I. 1, 39

Bumbac G. 277

Burgun F. 233

Butusov O. 251

Ciornei C. I. 257

Converti A. 95

De Benedetto L. 215

DeStefano A. 57

Dli M. 301

Dobos L. 105

Doví 95

Dumitrescu A. 277

Elmasides C. 47

Emtir M. 87

Fan L.T. 67

Flueraru C. 241

Foo D. C. Y. 29

Friedler F. 67

Galan M. I. 257

Grabarczyk R. 201

Grassi G. 295, 297, 299

Gwehenberger G. 310

Hektor E. 123

Hesseling W. 163

Hetland J. 309

Ipsakis D. 47

Jakob, M. 311

Jochem, E. 311

Kakatunova T. 301

Kapustenko P. 115

Karlopoulos E. 241

Karmacsi T. 305

Katsiadakis A. 241

Kenig E. Y. 77

Khavin G. 115

Klemeš J. 1, 209, 219,

267

Koppejan J. 163

Koukouzas N. 241

Kraslawski A. 306

Kubovics Stocz K. 305

Kukulka D. J. 57

Lam H. L. 209

Lauer M. 191

Maga L. 95

Majozi T. 153

Meshalkin V. 229, 251, 287,

x

291, 251

Mikhailov S. 229

Müller I. 77

Nagy E. 133

Napper D. 39

Narodoslawsky M. 310

Németh S. 105

Ng D. K. S. 29

Padinger R. 191

Panarin V. 287

Papadopoulou S. 47

Perego P. 95

Pereira S. 173

Perry S. 1

Pipio A. 171

Pleşu A. E. 257

Pleşu V. 257

Ponton J. W. 19

Rojas-Hernandes J. 143

Seferlis P. 47

Segurado R 173

Sénéchal S. 295, 297, 299

Sikos L. 267

Stehlik P. 11

Stergiopoulos F. 47

Szűcs A. 305

Tan R. R. 29

Tollas J. M 307

Urbaniec K. 201, 308

Varbanov P. 67, 209

Varga Z. 305

Vasen N. N. 183

Voutetakis S. 47

Wernik J 308

Zakhodiakin G. 291

Zeevalkink J. 163

Zhelev T. 143

ENERGY FOR SUSTAINABLE FUTURE Edited by Petar Varbanov, Jiří Klemeš, Igor Bulatov University of Pannonia, Veszprém, Hungary, 5-6 May 2008 Copyright © Manuscript Authors, ISBN 978-963-9696-38-9

1

The Integration of Waste and Renewable Energy Sources for Heating and Cooling Demands in Locally Integrated Energy Sectors

Simon Perry*1, Jiří Klemeš 2, Igor Bulatov1 1 Centre for Process Integration, CEAS, The University of Manchester

PO Box 88, Manchester, M60 1QD, UK, email: [email protected] 2 EC Marie Curie Chair (EXC) “INEMAGLOW”, Research Institute of Chemical

Technology and Process Engineering, FIT, University of Pannonia Egyetem ut 10, Veszprém, H -8200, Hungary

Abstract Increasing concern over global warming related to CO2 emissions, coupled with steeply rising energy prices, has resulted in massive societal interest in alternative non-carbon based energy sources and improved efficiency. Industrial energy improvements have focused on improvements to efficiency rather than the integration of renewable sources. Domestic energy use has focused on small scale renewable sources, and minor efficiency improvements. The design of a combined energy system for both industrial and residential/social buildings has been limited and ad-hoc, with no systematic design techniques employed to produce a symbiotic system. This paper proposes a method used extensively in the chemical process industries as a means to systematically design energy systems for Locally Integrated Energy Sectors (LIES).

Introduction At the start of the 21st century, rising energy costs coupled with increasing concern over global warming related to CO2 emissions have resulted in massive interest in alternative non-carbon based energy sources. In addition, there has been renewed interest in energy efficiency methods compared to only a few years ago when energy was relatively inexpensive and abundant. In the main, large scale renewable energy sources have been mainly directed towards satisfying domestic requirements, principally in the form of electricity. Small scale renewable energy sources have been put forward as means to satisfy heating and cooling demands, but have been directed in the main towards individual dwellings. Improving energy efficiency has been the main weapon in combating rising energy costs and carbon emission tariffs in the industrial sector, with relatively little interest in the possibilities of using large scale renewable energy sources. There has also been interest in improving energy efficiency in the domestic sector, but again has focused on improvements to insulation, low energy lighting, and more energy efficient appliances. Although all of these factors are moving societies towards lower carbon based energy use and reduced CO2 emissions, there has been little attempt at planning and producing a more integrated energy system that includes both industrial and domestic supply and demands. This paper will attempt provide an overall framework to initially meet the heating and cooling demands of a Locally Integrated Energy Sector (LIES) based on existing studies of integrated energy systems for industrial scale processes.

ENENERGY FOR SUSTAINABLE FUTURE

2

Local distributed renewable energy sources There are a number of renewable energy sources which are available to provide local heating and cooling, and in doing so can reduce the level of greenhouse gas emissions in comparison to fossil fuel based systems. There has been much interest and research over the last 5 to 10 years in reduced emission renewable energy sources, but in this section we are focusing on those sources that have been demonstrated or are operating on the local or regional scale. Additionally, there are other energy sources which predominantly or solely provide power, such as wind turbines, but in the current context these are outside the scope of this paper. In this section, renewable energy sources also include heat from waste, as in many cases waste contains a considerable proportion of biomass, and additionally the use of waste has a beneficial effect on greenhouse gas emissions due to the saving in additional carbon based emissions related to other forms of disposal. The use of heat pumps to provide low grade heat for space heating has become relatively widespread. Heat pumps are used to upgrade low temperature heat from sources such as ambient air, exhaust air, ground soil, ground rock, ground water and surface water, to higher temperature heat sources which can be used for principally residential or domestic purposes [1]. The two principal technologies used in heat pumps are compression and absorption. The compression heat pump consists of an evaporator, compressor, condenser, and an expansion valve. They are mainly driven by electricity, but can also be driven by gas or diesel engines. A coefficient of performance (COP) depends on the input temperature of the heat source, and the output temperature required, but under the most usual conditions the COP varies between 4 and 5. The majority of heat pumps currently installed for the provision of domestic (including office and municipal buildings) heating are closed-loop compression type systems driven by electricity. Most of these heat pumps work with water entering the system at between -5 OC and +12 OC, and exit at temperatures as high as 55 OC. The output heat can be used to provide underfloor heating (temperatures required 30 OC to 45 OC), low temperature heating radiators (45-55 OC), conventional radiators (60 to 90 OC) and air (30 to 50 OC). For systems using conventional radiators, other energy sources may be required to increase temperatures to the required level. Larger scale geothermal energy sources can also be used to provide local sources of heat. In Iceland, geothermal hot water at temperatures between 100 OC and 130 OC is provided by a series of deep production wells and natural hot springs and pools [2]. The current heat sources have a potential of 75-100 MW flow development. The high temperature geothermal water can be utilised for electricity production, space heating, domestic hot water, and various industrial applications. The geothermal water is used in a Kalina based CHP system. The steam raised in the process (120 °C and 32 bar) is used in a turbine system to produce 1.7 MW of power. The waste from the system, at 80 °C, is then utilised in the associated district heating system for households and industry. Further geothermal based heat and power systems have also been developed in Iceland [3]. The Hellisheiði geothermal system is planned to produce 300 MW of electricity and 400 MW of heat by 2009. Part of the plant producing power is already in operation.


3

Biomass based systems are another renewable energy source that can provide heat and can reduce the production of greenhouse gases. There are a number of methods that have been employed. First, and by far the most common, is direct firing of biomass. For example, in Ignalina, Lithuania, a 6 MW wood fired boiler was integrated into the district heating system [4]. The wood fired boiler produces 25,000 MWh/y of heat production, and is supplemented by the 10,000 MWh/y produced by the conventional heavy oil fuelled boilers. The biomass used in the boiler is a mixture of sawdust and wood chips from local woodworking industries. A similar biomass fired district heating system has been constructed in Catalonia, Spain. In this system the biomass based system provided hot water to 250 houses, and between January and November 2001 the use of 500 t of biomass had replaced the equivalent of 165 t of oil, and so avoided the production of 380 t of CO2 [5]. There are approximately 300 biomass fired district heating systems operating in Austria, producing around 450 MW of energy. The biomass is sourced from chipped wood from forestry, by-products of wood processing with bark, by-products of wood processing without bark, sawdust, bark, and straw [6, 7]. There are also 12 biomass district heating plants that are operated in combination with solar cells, which provide additional pre-heating of the water. The Biomass renewable energy source can also be used in gasification processes for the production of heat and power. In large scale gasification systems, the gasification system thermally converts the biomass to simple chemical building blocks, which can then be used to produce fuels, products, power, and hydrogen [8]. Small-modular gasification systems are also planned. These systems would operate in the range of 5 kW to 5 MW, and would provide heat and power from localised sources of biomass. The gasification product, composed primarily of carbon monoxide and hydrogen, is cleaned and then used in gas turbines or internal combustion engines. In addition to the power produced, waste heat can be directed to district heating based systems [9]. Although gasification processes have long been put forward as the basis of a biomass based energy and chemical industry, as yet few demonstration plants have been built. Solar heating and cooling is another renewable energy source that has found worldwide application. This is a mature technology that has proven to be reliable and cost-competitive since solar water heaters were introduced over thirty years ago. Although these systems have been mainly centred on producing heat and power for individual or large residential/commercial buildings, there has also been widespread interest in using concentrated solar heat systems for industrial applications [10]. In active solar heating systems, water, or another heat transfer fluid, is circulated through a duct and heated by transfer from direct solar radiation on the collector panel. Various designs of collectors are utilized in order to concentrate the solar radiation on the fluid duct and to maximize solar gains. The amount of heat energy captured per square metre of collector surface area varies with design and location but typically can range from 300 - 800 kWh/m2/y. Some designs use a heat transfer


4

fluid that when warmed flows to a storage tank or a heat pump where the heat is then transferred to water that can then be used as hot water or for space heating. Solar water heating uses the radiation from the sun to heat water in a panel often sited on the roof which in turn can supply that heat as a hot water or to a central heating system. If the system is sized correctly, it can provide at least 40-60 % of the household all hot water requirements throughout the year. Concentrated solar heating systems are usually used for medium scale heat application systems in industry, agriculture and food production. Similar to concentrating solar power (CSP) systems for electricity generation, a concentrating solar heater (CSH) device consists of a concentrator, receiver and transport-storage system. The concentrator captures solar radiation and directs it to the receiver where the heat energy is absorbed by a fluid –normally a special type of oil. The hot fluid is then transported in a pipe to enable the heat to be used directly via a heat exchanger or stored for later use at night or during less sunny days. A popular design consists of parabolic troughs in long arrays of identical concentrating modules, resembling trough shaped glass mirrors that track the sun daily from east to west. They concentrate the solar radiation on to the absorber pipe located in the focal line of the installation. Parabolic dish collectors also track the sun on two axes but the system units are usually smaller consisting of a dish and a receiver unit installed at its focal point.

Total Site Targeting The optimal design of large scale utility systems that serve large chemical processing sites, such as refineries, is extremely complex. Total Site targeting is an established method that has been used extensively in industry. The methodology forms part of the procedure required to synthesise the entire total site system, which includes process utility and heat exchanger network design integrated with the site utility system. Total Site Targeting was put forward as a means of integrating heating and cooling requirements between individual process units in a total processing plant [11, 12]. The method produces targets for the amounts of steam used and generated by the combined individual processes, the amount of heat recovery through the steam system, the boiler demand for steam, and potential cogeneration [13]. The method also forms the basis for a target based design sequence for the overall site utility system that is required to meet heating and cooling demands through the steam based system and power requirements. The Total Site Targeting method allows waste heat from processes to be used as a source of heat in other processes. The waste heat sources are converted to steam, and then passed to processes that are in heat deficit through the steam system infrastructure. The resulting Total Site Profiles (Figure 1) therefore depict the total heat source available from all the contributing processes in the processing plant (the curve on the left of the diagram) and the total heat sink that has to be satisfied from external heat sources (the curve on the right of the diagram).


5

Figure 1. Total Site Profiles

To meet the heating and cooling requirements of the contributing site processes, steam mains have been supplied. For graphical simplicity the steam mains have been shown at their real temperatures, and because of the shift in process temperatures, feasible heat transfer is ensured. When a horizontal line representing the steam main temperature touches the Site Profile, the shift in temperature of the extracted process stream segments and the real temperatures of the steam mains means that heat exchange will take place at the appropriate minimum approach temperature. The steam mains are represented by their saturation temperatures. The use and generation of steam and cooling water for heating and cooling purposes are easily extracted from the Total Site Profiles and are given in Table 1.

Steam Main

Saturation Temperature

[°C]

Used load [MW]

Generated Load [MW]

VHP 320 2.78 0 HP 190 0.72 0.87 MP 160 0.6 0.45 CW 25 4.83 0

Table 1. Utilities and their parameters Heat integration can be achieved via the steam mains by making use of the steam generation from the Total Site source profile to satisfy the heating requirements of the Total Site sink profile. The potential is illustrated in Figure 2. The generation of steam at any level can be used to provide heating at the same temperature level or at a temperature below by letting down steam to the reduced pressure (with the potential of producing cogeneration and shown by the arrows in Figure 2).

6 4 2 0 2 6

100

150

200

250

350

4 Enthalpy [MW]

50

0

300

T [o C

]


6

Figure 2. Total Site Profiles with potential steam heat recovery Without heat integration between the sink and source profile the boiler would need to provide 4.1 MW of heat. Final integration, as shown in Figure 3, shows a resulting heat requirement from the boiler of 2.8 MW, a saving of 1.3 MW due to heat recovery through the stream mains. These targets, together with the potential cogeneration associated with the expansion of steam, can form the basis of a utility system design.

Figure 3. Total Site Profiles with steam heat recovery

Locally Integrated Energy Sector case study Total Site Targeting has shown that there exists a method of analysing the heat sources and sinks from more than one process, and how heat can be transferred from one process to another via a carrying medium, in this case steam [13]. A similar

10 8 6 4 2 0 2 4Enthalpy [MW]

100

150

200

250

350

50

0

300

Sources Sinks

T [o C

]

6 4 2 0 2 6

100

150

200

250

350

4Enthalpy [MW]

50

0

300

T [o C

] Sources Sinks


7

concept can be adopted for the analysis of heating and cooling requirements in an enlarged geographical area, which here is referred to as a Locally Integrated Energy Sector (LIES). In the LIES heat sources and sinks can be derived from small scale industrial plants, large building complexes (such as hospitals), offices, and residential dwellings. The following example illustrates the concept. There are two small scale industrial processing plants, coupled with a hospital complex and a group of residential dwellings and office complexes. The Site Profiles of the LIES is given in Figure 4. The heat sources and heat sinks of the contributing processes have been combined to produce the overall Total Site sink profile and Total Site source profile. Without integration the LIES would need to dispose of 6.2 MW of heat, and 17.5 MW of heat would have to be supplied from external heating sources.

Figure 4. Site Profiles for the LIES

A possible scenario (Scenario 1) for integration in the LIES is shown in Figure 5. In this a hot water main at a temperature of 60 °C to 40 °C is provided for extracting heat from the Total Site source profile and supplying heat to the Total Site sink profile. The amount of generated heat for the hot water supply is 5.5 MW, and the amount of heat that has to be supplied by the hot water is also 5.5 MW. The remaining heat required by the LIES is 12.1 MW, giving a heat reduction of 5.5 MW due to heat integration in the LIES. The external heat required could be provided by a boiler using a carbon neutral fuel such as biomass or waste combustion unit. In his particular case the heat for the hot water system is supplied by the small scale industrial processing plants, and the recipients of the heat are the hospital and residential and office complexes. If this heat was no longer available, then this heat (5.5 MW) would need to be supplied from another source. In order to reduce the carbon footprint of the LIES this heat would need to be supplied by a renewable source. The possibilities could include solar hot water heating systems, heat pumps, or a combination [14]. However, the costs of the supply and installation of these renewable sources, in addition to the hot water distribution system that has already been supplied, are required to be economic.

Enthalpy [MW]

T [o C

]

5 10 15 20 5 10 020 40

60

80

100

120

140

160 Sources Sinks


8

Figure 5. Scenario 1 – Total Site Profiles

Oversizing of boilers or waste combustion units already supplied is, with current energy prices, likely to be a more economic option in case of failure in the heat source of the LIES. However with the changing energy price levels / fossil fuel taxing the balance is likely to change. The Site Profiles offer a quick and robust assessment of these changing conditions. A second scenario, Scenario 2, is shown in Figure 6. In this case a steam main at 125 °C has been added to the system. The heat source from the LIES has now been split between steam (3.1 MW) and hot water (2.4 MW). On the sink side, 8.4 MW of steam is supplied to the hospital and residential and office complexes, and 5.5 W of hot water.

. Figure 6. Scenario 2 – Total Site Profiles

Enthalpy [MW]5 10 15 20 5 10 0

T [o C

]

20 40 60 80

100 120 140 160 180 200

0

Sources

Sinks

Enthalpy [MW]5 10 15 20 5 10 0

T [o C

]

20 40 60 80

100 120 140 160 180 200

0

Sources

Sinks


9

The amount of heat supplied by the boiler or waste combustion system remains at 12.1 MW. Again, economics is a significant factor in the final design potential. The cost of installing both a hot water main and a steam main over the geographical area of the LIES may be an economical option, however this needs to be appropriately evaluated in relation to changing energy/capital cost ratios.

Conclusions The concept of a Locally Integrated Energy Sector for the distribution of heat involving small scale industrial plants and domestic, business, and social premises, and integrating renewable energy sources has been demonstrated using the Total Site Targeting methodology. This method is able to provide the basis for an overall design of a Locally Integrated Energy Sector that involves both heat and power as conceptualised in Figure 7. In this case demands for heating/cooling and electricity in units (e.g. dwelling, offices, hospitals, or schools) can be met locally by renewable energy sources such as wind, solar cells, or heat pumps as well as by some excess heat and power available from local industry.

Figure 7. Locally Integrated Energy Sector with heat and power Additionally locally installed boilers, consuming traditional fossil based fuels, biomass or waste, can also help to meet these requirements, when demand is high or other sources are unavailable. Heating/cooling and power not required by one unit can be fed to a grid system, and then passed to another unit that is unable to meet its demands locally. The grid system can distribute power (electricity) and heating in the form of hot water or steam. In geographic locations where air conditioning is required, a cooling distribution main could also be provided. If local sources are unable to provide the demands for all of the units in the system, then district renewable sources can be provided. These again would include larger scale wind turbines, solar cell

UNIT 2 UNIT 1 UNIT 3 UNIT 4

Electricity

Electricity Steam

Hot water Cooling utility

Fossil Fuels

Bio-fuels

UNIT 6 UNIT 5

Bio-fuels Bio-fuels

Fossil Fuels Renewables

Fossil Fuels Fossil Fuels

GT

Nuclear


10

systems, large scale heat pumps, and combustors using waste provided by the units or fossil or bio-fuels. The sources at this level would include power generating equipment such as steam turbines or gas turbines.

Acknowledgements The financial supports from the EC projects EMINENT2 – TREN/05/FP6EN/ S07.56209/019886 and Marie Curie Chair (EXC) MEXC-CT-2003-042618 are gratefully acknowledged.

References 1. EC FP6 SHERHPA Project – Sustainable Heat and Energy Research for Heat

Pump Applications, sherhpa.fiz-karlsruhe.de (visited 17/04/2008). 2. Hjartsson H, Gullev L. Húsavík, Iceland – A Model of Energy Efficiency Based on

Geothermal Energy, www.iea-dhc.org/download/H%FAsavik.pdf, (visited 17/04/2008).

3. Hellisheiði Geothermal Plant www.or.is/English/Projects/HellisheidiGeothermalPlant (visited 17/04/2008).

4. Biomass Wood District Heating, Ignalina (Lithuania), Energy-Cities.EU, www.energie-cites.org/db/ignalina_140_en.pdf, (visited 17/04/2008).

5. A biomass district heating in Molins de Rei, Energy-Cities.EU, www.energie-cites.org/db/molins-de-rei_569_en.pdf, (visited 17/04/2008).

6. ALTENER Biomass District Heating Plants Guidelines, www.biomatnet.org/secure/Ec/S1078.htm, (visited 17/04/2008).

7. Faninger G. Combined Solar-Biomass District Heating in Austria. Solar Energy 2000; 69(6):425-435.

8. Energy Efficiency and Renewable Energy, US Department of Energy. 2008 www1.eere.energy.gov/biomass/fy04/fundamental_biomass_gas.pdf (visited 17/04/2008).

9. Energy from Waste: A Good Practice Guide, The Chartered Institution of Waste management, 2003.

10. Rantil M. Concentrating solar heat - kilowatts or Megawatts? Seminar “Renewable heating and cooling - from RD&D to deployment”, International Energy Agency, April 2006 www.iea.org/Textbase/work/workshopdetail.asp?WS_ID=243 (visited 17/04/2008).

11. Dhole VR, Linnhoff B. Total site targets for fuel, co-generation, emissions, and cooling. Computers and Chem. Engng. 1993; 17 (suppl.) :S101-S109.

12. Raissi K. Total Site Integration. PhD Thesis, UMIST, Manchester, UK, 1994. 13. Klemeš J, Dhole VR, Raissi K, Perry SJ, Puigjaner L. Targeting and Design

Methodology for Reduction of Fuel, Power and CO2 on Total Sites. Applied Thermal Engineering 1997; 7:993-1003.

14. Pavlas M, Stehlík P, Oral J, Šikula J. Integrating Renewable Sources of Energy into an Existing Combined Heat and Power System. Energy 2006; 31: 2163–2175.


11

Waste to Energy Technologies and Biomass Utilisation

Petr Stehlik Brno University of Technology, Institute of Process and Environmental Engineering

(VUT-UPEI), Technicka 2, 616 69 Brno, Czech Republic, Email:[email protected]

Abstract This paper is aimed at showing and demonstrating progress in technologies and improvements in units for the thermal processing of various types of waste and biomass. It represents a challenge for potential investors and operators as well as researchers. It is necessary to take into account the following needs: • Waste disposal as a necessity • Waste and biomass as renewables • Waste-to-energy as the main aim • Limitations given by environmental legislation Some concrete latest achievements demonstrating this approach and strongly influencing the progress will be presented (e.g. low NOx burners, efficient special heat exchangers and waste heat boilers for heat recovery systems, newly developed equipment for flue gas cleaning, dioxin filters, systems for the treatment of sewage sludge etc.). Utilizing of up to date computational methods as CFD (Computational Fluid Dynamics) for an optimization of design or troubleshooting will be demonstrated on examples (flue gas ducts).

Introduction Serious problems are currently solved in the field of environmental protection [1, 2]. The challenge facing concerned citizens and decision-makers is a formidable one: To identify and implement long-term solutions that are safe, socially acceptable, and cost-effective. Such amount of wastes, which is produced either by inhabitants (municipal solid waste - MSW) or by industrial companies (industrial and hazardous waste - IHW) requires to use efficient ways of waste disposal. The recent focus on incineration has been on environmental consequences, not on performance [3]. In particular, the limitations, as well as the advantages, of incineration are being increasingly recognized. At present, incineration is not a waste disposal method but rather a waste processing technology. Up to date incinerators, though simple in concept, are highly complex units. Incineration of waste can be considered as a certain form of recycling energy contained in materials taking into account that the energy was consumed during their

12

production. Then we speak about waste-to-energy technology (WTE). WTE is also referred to as thermal processing of wastes including energy utilization. It means not only combustion of various types of wastes (incineration) in order to substantially reduce their volume but besides that, WTE systems can provide us with clean, reliable and renewable energy. What’s more, these sources can be easily integrated into existing power and heat distribution systems (district heating system, chemical and processing industry etc.).

Examples of progress in thermal processing of waste Progress in the field of thermal processing of wastes strongly depends on an efficient performing research and development consisting in convenient combination of theoretical approach utilizing sophisticated methods, experimental approach and feedback from industrial units including related know how. This approach can be called “from idea to industrial application” and can be specified as follows:

• Idea and basic design taking into consideration industrial demands • Design of experimental facility • Manufacturing the equipment, building the research facility • Putting into operation and testing functionality • Data acquisition for improving and validating the parameters of the design

calculation procedures and of the CFD model in case of equipment • Design of practical alternative arrangements following various

requirements coming from industrial practice • Solving industrial cases • Industrial applications • Feedback for further research

Recent achievements demonstrating this approach in the area of thermal processing of waste are introduced in the following sections.

Low-NOx burners Special low-NOx burners, firing typically natural gas (alternatively also mining gas, see e.g. reference [4]) are applied in secondary combustion chambers. The formation of nitrogen oxides, which belong among the most closely monitored pollutants, depends on the type of fuel, temperature and flow fields in the combustion chamber. When fuel does not contain any bound nitrogen, then the prevailing type of NOx is so-called thermal NOx. However, the rate of its formation is first of all substantially influenced by design of the burner. Therefore experience and know-how play a very important role. One of the possible ways how to decrease emission levels of nitrogen oxides is to use burners designed specifically to reduce their formation. In principle, the main idea in designing these burners is to decrease and homogenise temperature in the flame and to lower the amount of excess oxygen. Standard measures to achieve this goal include staged air and fuel supply and in some cases also in-burner flue gas recirculation. Specific burner design depends mostly on the fuel to be fired, requirements on the character of the flame (shape, length, width, momentum), and on the application

13

(boiler, secondary combustion chamber, etc.) that means first of all awaited conditions predominating in the furnace. Development of low-NOx burners is an art on its own; for more information see e.g. publications [5, 6]. An example of a novel design of low-NOx burner for firing natural/mining gas is shown in Figure 1.

Combustion air

Natural gas

Swirl generator

Mining gas nozzle

Mining gas

Second stage nozzles

First stage nozzles

Second stage combustion air

First stage combustion air

Figure 1 Dual low-NOx burner

Heat recovery system Heat recovery in units for the thermal processing of various types of wastes can be without any doubts considered as one of the most important parts of these processes. Design of equipment for utilization of energy contained in the flue gas from the thermal treatment of wastes (incineration) and their placement in the process is one of key factors in these technologies. In design and operation of heat recovery systems it is necessary to take into account characteristics of heat transfer equipment and their specific features as well as those of process fluids. The most common way of utilization of off-gas energy is steam generation in HRSG. Various types and arrangements are used in industrial practice - depending on a specific application and know-how. Selection of a convenient type of heat recovery steam generator (HRSG) depends on operating conditions and capacity of incinerator (throughput of waste). Often, special types of radiant recuperative heat exchangers must be used, preventing fouling and enabling easy access for mechanical cleaning. Using specialised database [7] for correct selection of suitable type of heat exchanger, taking into account specific features of the process fluids, can help to find optimum solution and to avoid possible serious operating problems (e.g. excessive fouling, thermal expansion, leakages). Example of specialised database for hot gas applications (HGA Database) [7] is shown in Figure 2.

14

Double-Pipe

Segmental baffles

Plate-Type

Double U-tubes

Sludge aplications

Simple

With regenerative layer

HGA DATABASE

Radiation recuperator

Helical baffles

Rod baffles

Heat-Pipe

Coaxial heat exchanger

Conventional types Special types

Shell-and-Tube

Orifice baffles

Twisted tubes Water – Sludge

Flue gas - Sludge

Figure 2 HGA Database – window for selection of suitable heat exchanger type

Dioxin filters Dioxins (collective name for polychlorinated - dibenzo-p-dioxins and dibenzofurans) belong among the most dangerous pollutants being considered as extremely toxic substances. Their elimination from the products of incineration (solid residues, flue gas) is vitally important. It is therefore very attractive to employ technologies that actually destroy dioxins instead of only collecting them. The latest achievement in this respect is the catalytic filtration technology REMEDIATM of the company W.L. Gore & Associates Inc., which combines fabric filtration (collection of particulate matter) with catalytic destruction of dioxins. Information on the performance of this technology recently applied in a municipal waste incinerator may be found e.g. in [8].

Flue gas cleaning systems Within the framework of research on cleaning of various off-gases, polluted by different classes of compounds, the following pollutants have been addressed:

• Volatile organic compounds (VOCs), halogenated organic compounds (HOC) and CO

• NOx, SO2, HCl, HF • Particulate matter (fly ash) and PCDD/F (dioxins and furans)

For the treatment of off-gases polluted by these three groups of pollutants, three novel approaches have been developed, respectively (see Figure 3):

• Equipment for thermal and catalytic treatment of waste gases containing combustible compounds

• Equipment for wet scrubbing, the so-called “O-element” homogenizer, used for the treatment of flue gases containing acidic compounds and heavy metals

15

• Equipment for combined process filtration / adsorption (chemisorption) – INTEQ

Description of these novel units may be found in a paper [9].

Figure 3 Progress in waste gas processing: a) Equipment for thermal and catalytic treatment of waste gases, b) Equipment for combined process filtration/adsorption

(chemisorption) – INTEQ and c) Equipment for wet scrubbing, the so-called “O-element” homogenizer

Thermal treatment of sewage sludge Sewage sludge (the residue of waste water treatment) makes an appreciable part of produced wastes. Utilization of sewage sludge as alternative fuel for cement and lime production industry is an example of material and energy utilization of waste. Co-firing in cement works is counted as wasteless technology. High temperature in the cement kilns and suitable retention time secures decomposition of all organic compound (including polychlorinated - dibenzo-p-dioxins and dibenzofurans). Ash from sludge is bounded to cement clinker due to its similar composition as other raw materials have. Waste heat of the process can be used for sludge drying [10], which is another benefit of this way of sludge treatment. Sludge incineration can be a complementary solution also for waste water treatment plants with large processing capacity. Two types of sludge are possible to incinerate in principle. The first one is mixed raw sludge, which is a mixture of primary sludge and activated excess sludge. The second one is digested sludge. This solution of sludge disposal has environmental benefit, because sludge is not land-filled as it is common in many countries. Incineration of suitably prepared sludge (mainly sufficient dewatering) can also be economical, when energy of off-gas is used for heat and/or electric power production. Ash can be used as a secondary building material (e.g. for road construction) as well, but its utilization depends on its quality.

a) b) c)

16

Computational modelling and optimization CFD is a very powerful tool that helps to provide insight in complex fluid flow phenomena not easily amenable to intuitive solutions. Its uses are multiple, including analyses of waste-to-energy processes. More on the application of CFD in thermal waste treatment technologies may be found e.g. in [11]. Generally, CFD may be applied to provide insight in cases of troubleshooting, it may be used to provide basis for selection of better alternatives during the design of new or modified devices (optimisation) and last but not least, it may provide both interesting and beautiful illustrations useful for marketing of new, better products. Improvements of performance and cost savings are at the focal point of any and all process equipment designers. Therefore, it is natural that equipment so widely used as plate type heat exchangers has attracted much attention of researchers over a large period of time. In spite of this fact, there are still new possibilities in obtaining better performance from this popular process equipment as shows recent publication [12]. Another example how CFD can be utilised for flue gas duct optimization is demonstrated in Figure 4 [13]. The flow pattern in this duct leads to fouling in a connected heat exchanger. CFD analysis is used to find what causes the fouling and to optimise the duct design in order to eliminate the undesirable phenomena.

Inlet (outlet from air preheater )

Heat exchanger “flue gas-water”,

Duct expansion element (with water injection nozzles)

Outlet (inlet into stack fan)

Manually optimized flow homogenizing swirl generator

Figure 4 Geometry of exhaust duct with heat exchanger

17

Conclusions The aim of this paper was to show the latest achievements in the field of thermal processing of various types of waste (municipal, hazardous, sewage sludge) and biomass. The thermal treatment units are included into Waste-to-Energy system, which can provide us with clean and reliable energy. Presented examples are related to key equipment of these units. Some introduced results of research and development have been implemented into industrial applications. A sophisticated approach (CFD, optimization) and its usefulness or even necessity in improved design was also demonstrated.

Acknowledgements We gratefully acknowledge financial support of the Ministry of education, youth and sports of the Czech Republic within the framework of research plan No. MSM 0021630502 "Waste and Biomass Utilisation focused on Environment Protection and Energy Generation".

References [1] Ludwig Ch., Hellweg S., Stucki S., Municipal Solid Waste Management,

Springer-Werlag, Germany, 2003 [2] Kiely G., Environmental Engineering, McGraw Hill, UK, 1997 [3] Kuehn T.H., Ramsey J.W., Threlkeld J.L., Thermal Environmental Engineering,

Prentice-Hall Inc., USA, 1998 [4] Kermes V., Stasta P., Sikula J., Oral J., Martinak P., Stehlik P., Substituting Fuel

by Mining Gas in Unit for Thermal Treatment of Sludge, 22nd Annual International Conference on Incineration and Thermal Treatment Technologies, Proceedings on CD ROM, Orlando, Florida, USA (12-16 May 2003)

[5] Baukal C. E., Industrial Burners Handbook, CRC Press LLC, USA, 2004 [6] Kermes V., Skryja P. and Stehlik P., Up to date experimental facility for testing

low-NOx burners, 10th Conference on Process Integration, Modelling and Optimisation for Energy Saving and Pollution Reduction PRES 2007, Proceedings on CD ROM, Ischia Porto, Italy (June 24 – 27, 2007)

[7] Kilkovsky B., Pavlas M., Jegla Z. and Stehlik P., Database HGA: Common and Specific Types of Heat Exchangers for Hot Gas Applications, Heat Transfer in Components and Systems for Sustainable Energy Technologies: Heat-SET 2007, Proceedings on CD ROM, France (April 18-20, 2007)

[8] Xu Z., Chang W., Sun H., Mongami Y., Application of a catalytic filter system in waste incinerators to meet dioxins emission regulations, 1st International Conference & Exhibition on Thermal Treatment and Resource Utilization of Wastes, Beijing, 2005

[9] Dvorak R., Bebar L., Stulir R., Stehlik P., Waste gas treatment and off-gas cleaning, 1st International Conference & Exhibition on Thermal Treatment and Resource Utilization of Wastes, Beijing, 2005

18

[10] Stasta P., Boran J., Bebar L., Stehlik P., Oral J., Thermal Processing of Sewage Sludge, Applied Thermal Engineering 26 (13) 1420-1426, 2006

[11] Hájek J., Computational fluid dynamic simulations in thermal waste treatment technology—Design, optimisation and troubleshooting, Energy, 2008

[12] Jegla Z., Stehlik , P, Kohoutek J., Alternative Approach in Optimization of Plate Type Heat Exchangers, Heat Transfer Engineering 25, pp. 6–15, 2004

[13] Hajek J., Kermes V., Sikula J. and Stehlik P., Utilizing CFD as Efficient Tool for Improved Equipment Design, Heat Transfer Engineering 26, pp. 15 – 24,2005


19

Biofuels: Thermodynamic Sense - and Nonsense Jack W Ponton

School of Engineering and Electronics, University of Edinburgh Edinburgh EH9 3JL, Scotland

e-mail: [email protected]

Abstract Much of the current enthusiasm for biofuels amongst politicians (who cannot be expected to know any better) and academics (who should) appears to ignore basic themodynamic and other constraints. The fundamental problem with growing fuel is that combustible plant matter is almost invariably solid, while the major demand for energy at present is in the form of gas or liquid fuels. All current conversion processes are of low efficiency even for the convertible parts of the plant: for example the energy which could be obtained from burning a kilogram of wheat grain is about twice that available from the ethanol into which it can be converted by fermentation. Furthermore most liquid fuel processes can use only part of the plant. This paper will seek to identify those biofuel technologies which make sense, such as co-firing straw with coal in power stations, and those which thermodynamic considerations show to be nonsense, such as making ethanol from grain in Europe. Since arable land is a scarce resource in most of Europe, biofuels are unlikely to become a major replacement for fossil fuels. We will suggest strategies which will help to maximise the contribution which they could make, and promising emerging technologies where it would make sense to concentrate research.

Introduction Over the last year or so the popular, and popular scientific, press has been full of articles about biofuels: biodiesel and bioethanol. The US has a major, and highly subsidised, corn to bioethanol programme which, enthusiasts claim, will produce fuel at a cost equivalent to 60 cents a US gallon, about 8p/l. The EU has, without apparently consulting informed scientific sources, imposed an `obligation' for us all to produce 5% of vehicle fuels from biological sources by 2010. However, more recently a note of scepticism, or realism, seems to have crept in. Some articles have suggested that the net energy yield from maize ethanol, once all energy costs are taken into account, is very low or even negative. Concerns have been raised that the EU's biofuels are not all being grown on CAP `set-aside' land (we note below that this would never have been possible) but are encouraging palm oil growers in Asia to destroy rainforest to grow oilseeds for export. It has also been suggested that using synthetic fertilisers to grow fuel crops in fact releases more damaging greenhouse gases, namely nitrous oxide, than the fuel produced can save in carbon dioxide.

20

In fact, most of these issues can be understood by elementary thermodynamic analysis, and indeed studies already exist which confirm that many of the concerns and reservations which have been expressed are entirely valid.

Is it Worth Growing Crops for Fuel? The answer to this question has to be `yes' unless one can devise some other long term source of energy. The alternatives, briefly are as noted below.

• Hydrocarbon resources are ultimately limited, and many of them are controlled by unpleasant and unreliable regimes.

• Nuclear is a pretty good bet for countries with effective safety standards, but hysterical politicians have made it hard to develop. Fusion, despites billions spent in research, is pie in the sky and probably always will be.

• Hydropower is limited in availability. using present techniques, geothermal is even more so.

• Wind and wave are becoming more plausible, but still have a long way to go to become seriously competetive without subsidies.

• Solar is a probably a good way of keeping cool, e.g. running air conditioning in hot countries, but pretty useless for keeping warm in cold ones.

Furthermore, nothing other than chemical fuels as yet provide portable energy sources adequate for the kind of transport which societies now expect. Biofuels are chemical fuels and so can, in principle, fill this bill. Agriculture, properly managed, is environmentally sustainable. It is fundamentally a way of capturing solar energy. Per square metre, it is much cheaper than any manufactured solar collection device. However, overall efficiencies are very low, typically fractions of one percent to usable chemicals. So biofuels are necessary, and so growing them has to make sense. However, this answer has to be qualified by consideration of what kind of fuel to grow, how to use it and where to grow it.

Initial Considerations In this discussion we shall consider primarilyy the UK. However, very similar conclusions should result from an analysis of the options for any part of the temperate zone. First, it is helpful to have an idea of the scale of our energy requirements. In 2005 the UK [1] used, reducing all energy sources to million tonnes of oil equivalent (Mte), 247 Mte, approximately in the following forms:

• 40 Mte of coal, • 90 Mte of oil, • 90 Mte of gas and • 20 Mte of `primary' electricity (nuclear, hydro and a tiny amount of windpower).

In terms of energy content, 1 te oil equivalent is about 42 GJ or 11.7 MWh.

21

Definitive energy balances for biofuel production are thin on the ground. However a report with estimates for biodiesel and ethanol in the UK was produced by Levington Agriculture [2] for BAFBO, an organisation representing producers of oilseeds and biofuels. A comparison of this and other estimates may be found in [11]. A hectare of arable land used to produce biodiesel would yield about 54 GJ worth of fuel, or about 74 GJ if used for grain based ethanol. So as biodiesel, one would get 1.28 te of oil equivalent. This is a bit less than the 1.5 te of biodiesel which the report predicts. To quickly put things in context, the total land area of the UK is 241,590 km2 or about 25 Mha. This includes cities and mountaintops, but if all of it could be used to grow biodiesel, it would still produce less than one fifth of the UK fossil fuel consumption. In fact the total amount of arable land in the UK is about 5.5 Mha. If we put all of this down to biodiesel, we would replace at best 8.25 Mte of oil, less than 10% of our total oil consumption! Ethanol yields are somewhat higher, but it has been estimated that growing grain for ethanol on all this land would provide only the equivalent of about one third of the UK's current gasoline (petrol) requirements. It seems unlikely that more than, say 20% of agricultural land could be divertied to fuel use, so we are looking less than 1% of the national energy requirements. Furthermore, these are the gross energy figures. The energy required to produce and process the crop is more than 50% of the energy available from biodiesel and nearly 90% for ethanol - so in practice nothing like this proportion of our energy requirements would be replaced. We can immediately draw some conclusions.

1. Liquid biofuels grown in the UK can only have a very minor impact.

2. If biofuels are going to have significant effect on the UK energy economy, they must either:

o have very much higher per hectare yields than above, or

o be grown on non-agricultural land, or

o be grown elsewhere than in the UK. Others [5] have pointed out the impossibility of producing the EU's biofuel `obligation' on set-aside land, and in general confirm these conclusions on an EU-wide basis.

What Kind of Biofuels? The next most important thing to appreciate is that chemical transformations of fuel from one form to another normally either involve a reduction of the energy value of the material or the input of energy, or both. Consider the figures from [2] for ethanol from wheat. For each 1 te of grain we get 0.276 te of ethanol. The calorific value of the grain is about 17 MJ/kg, so if the grain were burned it would produce 17GJ of energy. However, the energy value of the ethanol is only 8.3 GJ, a loss of more than 50% without even taking into account the

22

processing costs! In fact the energy value of the wheat straw is 10.9 GJ, greater than that of the ethanol. If the entire plant were used as solid fuel it would yield nearly 28 GJ of energy instead of 8.3 GJ and the production cost would be substantially less, so the net energy yield would be 4 to 5 times greater. Plant products are normally solids. It is therefore most efficient to use them in this form unless transforming them leads to increased efficiency elsewhere. In this case it is by no means obvious that i.c. engines running on ethanol will be more efficient than, say, power stations fuelled with straw. The conclusion here is that biofuel crops have to a) allow the whole plant to be used and b) involve minimum additional processing. A possible exception to this might be gasification. Existing well-established coal gasification technology can be used for biomass on both large and small [20] scale. The gas could be used in a gas turbine which does have a higher thermodynamic efficiency than a steam plant

A Glycerol Lake? It is worth pointing out that the UK Government's preferred biodiesel strategy, as well as involving unnecessary processing and energy losses, suffers from a further drawback. Although it is perfectly possible either to use straight vegetable oil [10] (called SVO or pure plant oil, PPO) after only physical processing in slightly modified diesel engines, or to process it along with petroleum oils in existing refineries, a more complex process is being promoted by tax breaks. The vegetable oil is reacted with methanol to produce methyl ester, see e.g. [18]. This is about one third of the molecular weight of the oil and so can be used in an unmodified engine. In the reaction below R has typically 20 carbons. C3H5(COOR)3 + 3 CH3OH = 3 CH3COOR + C3H5(OH)3 The byproduct, glycerol, C3H5(OH)3, is a potentially valuable chemical used in the food, cosmetic and pharmaceutical industries. However, it is not a bulk chemical like methanol or diesel fuel. In fact the US market (2000) for glycerol was around 200 kte/yr [8] and was less than installed capacity. Byproduct glycerol from 5% of the UK diesel market alone would provide an additional 150 kte/yr. Excess glycerol from European biodiesel production by 2000 had already dropped US by up to 45% in the late 1990s.

Byproduct Biomass It is perhaps understandable that the US is obsessed with transportation fuels, but of the 247 Mte oil equivalent used in the UK in 2005, just over one fifth, 59 Mte was used for transportation [1]. This is comparable with 51 Mte used in electricity generation, of which 37 Mte is gas and petroleum, both of which could alternatively be used as transportation fuels.

23

At present the UK has about 1.9 Mha of wheat and 1 Mha of barley [3]. The straw from this is essentially a waste product. More than 40% [4] is currently ploughed in as it has some minor value as a fertiliser. Taking the the heat value of wheat straw [2] to be applicable also to barley suggests that the UK has at the present the potential to produce (1.9 x 97.5) = 185 million GJ of energy from waste straw. This corresponds to 4.4Mte of oil equivalent to which could be added the straw from 0.58Mha of current rape production, at 60 GJ/ha, a further 0.8Mte, giving a total of 5.2Mte is equivalent to nearly 10% of the energy used in transportation without reducing the land area available for food production. Of course, there are other issues. Straw is produced over a limited period, although conveniently this is in the autumn, before the winter peak of energy demand. It has to be transported, stored and processed, and the technology of efficient combustors developed. At least one coal burning power station, in Denmark [12], has successfully `co-fired' straw. The last of these problems would appear to already be solved. In Scotland Cockenzie power station, convenient to the agricultural are of East Lothian, is in fact already co-firing byproduct biomass. However, this presently is olive oil waste from southern Italy. It is hard to see the justfication of transporting a low value fuel more than half the length of Europe.

Biomass Fuel Crops The relatively high energy content of wheat straw, a plant that has been bred for thousands of years to maximise its grain yield, suggests that there may be other species which could be grown to produce still larger amounts of biomass cellulose. This is indeed the case. Grasses, reeds and some rapidly growing softwoods such as willow and poplar [17], are all crops which could yield biomass in comparable or greater quantities. A particularly promising plant is Miscanthus [14] which offers high yields in most of the UK. A recent study in the US [21] claims that switchgrass, a fast growing perennial could produce more than five times the energy required to grow it. However, this appears to only marginally better than for oilseed rape for which the factor is about 4.5 [2]. (The yield energy figure in fact applies to a notional and as yet nonexistent process for conversion to ethanol, with uspecified costs and efficiency!). Effective fuel crops must share the following characteristics:

• They are fast growing. • They are perennial, and thus once planted do not require an annual input of

fuel for ploughing, which requires typically around 10% of the crops annual gross energy content.

• They can be grown on marginal rather than prime agricultural land. • They require minimal fertiliser.

Provided a strategy for using them directly in power generation can be developed, such plants are possible energy sources to replace fossil fuels. Furthermore, these are plants which have largely evolved naturally. Applying modern plant breeding

24

methods, especially genetic modification, could increase yields still further. One might hope that public hysteria about GM plants will be assuaged if they are to be used for fuel rather than food. Power generation companies are already carrying out a significant number of tests with biomass fuels, normally co-fired with coal [13]. In the UK, Drax, a large UK coal fired station, has tested co-firing with willow [15]. The above calculation for the total energy available from wheat on a per hectare basis, with a yield of 8.96 te of grain per ha, implies an energy yield of 6.2 te oil equivalent per hectare. Assume that genetic modification of selected varieties of fuel crop could increase this by about 50% to around 9 te/ha, and that land equivalent to 30% of the current UK arable total were to be used for fuel crops. This could be comprised of 10% set-aside, 10% diversion and 10% of additional marginal land suitable for fuel but not food. The UK could produce about 0.3 x 5.5 x 9 = 14.8 Mte of fuel crops. Add to this the energy available from straw and other agricultural byproducts and a figure of 20Mte is plausible. This could either replace about two thirds of the coal used in power generation or two thirds of the oil and gas. Replacing coal, the worst fuel from a CO2 emission standpoint, would result in a significant reduction in the UK's emissions, providing politicians and concerned members of the public with a warm green glow and meeting everyone's `targets' for greehouse gas reduction. The practical significance of of this, given that China will be continuing to add several times this amount every year is perhaps questionable. Replacing oil and gas would free these fuels for use in transportation. This would be both ecomomically and environmentally more efficient than making liquid fuels from biomass using any presently available technology.

Prospects for Liquid Fuels As every chemical engineer knows, fluids, particularly liquids, are much more convenient than solids. It would be satisfactory if we could develop efficient routes from biomass to liquids, and indeed significant effort has gone into the investigation of improved pathways. Much work has focussed on conversion of cellulose to ethanol. This would enable all of a plant rather than just its sugar or starch content, to be converted. Efficiencies are still too low for this to be a realistic process. Development of the 1919 Weizmann process for bacterial conversion of carbohydrates to acetone to produce butanol have been investigated since the 1980s. Recent developments, e.g. [6] claim good yields and DuPont [7] claim to have a viable commercial process. Although precise information is hard to come by, it may be that conversion of cellulose using Clostridium bacteria is possible by this route. If

25

so, this is certainly a better propect than ethanol, since butanol may be more readily mixed with gasoline or used as a straight replacement without engine modification. However, it all hinges on the efficiency of conversion of biomass. Structured plants are very inefficient converters of sunlight to biomass. Much less than 1% of the incident solar energy is released even if the entire plant is used. However, algae, which are much simpler organisms display greater efficiency, and some of them even produce hydrocarbons. Their faster growth rates, and the fact that their production can be essentially continuous rather than seasonal, could make relatively low efficiencies of hydrocarbon production acceptable [16]. Moreover, dried algae are a fine particulate material which could be slurried with liquid fuels and used directly in suitably designed burners or even i.c. engines. However, present technology does not offer an efficient route to liquid fuels. Much more research and development is needed.

Grow it Elsewhere? The basic problem with fuel crops in northern Europe is that, since plants are inefficient converters of sunlight, when there isn't much of this, i.e. for about half the year, little growth takes place. In the tropics, on the other hand, it is possible to grow two or three crops a year. From the standpoint of ecomomics, sugar cane grown for ethanol in e.g. Brazil, appears to make sense. Bioethanol in Brazil is actually sold at a competitive price without any overt subsidies. However, there must be provisos. If the whole plant is not used then the process is inefficient from either a thermodynamic or environmental viewpoint, or both. Bagasse, the residual plant material from sugar cane, is traditionally burned to provide energy for refining the raw sugar, and it is claimed that it can be used to provide all the energy required in ethanol production. What are the long term effects of intensive sugar cane cultivation? There is evidence that it can be severely damaging to the environment. Is cane cultivation really not competing with rainforest? Even if it isn't now, one must fear that if biofuels become a lucrative industry, it will in the future. Similar arguments apply to e.g. palm oil production in Malaysia and Indonesia. Indeed the developed world's enthusiasm for biofuels may seriously threaten such countries [19].

Waste Not? The UK produces more than 30 Mte/yr of domestic waste. Most of this has gone to landfill, occupying space near cities and towns with which the country is by no means well supplied, and producing, by anaerobic fermentation to methane, a significant contribution to the greenhouse effect. There are various ways in which this material can be used as an energy source, the simplest being just to burn it. It is fairly simple to remove glass and metal, and most

26

of what remains, about 85%, although not high grade fuel, is combustible, and in 2000 about 6.5Mte was in fact combusted to produce heat and/or power [24]. There are both practical and political difficulties with combusting a larger proportion. Practically, it is a poor fuel, and, with a significant amount of animal material from food waste, may need carefully designed combustors to avoid the production of toxic byproducts. EU directives, for no good scientific or economic reason, also limit the amount of waste which may be combusted to 25% of the national total. The public also dislike the idea of `incinerators'. However, waste includes 8-10% of plastics, much of which is already separarated by consumers and collected separately. At present this is recycled, in some cases after a complex, expensive and not very efficient further separation, into low value, low grade material with limited uses. Although it is possible in principle to depolymerise waste plastic to produce a syncrude which can be used as a chemical feedstock, it is questionable as to whether this is worthwhile. Of the 180Mte of oil and gas consumed in the UK in 2005, only about 10Mte was used as a chemical feedstock. Much more sensible would simply be to use waste plastic as high grade fuel in existing power stations. Since mixed plastic waste is traded at well below its fuel value this makes economic sense. Since it will substitute for new oil or gas it also makes environmental sense. That the most efficient way to recycle plastics is as energy has been known for more than 10 years [25]. There are other proven technologies, such as anaerobic fermentation to methane for fuel which could be more widely used, as well as emerging technologies of algal digestion which can be applied to waste as well as new biomass.

Conclusions When work on this paper was started (January 2007) the general public and political climate was strongly and uncritically in favour of biofuels. Even from the outset of the work, the author was rather more sceptical [9]. Since that time, a number of better informed sources, e.g. [19, 22] have sounded serious notes of caution. The UK Parliament's Environmental Audit Committee [23] has called for the EU to abandon its renewable transport fuels directive. In addition to the purely thermodynamic and logistical issues considered here, there are concerns about the net balance of greenhouse gases. Nitrogenous fertilisers can produce nitrous oxide, and ploughed in biomass methane. Both of these produce many times the greenhouse effect of carbon dioxide. Even if biological energy sources are not a panacea, they are potentially of significant value. However it is important to take all factors into account, and choose appropriate rather than fashionable strategies. We will end with a summary of what we think makes sense, or otherwise, in the development of this type of energy.

27

Nonsense: • Attempting to produce ethanol from grain in northern climates. • Diverting useful arable land to the production of fuel oilseeds. • Making special `biodiesel' rather than using vegetable oil more directly.

Questionable: • Encouraging large scale production of biofuels in Brazil, Indonesia, etc. • Expecting biofuels ever to support transport at its present scale.

Sense: • Burning agricultural byproducts such as straw, and plastic waste, in existing

power stations.

• Development of solid biomass technologies.

• Growing rapeseed for oil on set-aside to blend with mineral oils.

• Carrying out research on:

o genetic modification to produce high biomass yielding plants for marginal land,

o algae etc. as a source of biomass and

o processes for effficiently converting lignin and cellulose to liquids.

• Practically any energy conservation policy. Investment in e.g. building insulation yields better returns than any renewables generation or fuel development.

References [1] "Energy Consumption in the UK", DTI publication: www.dti.gov.uk/energy/statistics/publications/ecuk/page17658.html [2] IR Richards, `Energy balances in the growth of oilseed rape for biodiesel and wheat for ethanol", Report for BAFBO, June 2000. www.senternovem.nl/mmfiles/27781_tcm24-124189.pdf [3] UK Agriculture web site www.ukagriculture.com/index.cfm [4] www.nnfcc.co.uk/nnfcclibrary/cropreport/download.cfm?id=10 [5] M Frondel and J Peters, `Biodiesel, a new Oildorado?', Energy Policy 35, 1675-1684, 2007 [6] DE Ramey, US patent 5,753,474, 1998 [7] DuPont website www2.dupont.com/Biofuels/en_US/ [8] www.the-innovation-group.com/ChemProfiles/Glycerine.htm [9] JW Ponton, `Biofuels, Solution or Snake Oil?' seminar to IES, University of Edinburgh, 2007. [10] The SVO Users and makers Association www.bio-power.co.uk/svoa/index.htm [11] AP Armstrong et al, `Energy and greehouse gas balance of biofuels for Europe', report 2/02 for CONCAWE, Brussels, 2002 www.senternovem.nl/mmfiles/26601_tcm24-124161.pdf [12] P Overgaard et al, `Two years operational experience and further development of full scale co-firing of straw', 2nd World Conference on Biomass for Energy, Industry and Climate Protection, Rome, 10-14 May 2004

28

[13] WR Livingston, Àdvanced biomass co-firing technologies for coal fired boilers', keynote at International Conference on Coal science and Technology, Nottingham, Sept 2007 [14] MAFF report NF0403 Miscanthus Agronomy: www.defra.gov.uk/farm/crops/industrial/research/reports/rdrep14.pdf [15] Drax Power Ltd web site, www.power-technology.com/projects/drax/ [16]NERC Knowledge Transfer Project web site: www.bluemicrobe.com/home.htm [17] `Yield Models for Energy Coppice of Willow and Poplar', Energy crops and biofuels web site: http://forestry.gov.uk/srcsite/infd-5l8hhr [18] V Hofman, `Biodiesel Fuel', AE-1240, Feb 2003, www.ag.ndsu.edu/pubs/ageng/machine/ae1240w.htm [19] R Clift and Y Mulugetta, À plea for common sense (and biomass)', TCE 24-26, October 2007 [20] Ankur Scientific, India, sell small scale biomass gasifiers: http://ankurscientific.com/ [21] M. R. Schmer, K. P. Vogel, R. B. Mitchell, and R. K. Perrin, `Net energy of cellulosic ethanol from switchgrass' Proc Nat Acad Sci, 105, 2, 464-469, 2008 www.pnas.org/cgi/content/short/0704767105v1 [22] J. Picket, et al, `Sustainable biofuels: prospects and challenges' The Royal Society, London, 2008 royalsociety.org/document.asp?id=7366 [23] Àre biofuels sustainable?', Report by UK Parliamentary Environmental Audit Committee, 2008, www.parliament.uk/eacom/ [24] Ìncineration of household waste', UK Government Post Note 149, December 2000 [25] W Reid Lea, `Plastic incineration versus recycling: a comparison of energy and landfill cost savings', Jl of Hazardous materials, 47, 295-302, 1996

ENERGY FOR SUSTAINABLE FUTURE Edited by Petar Varbanov, Jiří Klemeš, Igor Bulatov University of Pannonia, Veszprém, Hungary, 5-6 May 2008 Copyright © Manuscript Authors, ISBN 978-963-9696-38-9E

29

Automated Targeting for Segregated Energy Planning Denny Kok Sum Ng1, Dominic Chwan Yee Foo1, Raymond R. Tan2

1University of Nottingham Malaysia, Broga Road, 43500 Semenyih, Selangor, Malaysia, email: [email protected]; [email protected];

Tel: +60 3 8924 8130; 2De La Salle University, 2401 Taft Avenue, 1004 Manila, Philippines,

email: [email protected];

Abstract Due to growing international concern about climate change, the management of carbon dioxide emission from different human activities has become very important. Many countries now seek to control their carbon dioxide emissions by having proper energy planning. This paper presents a novel automated targeting approach to determine the optimum allocation of energy sources in a segregate planning scenario where energy targets for different sub-sectors in the planning horizon are located.

Introduction There is now a strong consensus in the global community that emissions of greenhouse gases such as carbon dioxide (CO2), methane (CH4) and nitrogen dioxide (NO2) are causing global warming or climate change. Such changes are projected to cause a rise in average global temperature over the coming decades; at the same time, potential changes in local climatic conditions pose different risks for human communities that rely on predictable weather for their sustenance. In responding to the global warming issue, many countries have signed and ratified the 1997 Kyoto Protocol [1] which is aiming to reduce of the emission of greenhouse gases by 5% of their 1990 emission levels. Furthermore, preparations are under way for a new international agreement when the Kyoto Protocol expires in 2012. Thus, energy planning is crucial at both national and regional levels. Different planning tools, usually based on optimisation models, are often used by governments to meet emission limits when satisfying energy demands for different geographic or economic sectors at reasonable costs. Pinch analysis was developed since the 1970s as a systematic technique for the synthesis of heat exchanger network for process plants. This technique is based on thermodynamic principles and its basic concept is to maximise energy recovery by matching the available internal heat sources with the appropriate heat sinks, which eventually leads to the minimum usage of external heat sources [2-4]. Analogy between heat and mass transfer in thermodynamic principles lead to the introduction of mass integration, which is widely used in the areas of resource conservation, e.g. solvent, water and utility gases [5-18]. Recently, novel applications of pinch analysis in non-conventional areas such as production planning [19-22], financial management [23], and property-based material recovery network [24-25] have also been reported. All of these applications shared a

E

30

common principle, which is making use of information about stream quantities (flowrate) in conjunction with their data on quality (temperature, concentration, etc.). Emission targeting by pinch analysis was previously reported in the framework of total site analysis [26-28]. Total sites in these early works refer to industrial facilities that incorporate several processes which are serviced by a central energy utility system. Although emissions targeting by pinch analysis was introduced in those studies, the early applications were limited specifically to optimisation within industrial facilities, and not to regional or national energy sectors. The latter application covers broader geographic and temporal scales, and also includes different energy demand sectors, such as residential consumption, transportation and industry. More recently, Tan and Foo [29] addressed this latter application by utilising the pinch-based energy planning composite curve in locating the minimum amount of zero-carbon energy source during energy planning. However, this early work was restricted to cases where mixed low- and zero-carbon energy sources are used to achieve the overall emissions target for a specific country or region, without having to differentiate the various carbon sources. Note that, in practice, energy sources can only approach carbon neutrality. For example, even though biomass may be viewed as a potentially carbon-neutral or zero-carbon source in the life cycle analysis perspective, its production and use can still generate small amounts of net CO2, or other greenhouse gases (CH4 or N2O) which can be expressed in terms of CO2 equivalents through indirect activities such as the use of agricultural equipment or chemicals. The method of Tan and Foo [29] assumes that net carbon emissions are so low when compared to fossil energy sources that they can be treated as being approximately zero. A more exact solution can be found by using the true value of carbon intensity. Furthermore, such an approach allows for the differentiation of different low-carbon energy sources from each other. More recently, Crilly and Zhelev [30] proposed to extend the use of energy planning composite curve to handle dynamic nature of energy demand-supply. In general, it is desirable to maximise the use of low- or zero-carbon sources to replace conventional fossil fuels with high carbon content e.g. oil, coal, diesel, etc. However, such technologies are either more expensive (as with renewable energy) or more controversial (as in the case of nuclear energy) than conventional fossil fuels. In addition, in the short term it is often necessary to manage the transition to increased low-carbon energy utilisation to minimise disruptions in energy supply or price. For example, current interest in liquid biofuels has led to the allocation of agricultural land for production of energy crops, which has the undesired result of increasing the price of food crops. Thus, in many planning scenarios there are some interests in identifying the minimum amount of low- or zero-carbon energy sources needed to meet regional CO2 emission limits, which is the subject of this work.

Problem statement The problem definition of carbon constrained energy planning is stated as follows:

• Given a set of energy demands (regions), each requires energy consumption of Dj and at the same time, restricted to a maximum emission limit of ED, j. Dividing the emission limit by the energy consumption yield the emission factor for each demand, CD, j.

• Given a set of energy sources to be allocated to energy demands. Each source has an available energy of Si and is characterised by an emission

E

31

factor, CS, i. Product of the available energy and the emission factor gives the total emission of ES, i.

The problem may be summarised as a source-sink representation, as shown in Figure 1. In this work, a more practical scenario of energy planning is presented, where energy planning is broken down into two sub-sectors, namely industry and transportation. In this segregated planning scenario, a newly proposed automated targeting technique is presented to optimise the allocation of energy sources, while observing the emission limit in each region.

Figure 1 Sink-source representation for energy planning

Automated targeting To demonstrate a realistic situation of energy planning, energy demand in each region is divided into the sub-sectors of industry (i.e., stationary applications, including power generation) and transportation (i.e., land, air and maritime transport). Although the conventional fossil fuels are applicable to both sub-sectors, the use of low- and zero-carbon energy sources is limited by compatibility of the different fuels with the respective applications. For instance, biodiesel is normally not used for power generation due to its high production cost (although it can be used as fuel for diesel-powered generators and boilers), while hydropower will not be used directly for transportation. In this case, an optimum allocation of fossil fuels among the sub-sectors is needed to ensure an overall minimum low- and zero-carbon energy sources to be used. For ease of illustration, this scenario is termed as segregated planning. Due to the use of fossil fuels are used among the various sub-sectors, their optimum allocation in these sectors are beyond the capability of the previous-described graphical targeting approach. This calls for the use of mathematical optimisation-based automated targeting technique. The automated targeting technique was originally developed for targeting minimum mass separating agent in mass exchange network synthesis [5,8]. It is extended here based on the cascade analysis concept [29] to locate the minimum amount of low- and zero-carbon energy sources. First step of the procedure calls for the construction

E

32

of an emission interval table, where all emission factors (C) are arranged in an ascending order, with the lowest emission level being denoted as k = 1 and the highest being k = n. Hence, the number of emission interval is determined as n – 1. At each emission level k, net energy is given by the difference between the available energy source with the energy demand, i.e.: ∑i Si, k – ∑j Dj, k (1)

On the other hand, the total emission load (Δmk) within each emission interval is given by the product of the residue energy from interval k (δk) with the difference between two consequent emission levels, i.e.:

Δmk = δ k(Ck+1 – Ck) (2)

In order to minimise clean (low- and zero-carbon) energy sources (SCS), the segregate targeting problem may be formulated as follows:

min Σi SCSi (LP1) subject to,

δk = δk–1 + ( ∑i Si, k – ∑j Dj, k ), k = 1, 2, … n εk = εk–1 + δ k–1(Ck – Ck–1) , k = 1, 2, … n 0 ≤ Si, j ≤ Si , i ∈ SOURCES ε0 = 0, εk ≥ 0, k = 1, 2, … n

The first sets of equality constrains represents the overall energy balance around each emission level k, where residual energy that leaves an emission level k (δk) is given by the sum between the net energy balance at level k (given as Equation 1) with the residual energy of an earlier emission interval (δk–1). The second sets of equality constrains indicate the residue emission load at each emission level k (εk) that is given by the sum between emission load of an earlier emission level k–1 (εk–1) with the total emission load at level k (given by Equation 2). Third sets of inequality constrains indicate that the allocated energy source i to the sink j (Si, j) is subjected to the maximum availability of the source. The last two sets of constraints ensure an overall emission balance to be realised and the emission loads to be cascaded down the emission levels. The formulation of LP1 is a linear programming model that can be easily solved to yield a global optimum. The location of the pinch point is indicated by the emission level where the residue emission load vanishes. A hypothetical case is used to illustrate the proposed automated technique for segregate targeting.

Example Tables 1 and 2 show the energy source and demand data for a hypothetical case that use to illustrate the concept of segregate planning with automated targeting respectively. Regions I, II and III (labelled as R-I, R-II and R-III respectively) each has its specific energy demand and CO2 emission limits. All regions are further divided into two sub-sectors of industry and transportation, with the energy demand ratio between them given as 8:2, 4:6, and 1:9 respectively. The available energy sources include the conventional fossil fuels (coal, oil, and natural gas), low-carbon (biodiesel) and zero-carbon sources (hydropower and biogas). The fossil fuels may be used in both sub-sectors. However, the use of

E

33

certain energy sources is limited since some fuels will be incompatible with certain applications. Two scenarios are analysed for this example.

Table 1 Energy source data for the hypothetical case.

Energy source Emission factor, CS, I (t CO2/TJ) Available source, Si (TJ)Biogas 0 600,000

Hydropower 0 1,000,000 Biodiesel 16.5 900,000

Natural gas 55 800,000 Oil 75 1,000,000

Coal 105 5,000,000 Total 9,300,000

Table 2. Energy demand data for the hypothetical case. (a)

Energy demand

Total emission limit Σj Ej (106 t CO2)

Total expected consumption Σj Dj (TJ)

Emission factorCD, j (t CO2/TJ)

R-I 60 2,000,000 30 R-II 48 1,200,000 40 R-III 40 800,000 50 Total 148 4,000,000

(b) Industry Transportation Energy

demand Dj (TJ) Ej (106 t CO2) Dj (TJ) Ej (106 t CO2) R-I 1,600,000 48 400,000 12 R-II 480,000 19 720,000 29 R-III 80,000 4 720,000 36 Total 2,160,000 71 1,840,000 77

Scenario 1 In Scenario 1, biodiesel is limited to transportation sector; while hydropower and biogas are only used in the industry. The energy planning task is to locate the minimum amount of clean energy sources (include both low- and zero-carbon sources) needed to meet both energy demand and emission limits for each sub-sectors by maximising the use of fossil fuels. In order to apply the automated targeting technique, two emission interval tables are individually constructed for the sub-sectors of industry and transportation (Table 3). As shown, all energy demands and allocated energy sources for both sub-sectors of industry (Si, P) and transportation (Si, T) are located at their respective emission levels. To minimise the use of biogas (BG), hydropower (HP) and biodiesel (BD), the LP1 problem may be modified as follows:

min SBG + SHP + SBD (LP2) subject to,

δk, P = δk–1, P + ( Σi Si, P, k – Σj Dj, P, k ), k = 1, 2, … n δk, T = δk–1, T + (Σi Si, T, k – Σj Dj, T, k ), k = 1, 2, … n

E

34

εk, P = εk–1, P + δ k–1, P (Ck – Ck–1) , k = 1, 2, … n εk, T = εk–1, T + δ k–1, T (Ck – Ck–1) , k = 1, 2, … n 0 ≤ Si, P + Si, T ≤ Si , i ∈ SOURCES ε0, P = ε0, P = 0, εk, P ≥ 0, εk, T ≥ 0, k = 1, 2, … n SBD, P = SHP, T = SBG, T = 0

The first six sets of constraints in LP2 are essentially similar to the four constraints in LP1, with an extension to cover for both industry and transportation sectors. The last sets of constraints indicate that the use of biodiesel in industry is not allowed and that hydropower and biogas are not used in the transportation sectors. Note that additional constraints may easily be imposed to reflect case-to-case restrictions on energy use. The optimum allocation of various energy sources for this example is obtained by solving LP2, with the results shown in Figure 2. From Figure 2, the pinches for industry and transportation sectors correspond to the emission factors of 105 and 75 t CO2/TJ respectively. Minimum low and zero-carbon sources targets are identified to fulfil the overall energy demand; while meeting the CO2 emission limit of each geographic region.

Table 3 Emission interval table for Example 2 (Scenario 1)

Industry sector Transportation sector k Emission factor, C Demand, Dj Source, Si, j Demand, Dj Source, Si, j 1 0 SBG, P + SHP, P SBG, T + SHP, T 2 16.5 SBD, P SBD, T 3 30 1,600,000 400,000 4 40 480,000 720,000 5 50 80,000 720,000 6 55 SNG, P SNG, T 7 75 SOIL, P SOIL, T 8 105 SCOAL, P SCOAL, T

As shown in Figure 2, 1,272,576 TJ of zero-carbon energy source (biogas and hydropower) are used to fulfil the energy requirement of industry. Besides, 732,650 TJ of oil and 154, 774 TJ of coal are allocated in industry sector. On the other hand, all the natural gas (800,000 TJ) is used in transportation sector; however, no coal is required. Note that, a small portion of biodiesel (900,000 TJ – 772,650 TJ = 127,350 TJ) and a larger portion of coal (5,000,000 TJ – 154,774 TJ = 4,845,226 TJ) are left unutilised. Furthermore, the so-called “golden rule” of pinch analysis, whose implications in energy planning was discussed by Tan and Foo [29] applies here as well. From Figure 2, all of the zero- (biogas and hydropower) and low-carbon (biodiesel) sources are to be consumed only above the pinch point in the cascade diagram. Any allocation beyond the pinch point will require more zero- and/or low-carbon sources to be consumed than the minimum targeted demand determined from the model.

E

35

(a) Industry sector

Emission factor, C Energy cascade (TJ) CO2 cascade (106 t CO2)

0 SBG, P+SHP, P = 1,272,576

1,272,576 16.5 SBD, P = 0 21

1,272,576 30 -1,600,000 38

-327,424 40 -480,000 35

-807,424 50 -80,000 27

-887,424 55 SNG, P = 0 22

-887,424 75 SOIL, P = 732,650 4

-154,774 105 SCOAL, P = 154,774 0

0 (PINCH)

(b) Transportation sector

Emission factor, C Energy cascade CO2 cascade (106 t CO2) 0 SBG, T+SHP, T = 0

16.5 SBD, T = 772,650 772,650

30 -400,000 10 372,650

40 -720,000 14 -347,350

50 -720,000 11 -1,067,350

55 SNG, T = 800,000 5 -267,350

75 SOIL, T = 267,350 0 0 (PINCH)

105 SCOAL, T = 0 0

Figure 2. Cascading results of example (Scenario 1) Scenario 2 In Scenario 2, a situation where 10% of total energy demand expansion experienced in Region II is analysed. In this case, the total energy demand is increased to 1,320,000 TJ, with the energy demand ratio of the industry to transportation sub-sectors being remained as 4:6. However, it is desired to maintain the current emission limit during the expansion. Therefore, the emission factor for Region II is decreased from 40 t CO2/TJ to 36.36 t CO2/TJ (= 48 × 106 t CO2 / 1,320,000 TJ). Furthermore, the limitation of biodiesel energy use in industry (as in Scenario 1) is relaxed. It is assumed that 90% of biodiesel is to be used in the transportation sub-

E

36

sector, while at most 10% can be used for stationary applications e.g. diesel generators or small boilers. This corresponds to the new constraint as follow:

9 SBD, P ≤ SBD, T

(a) Industry sector Emission factor, C Energy cascade (TJ) CO2 cascade (106 t CO2)

0 SBG, P+SHP, P = 1,314,778 1,314,778

16.5 SBD, P = 0 22 1,314,778

30 -1,600,000 39 -285,222

36.36 -528,000 38 -813,222

50 -80,000 27 -893,222

55 SNG, P = 0 22 -893,222

75 SOIL, P = 753,007 4 -140,215

105 SCOAL, P = 140,215 0 0 (PINCH)

(b) Transportation sector

Emission factor, C Energy cascade CO2 cascade (106 t CO2) 0 SBG, T+SHP, T = 0 0

16.5 SBD, T = 865,007 0 865,007

30 -400,000 12 465,007

36.36 -792,000 15 -326,993

50 -720,000 10 -1,046,993

55 SNG, T = 800,000 5 -246,993

75 SOIL, T = 246,993 0 0 (PINCH)

105 SCOAL, T = 0 0

Figure 3 Cascading results of example (Scenario 2) Resolving LP2 yields the results in Figure 3. Note that, the pinch points occur at the same emission factor levels as in the previous case; however, the allocation of energy sources have changed. A total of 1,314,778 TJ of biogas and hydropower is used to meet the overall industry demand. In additions, 753,007 TJ of oil and 140, 215 TJ of coal are allocated to this sector. As compared to Scenario 1, the increase of total energy demand is accompanied by more zero- and low-carbon sources and less coal to be used, in response to the aforementioned 10% increase in energy use

E

37

within Region II. Similarly, all the natural gas (800,000 TJ) is consumed in the transportation sector and the demand for biodiesel increases from 772,650 TJ to 865,007 TJ. As presented previously, pinch analysis principles imply that the biogas, hydropower and biodiesel must only be used above the pinch point in Figure 3.

Conclusions In this work, a mathematical optimisation-based targeting approach is then presented to locate the optimum allocation of energy sources to the sub-sectors in each region while meeting emission limits. The model is also able to account for incompatibilities between certain energy sources and applications through the addition of case-specific constraints, which could not be dealt with in the original, graphical pinch analysis approach. These extensions provide new tools for energy planning by government agencies under carbon emission limits; furthermore, these same techniques can be readily extended for use in applications to meet limits of other energy-related emissions, such as sulphur dioxide.

References [1] UNFCCC. A Summary of the Kyoto Protocol. http://unfccc.int/kyoto_protocol/background/items/2879txt.php. Accessed on 25 March 2008. [2] Linnhoff B, Townsend DW, Boland D, Hewitt GF, Thomas BEA, Guy A R, Marshall RH. A user guide on process integration for the efficient use of energy. Rugby: Institute of Chemical Engineers; 1982. [3] Smith R. Chemical process design. New York: McGraw Hill; 1995. [4] Smith R. Chemical process: design and integration. New York: John Wiley & Sons, Inc; 2005. [5] El-Halwagi MM, Manousiothakis V. Simultaneous synthesis of mass-exchange and regeneration networks. AIChE J 1990; 36: 1209-1219. [6] Wang YP, Smith R. Wastewater minimisation. Chem Eng Sci 1994; 49: 981-1006. [7] Towler GP, Mann R, Serriere AJ-L, Gabaude CMD. Refinery hydrogen management: Cost analysis of chemically integrated facilities. Ind Eng Chem Res 1996; 35: 2378-2388. [8] El-Halwagi MM. Pollution prevention through process integration: Systematic design tools. San Diego: Academic Press; 1997. [9] El-Halwagi MM. Process integration. Amsterdam: Elsevier Inc.; 2006. [10] Alves JJ, Towler GP. Analysis of refinery hydrogen distribution systems. Ind Eng Chem Res 2002; 41: 5759-5769. [11] Hallale N. A new graphical targeting method for water minimisation. Adv. Env. Res. 2002; 6 (3): 377-390. [12] El-Halwagi MM, Gabriel F, Harell D. Rigorous graphical targeting for resource conservation via material recycle/reuse networks. Ind Eng Chem Res 2003; 42: 4319-4328. [13] Manan ZA, Tan YL, Foo DCY. Targeting the minimum water flowrate using water cascade analysis technique. AIChE J 2004; 50(12): 3169-3183. [14] Prakash R, Shenoy UV. Targeting and design of water networks for fixed flowrate and fixed contaminant load operations. Chem Eng Sci 2005; 60(1): 255-268.

E

38

[15] Agrawal V, Shenoy UV. Unified conceptual approach to targeting and design of water and hydrogen networks, AIChE J 2006; 52(3): 1071-1082. [16] Foo DCY, Manan ZA. Setting the minimum utility gas flowrate targets using cascade analysis technique. Ind Eng Chem Res 2006; 45: 5986-5995. [17] Ng DKS, Foo DCY, Tan RR. Targeting for total water network – Part 1: Waste stream identification, Ind Eng Chem Res 2007; 46, 9107-9113. [18] Ng DKS, Foo DCY, Tan RR. Targeting for total water network – Part 2: Waste treatment targeting and interactions with water system elements. Ind Eng Chem Res 2007; 46, 9114-9125. [19] Singhvi A, Shenoy UV. Aggregate planning in supply chains by pinch analysis. Trans. IChemE (Part A) 2002; 80: 597-605. [20] Singhvi A, Madhavan KP, Shenoy UV. Pinch analysis for aggregated production planning in supply chains. Comp Chem Eng 2004; 28: 993-999. [21] Foo DCY, Ooi MBL, Tan RR. A heuristic based algebraic targeting technique for aggregate planning in supply chains. Comp Chem Eng 2007 (in press). [22] Foo DCY, Hallale N, Tan RR. Pinch analysis approach to batch reactor scheduiling in multipurpose plants. Int J Chem Reactor Eng 2007; 5: A94. [23] Zhelev TK. On the integrated management of industrial resources incorporating finances. J Cleaner Pro 2005; 13: 469-474. [24] Kazantzi V, El-Halwagi MM. Targeting material reuse via property integration. Chem Eng Prog 2005; 101(8): 28-37. [25] Foo DCY, Kazantzi V, El-Halwagi MM, Manan ZA. Surplus diagram and cascade analysis technique for targeting property-based material reuse network. Chem Eng Sci 2006; 61: 2626-2642. [26] Dhole VR, Linnhoff B. Total site targets for fuel, co-generation, emissions, and cooling. Comp Chem Eng 1992; 17: S101 – S109. [27] Linnhoff B, Dhole VR. Targeting for CO2 emissions for total sites. Chem Eng Tech 1993; 16: 252-259. [28] Klemeš J, Dhole VR, Raissi K, Perry SJ, Puigjaner L. Targeting and design methodology for reduction of fuel, power and CO2 on total sites. Applied Thermal Engineering 1997; 17(8-10): 993-1003. [29] Tan RR, Foo DCY. Pinch Analysis Approach to carbon-constrained energy sector planning. Energy 2007; 32 (8): 1422-1429. [30] Crilly D, Zhelev T. Current trends in emissions targeting and planning. 10th Conference on Process Integration. Modeling and Optimisation for Energy Saving and Pollution Reduction (PRES 2007) 2007.

ENERGY FOR SUSTAINABLE FUTURE Edited by Petar Varbanov, Jiří Klemeš, Igor Bulatov University of Pannonia, Veszprém, Hungary, 5-6 May 2008 Copyright © Manuscript Authors, ISBN 978-963-9696-38-9E

39

Beyond Production Integrated Environmental Protection (PIEP) in Food Industry

1David Napper, 2Igor Bulatov

1Euroteknik, Ltd., Nr. Lutterworth, UK, email: [email protected] 2CPI, CEAS, The University of Manchester, UK, email: [email protected]

Abstract There has been a lot of good research work on pinch process for water and energy plus many efforts to comply with legislation. In many of the best efforts we see today a holistic approach is being taken to identify the individual process points within food factories and see what changes can be made there which will change the total impact the facility has on the working environment, the impact the factory has on the environment, what economical advantages the approach has and how improvements can change the contribution to the value chain and public acceptance/support of the effort. A software platform has been developed to collect information from the individual points of intervention and communicate it to the technical staff who are usually involved. This information forms the basis for continual monitoring and documenting HACCP procedures. In planning stages it supports the evaluation of the probable intervention for impact on the bottom line of the company and return on investment of the individual changes with reports that can be passed upstream to management. Finally when in operation it offers the opportunity for giving relevant authorities constant output of evidence of compliance to applicable legislation.

Introduction The chemical companies began to practice Production Integrated Environmental Protection (PIEP) as much as 20 years ago. It was simply not enough to just mix all waste/by-products together and hope to ever solve the problems their presence in the environment create. This is a direct quote from the BASF web site [1]:

‘Production-integrated environmental protection

Along with the desired end products, chemical processes usually also generate by-products. It pays to reduce or recycle these. Right from the plant planning stage we therefore pay special attention to ensuring that any by-products are avoided or optimally recycled. Wherever we can, we also improve existing processes, and not only our own but those of our customers, too. For example, we developed an optimized recycling procedure for N-dimethylacetamide (DMAC), which our customers use as a solvent in the production of spandex fibers. The new procedure makes it possible to recover DMAC in a high quality and at the same time minimizes solvent losses.’

40

Over the years many environmental efforts have been added to the picture with pinch process on water and energy and point source separation and treatment has become a watchword in many industries. That does not apply to many food factories. Now many drivers (legislative, financial,) make it imperative to do something urgently but the picture needs to be taken beyond PIEP. Paramount in the new focus is climate change, the IPPC directive as well as what can be done under the new hygiene regulations that recognize the need to conserve water, if it can be done in a responsible way. This paper takes the processing of fish as an example of state of the art and beyond the state of the art.

Description of the system In quite a number of industries, including food factories, most waste streams from production are simply mixed together to be treated “end-of-pipe” (as it comes out of the factory) prior to sending to communal waste water treatment plants (Figure 1). Chemicals such as polymers and chemical coagulants are added to gather as many particles and organic material as possible. Flotation using compressed air is used to take the sludge to the top where it is collected, sometimes dewatered. Depending on what is in the water it can be sent to biogas if that facility is available. Otherwise it often must be deposited as hazardous waste. In specific cases it can be spread on the earth. It is always a major expense. All of this is done to reduce the amount of organics in the process water.

Figure 1. “End-of-pipe” treatment of process streams

Environmental legislation and process cost at the waste water treatment plants make it necessary to pre-treat the process water and those tariffs are being constantly increased. In most European countries the company must pay a surcharge for the extra organic load the waste water treatment plants must process. The companies have ever more stringent requirements to avoid placing organics in landfill/deposits.

41

In fact the IPPC Directive (Article 16.2 of Council Directive 96/61/EC) [2] specifically recognizes and mandates that the quantities being deposited be minimized. When deposited they are responsible for the creation of large amounts of methane gases which are part of the problem of global warming. If the organics did not have the chemicals in them they could at least be burned for energy if they have been dewatered as they often contain good quantities of oils/fats but if they have been collected with chemicals burning of the waste can create toxic gas. At each process point in a food factory different material comes into the process water. If all of the water from all of the process points is mixed together there is little chance of getting maximum utilization of the organics that are removed. If, however, separation takes place at each process point the material can have a much better value. European legislation recognizes that by-products from any animal slaughtered for human consumption is automatically suitable for pet food if the material is handled in the correct way. In a typical small fish slaughter facility for trout (Danforel in Vejle Denmark) the technical/R&D responsible reports that only 40% of the quantity of fish processed emerges as product. A small amount of material such as heads and fins are dumped into bins and sold and mink food (for very little) the organics which are removed with chemicals have been spread on farmland until recently but that is now being forbidden. That organic material as well as the organics in the process water are full of omega 3 and 9 oils which have a good market value for cosmetics and pharmaceuticals and the proteins can go to higher cost products such as food additives or for pet food. An estimated additional 40% of the fish could be utilized, if the noble factions of the waste is processed properly. The Norwegian fish foundation RUBIN [3] estimates that 615,000 tons of fish byproducts are dumped into nature each year. The unexploited value of this current waste is estimated to be between 0.5 and 1 billion Euros per year.(see end of document) It is interesting to note that the hygiene regulations related to fish slaughter requires all waste from fish processing points be mixed with 50ppm of chlorine and held for something like 20 minutes to be certain that any pathogens are killed. The result of this action is that very large quantities of disinfection by-products created by the combination of chlorine and the organic material are being dumped into nature each day. In a similar analysis of a small chicken processing facility [4] shows 2000 kroner per day being used for purchasing chemicals and 1200 kroner per day is paid to a truck to haul 28.8 tons of chemical sludge away. Again this “waste” is full of proteins and fat. The company even owns their own pet food company but the lack of technology forces them to pay so much money to get rid of what could be used as product. The fats could be burned to warm the factory and the proteins could be put into pet food. The problem of the organics in the water also gives added expenses in operation. Heat exchangers and other infrastructure that transport the material more easily builds biofilm which is estimated to cost 0.5% of GNP according to the “Biofilm in technical systems” (BioTekS) research project conducted by Danish Technical Institute [5]. Often the process water has been heated or cooled by or for a process and if the water gets a heavy organic load it cannot be used again. All of the energy used goes out of the plant as waste. The water with heavy organic load is ideal for the growth of bacteria which can compromise the quality of the production of the facility.

42

A system is needed to optimize the several problems which have been outlined but it must be a holistic solution that takes the many problems typical to a food factory into consideration. If all of the problems of process water from the individual process points are mixed together and solution is tried end-of-pipe the options are limited. If, on the other hand, each process point is examined individually to see what the operation parameters are, what is the quality of water entering the process point, what is the quality exiting, what by-products can be removed at the individual point, what remediation might be made individually in order to prevent causing problems that cannot be economically solved collectively, it can be possible to improve not only the economy of the plant, but also the impact on nature and carbon footprint. One of the major requirements of a system to solve such a combined problem is one that is not only scalable to serve the individual processing points but also is easy to change the configuration on as needed. There could be two basic options considered. One to provide disinfection when needed, the other for polishing water for the last organic materials, always avoiding the use of chemicals where possible. The most difficult aspect of these approach is the widely varying qualities of water parameters that can be encountered from a functioning factory. The presence or absence of salt in the water, levels of organic material that can be removed by traditional physical methods, Necessary use of soaps in cleaning which cause wide pH variation. All make the process more complicated. During the cycle of a day or a week the qualities and quantities of water vary constantly according to the processes that are taking place. In the fish industry, if the bleeding is taking place, the characteristics of the water is full of blood and fish oil, at the point where the fish is cut into fillets the organic load is of fish protein and oil, where the viscera is removed there is a combination of organics including excrement, guts, oils, blood, and protein. If this material is mixed and allowed to stand for much time the enzymic action makes much of the noble contents unsuitable for many uses. To take the ingredients individually, where they come into the process water requires a flexible system but enables the realization of the value of the by-products. Whether large or small, food factories need to take this “process integrated environmental protection” approach. For larger companies the economies of the scale they are working on make the approach less financially critical (but no less important for the environment) but for small companies legislative barriers can mean life or death. Smaller companies need units that are modular planned which can be quickly configured and modified to accomplish a number of functions with minimal cost. Quite a good amount of organic load can be taken from process water from the several process points within a factory using traditional separation techniques or combination of these in innovative bundles with emerging technologies. The advantage of this approach is that the equipment is often available in many sizes and can be adapted flexibility as needed.

43

< S

T M

icro

scop

e<

Sca

nnin

g E

lect

ron

Mic

rosc

ope

< O

ptic

al M

icro

scop

e<

Vis

ible

to N

aked

Eye

Ioni

c R

ange

Mol

ecul

ar R

ange

Mac

ro M

olec

ular

Ran

geM

icro

Par

ticle

Ran

geM

acro

Par

ticle

Ran

geM

icro

met

ers

0.00

10.

010.

11.

010

100

1000

Ang

stro

m U

nits

1010

010

0010

410

510

610

7

Sug

ar

Viru

s

F

ine

Test

Dus

tG

rand

ular

Met

alSy

nthe

ticTo

baco

Sm

oke

Mille

d Fl

ower

Activ

ated

Car

bon

Ion

Dye

Late

x / E

mul

sion

Aque

ousS

alts

Alb

umen

Pro

tein

Bac

teria

Beac

h Sa

nd

Pes

ticid

eAs

best

osPo

llen

Her

bici

deG

elat

in

Sep

arat

ion

Tech

nolo

gies

Mec

hani

cal F

iltra

tion

Rev

erse

Osm

osis

RO

Ultr

afilt

ratio

n U

FPa

rticl

e Fi

ltrat

ion

Driv

ing

Forc

e (D

F)

P

ress

ure

diffe

renc

eB

ar 1

0 - 1

00ba

r 1 -

10N

anof

iltra

tion

NF

Mic

rofil

tratio

n M

FP

ress

ure

diffe

renc

eba

r 10

- 70

Bar 0

.5 -

5C

ontin

ual s

elfc

lean

ing

filte

rtype

Con

t. se

lfcle

anin

g fil

terty

pe

Con

t. se

lfcle

anin

g fil

terty

pe S

UPP

OR

TED

Con

t. se

lfcle

anin

g fil

terty

pe S

UP.

Dec

ante

r cen

trifu

geD

ecan

ter c

entri

fuge

DF

for c

entri

fuge

is 5

00 to

300

0 tim

es th

e fo

rce

of g

ravi

tyD

ecan

ter c

entri

fuge

SU

PPO

RTE

DD

ecan

ter c

entri

fuge

SU

P.

Dis

c st

ack

sepa

rato

rD

isc

stac

k se

para

tor

Ele

ctro

dial

yais

Ele

ctro

nica

lly d

riven

mem

bran

es (E

lect

ro o

smos

is)

DF

Ele

ctric

al p

oten

tial d

iffer

ence

Dia

lysis

mol

ecul

es re

tain

ed to

< 1

nm

DF

conc

entra

tion

diffe

renc

e

Com

bine

d E

lect

roch

emic

al p

roce

sses

< P

PT

Rem

oval

is c

urre

nlty

bei

ng d

emon

stra

ted

to P

PT

rang

esC

opyr

ight

© -

Eur

otei

nik

Ltd.

44

The full process water recycling strategy is based on a number of innovations, first the points where organic material comes into process water are identified and their parameters defined. A priority is made for the testing and application for each water quality and identified process water source. Required water qualities to serve that point are also defined (e.g. can water fit for use be substituted for drinking water quality.) A configuration is put together for that process point. The system is modular and fully scalable. The modular construction can also be a good strategy for system integrity. Even with a flow of only 30 m3 it can be an advantage to use three 10 m3 units in order to be able to take one out of service in off peak periods for service or cleaning. When all of the defined points of intervention in a factory have been tested and the quality of the water validated as is required, a water quality management strategy is possible. Qualities of water produced after the PRDS are matched with qualities required, keeping in mind any HACCP principles that could effect decisions. A software package has been developed to provide Drag And Drop (DAD) insertion of the various configurations of points of intervention at a given factory. This facility is self configuring and matches qualities of water provided with qualities indicated as fit for use. The system can be adjusted for all parameters including energy used (heat, fuel and electricity), organics recovered or removed and their fate plus economical impact, variations in the CO2 impact traceable to the individual point of technological intervention. If there are, for example 12 process points in a factory with various technical innovations considered (variable speed pumps vs use of choke valves, re-circulation of water after it has been processes, the use of fats removed from process water to be burned as fuel, cost of trucking byproducts to point of sale and economical yield that is a result) the DAD system can maintain an overview of what the total system changes can be, help suggest priorities for intervention and track achievements as they are implemented. There are a number of synergies in the system which is being submitted to patenting at the moment and will be subsequently made public except on a need-to-know basis. Often there can be sector specific legislation to be considered. In 1997 the Norwegians passed a law which was intended to control the problems related to the organic materials being dumped into the environment. In addition to the BOD/TOC problems there was a major problem with pathogens. This legislation has been modified a few times and now is the model for controls in Denmark and the rest of Europe. (FOR 1997-02-20 nr. 192: Forskrift om disinfeksjon av inntaksvann til or avløpsvann fra akvakulturrelatert virksomhet) both in Norwegian and English FAO site [6]. Briefly, the legislation calls for the collected effluents to be filtered with a 300µ filter. The material gathered can be classified as category 2 by-products which can be treated with heat and pressure under a prescribed formula for disinfection to be converted into, for example animal food. The process water which passes through the filter with its organic load is mixed with chlorine until the level of 50 ppm is reached. There must be a retention time of 20 minutes with constant mixing. Polymers are and flocculating chemicals (which are both classified as hazardous for transport and handling) are added to take as much of the organic load as possible. The mixture of material that has been gathered has bubbles pumped into it that float the coagulated portions to the top and this chemical sludge, classified as hazardous waste must be disposed of: In Denmark that cost a minimum of 100 dkk per ton. When the water that has been disinfected and chemically treated is released into the fjord it must have a residual level of 8 ppm to assure the overall log 3 reduction of

45

pathogen bacteria and virus prescribed by the mentioned legislation. In tests certified by the authorized lab of Bergen the PRDS system showed that it could give up to log 8+ reduction of pathogens, far exceeding the legal requirements, in a continual flow, multiple process configuration which required no residence time. When the traditional fish waste water system has a flow of between 20 and 70 m3/hour in a typical system, the infrastructure required for holding, stirring, dosing the various chemicals for 20 minutes, using compressed air for flotation, transport storage plus handling of hazardous chemicals, transport and disposal of chemical sludge you can see the advantage of our system which takes the by-products without residence time, without using chemicals in a condition that is at least category 2 and possibly category 3 rated by-products but may have more noble value for extraction of enzymes or other valuable material when it is extracted in these ways. Alone if the oil is taken and burned to furnish energy to the factory and meal is used for pet food, there is a great advantage. The chemical sludge created by the traditional method cannot be easily burned because of the creation of toxic substances created by the burning of the chemicals. This is specifically the chlorine residues from FeCl used for flocculation as well as the 50 ppm of chlorine that is used for disinfection. When burned, this chlorine is converted into HCl. The system described also extends the life of the infrastructure because of the reduction of use of chemicals. All of the chemicals in question are highly corrosive according to their labels. Pumps, pipes and containers that function in the chemical heavy environment are subject to a lot of corrosion. The elimination of the chemicals greatly decreases the cause of the corrosion. There are other case studies available that show the saving of infrastructure, water, energy, transportation with the conversion of waste bound for biogas with transport cost carried by the waste generator compared to the immediate conversion of the sludge to category 2 and 3 by-products for a company that already has an animal food production where they can sell the by-products. They also have equipment for rendering organic material that makes it possible to take the oils and use them for energy, reducing dependence on fossil fuels. The case study is at a chicken slaughter factory where the additional organic load makes the waste water treatment plant of the commune go beyond their capacity during peak periods. The traditional wastewater treatment plant cannot handle the additional load of Coli bacteria from the chicken slaughter so the water released into the stream leading to the beach of our neighbour country is polluted. The new system will give effluents to the wastewater treatment system that are without the pathogen bacteria and virus load. The treatment of the water at the respective process points will reduce the chance of cross contamination of batches of chickens some of which could be infected by salmonella others which are free from it. Additionally pathogen virus such as those causing bird flu can be eliminated in the process water. The balance of the units in the flow chart must be achieved to get optimal performance. The removal of organic material in the primary separation has a knock on effect on the amount of energy necessary for the integrated electrochemical removal disinfection phase from the PRDS unit which allows the electrochemical process to be accomplished efficiently (Figure 2). The final polish makes a pre-separation of organic material that now is without chemicals so that better use of it can be achieved. The known use of the heat exchanger in the flow is enhanced as the formation of biofilm with attendant loss of efficiency is avoided or reduced because

46

the bacteria needed to attach and grow the biofilm have been eliminated as is the organic material which can add to it. This feature of the system is also subject to patenting.

Figure 2. Particle Removal Desinfection System (PRDS)

In both of the case studies made so far the CO2 footprint of the facility is significantly improved so that carbon trading will be an additional advantage created by the technology as well as the contribution of better environment.

Conclusions and Discussion There is a very interesting dichotomy in the food value chain. Prices are going up because primary production is being used to make fuel. Too much water and energy is being used. By-products are being treated as waste and end in landfills creating methane gas bad for our environment. The project we are involved in attempts enables to find concrete solutions to some of these issues.

References 1. BASF website www.corporate.basf.com/en/sustainability/oekologie/umweltschutz. htm?id=V00-eLEJjC373bcp3hH, (visited 03/04/2008) 2. Council Directive 96/61/EC of 24 September 1996 concerning integrated pollution prevention and control http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri =CELEX:31996L0061:EN:HTML (visited 03/04/2008) 3. RUBIN www.rubin.no (visited 03/04/2008) 4. Personal communication with technical chief of Rose Poultry Bov, Denmark (17/10/2005) 5. BioTekS, Danish Technical Institute, Reference www.bioteks.dk (visited 17/11/2004) 6. www.fao.org/fi/website/FIRetrieveAction.do?dom=legalframework&xml=nalo_norway.xml (visited 03/04/2008)

Particle Removal Desinfection System (PRDS)

Fine particle recovery

desinfection system

Centrate Centrate

Org. DS (Fines)


47

Design and Operation Optimization for a Stand-Alone Power System Using Renewable Energy Sources and Hydrogen Storage

Dimitris Ipsakisa,b, Spyros Voutetakisa, Panos Seferlisa,c, Fotis Stergiopoulosa, Simira Papadopouloud, Costas Elmasidese

aChemical Process Engineering Research Institute (C.P.E.R.I.), CEntre for Research and Technology Hellas (CE.R.T.H.), P.O. Box 60361, 57001, Thermi-Thessaloniki,

Greece bDepartment of Chemical Engineering, Aristotle University of Thessaloniki, P.O. Box

1517, 54124 Thessaloniki, Greece cDepartment of Mechanical Engineering, Aristotle University of Thessaloniki, P.O. Box

484, 54124 Thessaloniki, Greece d Department of Automation, Alexander Technological Educational Institute of

Thessaloniki (A.T.E.I.TH.), P.O. Box 141, 57400 Thessaloniki, Greece e Systems Sunlight SA, Neo Olvio, Xanthi, 67200, Greece

Abstract The stand-alone power system (SAPS) under consideration is based on the exploitation of solar and wind energy via a photovoltaic array (pv-array) and three wind generators respectively. The produced power is used to meet the demand of a constant load and any excess power can be used for the production of hydrogen through a PEM electrolysis unit. The produced hydrogen is compressed and stored in cylinders under medium (<50bar) pressure for future use in a PEM fuel cell in case of shortage of power. A lead-acid accumulator is also used to account for the short-term needs of the system due to the wide fluctuations of the weather conditions. The developed power management strategies for the above system are based on the State-of-Charge (SOC) of the accumulator. The various subsystems (electrolyzer and fuel cell) can operate in such a way that the load electrical needs are always met and depletion of hydrogen is avoided. Specifically, the lower limit on the SOC zone indicates the regime of fuel cell operation while the upper limit designates the electrolyzer operation regime. Simulation studies based on real meteorological data revealed that hard bounds on the SOC zone might lead to excessive usage of the various subsystems (electrolyzer and fuel cell) with possible deterioration on their operation. Therefore, a modified power management strategy that employs flexible SOC bounds resulted in significant improvement of the overall system performance. Keywords: Stand-alone power system, power management strategy, renewable energy sources, hydrogen, fuel cell, electrolyzer. Introduction The use of renewable energy systems (RES) for the production of electrical energy can contribute significantly to the reduction of greenhouse emissions as compared to conventional methods of energy production from fossil fuels. However, the large variations of the weather conditions even during the day, lead to the need of a more

48

complicated system that consists of many components which can contribute to the electrical production. Therefore, the pv-array and the wind generators should be complemented by an energy storage system to absorb the excessive energy and to meet the energy deficit of the system during periods of low energy production. Traditionally, deep-cycle lead-acid accumulators have been used as the means of energy storage but recently fuel cells in combination with an electrolyser for hydrogen production and hydrogen storage units have been considered for energy storage [1,2]. The advantages of using subsystems that exploit solar and wind energy to generate electricity include the avoidance of pollutants emissions, long lifetime and low maintenance requirement [1]. Moreover, solar and wind energy is abundant, free, clean and inexhaustible. The production of hydrogen through water electrolysis is considered as an eco-friendly procedure that does not lead to harmful gases as in other processes that involve hydrogen production from hydrocarbons [3, 4]. The design, analysis and optimization of such systems require the development of accurate models for all subsystem components. Previous studies on modeling stand-alone power systems considered each subsystem as individual. Accurate models that predict the daily profiles of energy produced from photovoltaic systems and wind generators were based on meteorological data [5]. Dynamic PEM electrolyzer, hydrogen storage and fuel cell systems, determined the hydrogen production and consumption according to their current-voltage characteristic [6,7]. Optimization strategies based on cost minimization of the integrated system utilizing a short-term and a long-term storage system can be proved quite efficient, while numerous power management algorithms that use models to predict the performance of stand-alone power systems have been developed and evaluated [8-10]. The main conclusion is that power management strategies strongly affect the lifetime of the various subsystems and in particular the lifetime of the accumulator, the electrolyzer and the fuel cell. Such information can guide the designer to suitable decisions on enhancing the performance of the system for an economical and reliable operation. In this study, a power management strategy is introduced and modified with the use of a variable width hysteresis band on the SOC of the accumulator in order to determine the most efficient way of controlling the energy flow in the system. In the first proposed strategy, the SOC of the accumulator is held within two limits: the minimum SOCmin, and the maximum SOCmax. Nevertheless, by using only two boundary limits for the SOC, the electrolyzer and the fuel cell cannot operate for a long time and the continuous start-ups and shut-downs on their operation may prove to be detrimental to their overall lifetime. The advantages of using the hysteresis band in the second strategy include the protection of the accumulator, because its operation cycles can be reduced, whereas the electrolyzer and the fuel cell can continue to operate with respect to the energy produced from RES and protect them from continuous start-ups and shut downs. The key characteristic of the hysteresis band, such as the limits that determine the boundaries of operation for the various subsystems, will be studied with the ultimate goal of building a reliable and effective control strategy.

49

Stand-Alone Power System Description The present study is based on an existing SAPS located at the facilities of Systems Sunlight S.A in Neo Olvio, Xanthi in Greece, developed within the framework of a research project with the participation of the Chemical Process Engineering Research Institute and Systems Sunlight S.A.. This application utilizes solar and wind energy with hydrogen production through water electrolysis, storage and utilisation in a fuel cell. The renewable energy system (RES), consists of a pv-array and three wind generators. Surplus energy is to be supplied to a PEM electrolyzer, after the demand of a 1kW constant load have been met. The produced hydrogen is to be stored in cylinders under pressure. In case there is a lack of energy, a PEM fuel cell is used to provide power. Also, in order to account for short-term needs, a lead-acid accumulator is also used and charged by the renewable energy sources or the fuel cell, depending on the availability of the renewable energy sources. In case of energy excess that can’t be used in any other subsystem, then the auxilliary units (e.g. compression of hydrogen) of the integrated system may utilize this energy. Furthermore, power electronic converters are employed for power management and for the integration of the various subsystems. Figure 1 represents a layout of the stand-alone power system and Table 1 shows the basic specifications of the various subsystems.

Figure 1: Block diagram of the SAPS

Table 1. Subsystems specifications in the SAPS

Photovoltaic Array

Wind Generators

Lead-Acid Accumulator

PEM Electrolyzer

PEM Fuel Cell

Storage Unit

5kWp 3kWp 144kWh 4.2kWp 4kWp 4m3

Modelling of the subsystems In this section the mathematical models for the operation of the various subsystems of the SAPS will be presented and discussed. The relationship that describes the I-V characteristic in a pv-array was based on an equivalent electrical circuit that depicts the electrical phenomena during the solar absorption from the module. [2]:

50

]1)[exp( −⋅+

⋅−=−−=α

spvpvOLshDLpv

RIVIIIIII (1)

where IL, ID, ISH denote the light current, diode current and shunt current respectively in A, IO the diode reverse saturation current in A, Rs the series resistance in Ω, α the curve fitting parameter in Volt, and Vpv, Ipv the operation voltage and current in Volt and A, respectively. Thus the produced power, Ppv, from the pv-array is given by:

convpvI η⋅⋅= pvpv VP (2) where ηconv denotes the efficiency of the DC/DC converter (~90-95%). Respectively the equation that describes the output power of the wind generator for a specific value of the wind speed is given from [11]:

conv3 η

2),( ⋅

Α⋅⋅= windpm vcP ρβλ (3)

where Pm denotes the mechanical output of the turbine in Watt, cp the performance coefficient of the turbine, ρ the air density in kg/m3, Α the turbine swept area in m2, vwind the wind speed in m/s, λ the tip speed ratio, β the blade pitch angle in (deg) and ηconv the efficiency of the AC/DC converter (~90-95%). A very important parameter that needs to be studied in detail during the power management strategies is the SOC of the accumulator, as it influences the operation of the accumulator, of the electrolyzer and of the fuel cell. The SOC is given by [2]: SOC(t)=SOC(t-1) +Ιbat·ηbat·(∆t) (4) where ηbat is the efficiency factor, ~95%, Ιbat is the charging/discharging current in Α, and ∆t=(t)-(t-1) is the sample time in h. The production or consumption rate of hydrogen is given by the Faraday’s Law [2]:

FnIn

nne

elecfccFH ⋅

⋅⋅= /

2 (5)

where nH2 denotes the hydrogen flow rate in mol s-1, nc the number of cells, Ifc/elec is the operation current of the fuel cell/electrolyzer in A, ne is the number of electrons and F is the Faraday’s constant in Cb/s. The Faraday’s efficiency, nF, is defined as the ratio between the actual and the theoretical amount of hydrogen produced and is usually around 80-100%. In the case of a fuel cell the Faraday’s efficiency is defined as 1/nF. It is noted that the minimum and maximum power levels that the electrolyzer is allowed to operate in 1050Watt (Pmin,elec) and 4200Watt (Pmax,elec), respectively. These levels have been provided by the manufacturer and correspond to the 25% and 100% of the nominal power of the electrolyzer respectively. Power Management Strategies Based on the input data (weather profiles) during the simulated four-month period, the net power from the renewable energy system after meeting the load requirements

51

is calculated as the difference between the output power from the RES, Pres, (i.e. the sum of the output power from the photovoltaic array and the wind generators) and the constant power demand, Pload. loadRES PPP −= (6) Figure 2 shows the net power in the system during the simulated time period where it can be concluded that in 56.4% of the total time period an excess of power is available (positive values) and in 43.6% of the time a shortage of power is obtained (negative values). Consequently, it is essential for such a complicated system to employ a power management strategy that will handle efficiently the surplus or shortage of energy.

0 500 1000 1500 2000 2500 3000-1000

0

1000

2000

3000

4000

5000

Exce

ss o

r Sho

rtag

e of

Pow

er, W

att

Time, h

Figure 2: Net power on the SAPS during the four-month simulated time period

The first power management strategy (PMS 1) is presented in Figure 3. As it can be seen there are two limits on the SOC whithin which the electrolyzer and the fuel cell are allowed to operate with respect to the power produced from the RES.

• If SOCmin<SOC<SOCmax, then the accumulator is charged or discharged by the RES in order to meet the system’s energy needs in respect to the available energy sources.

• If SOC≤SOCmin, then the fuel cell meets the load demand and charges the accumulator if shortage of power exists (P<0). In case the output power of the fuel cell is higher than the power deficit, then the excess power is utilized by the accumulator.

• If SOC≥SOCmax then the excess power is used by the electrolyzer as long as Pmin,elec≤P≤Pmax,elec. If P<Pmin,elec then the accumulator is charged by the RES beyond the maximum limit and if P>Pmax,elec then the electrolyzer utilizes power equal to Pmax,elec and the rest (P- Pmax,elec) is used to charge the accumulator [9].

52

SOCmin SOCmax

OPERATION OF THE PEM FUEL

CELL

OPERATION OF THE PEM

ELECTROLYZER

LEAD-ACID ACCUMULATOR

USAGE FOR ENERGY

FLUCTUATIONS

Figure 3: Representation of the operation of the SAPS for the PMS1 [9]

By using only two limits the electrolyzer and the fuel cell cannot operate for a long time and the frequent start-ups and shut-downs in their operation might lead to their performance deterioration. In the second strategy (PMS2), a hysteresis band will be introduced for the operation of the electrolyzer and the fuel cell. Figure 4 presents the basic philosophy of the strategy.

• If SOC ≤ SOCmin, then the fuel cell provides the necessary power to meet the power deficit. In case the output power of the fuel cell is higher than the power deficit, then the excess power is utilized by the accumulator.

• If SOCmin<SOC<SOCfc and in the previous time step the fuel cell was operating and still shortage of power exists (P≤0), then the fuel cell doesn’t shut down and continues to operate until the SOC reaches the limit SOCfc or until the renewable energy sources can meet the systems demands (P>0).

• If SOCfc ≤ SOC ≤ SOCmax the lead-acid accumulator is charged (P>0) or discharged (P≤0) with respect to the renewable energy sources.

• If SOCmax < SOC < SOCmax_charge and P<Pmin,elec then the power P, is used to charge the battery until the limit SOCmax_charge and after that limit the accumulator cannot be further charged and the extra energy is utilized by the auxilliary units of the system (Eextra). Similarly, if P>Pmax,elec then the electrolyzer utilizes power equal to Pmax,elec and the extra power P-Pmax,elec is used to charge the battery, but not further than SOCmax_charge. It is also highlighted that if SOC>SOCmax and Pmin,elec ≤ P ≤ Pmax,elec the RES fully support the operation of the electrolyzer. Obviously, with suitable auxilliary load practices the accumulator can be protected from overcharging.

53

Figure 4: Operation of the SAPS with the use of the hysteresis band for the PMS2 [9] Simulation Results and Discussion The following values for the operating parameters of the SAPS are selected (unless stated otherwise): SOCmin: 84%; SOCmax: 91%; load: 1kW; fuel cell output power: 1kW; initial accumulator capacity: 2700Αh (SOCo=90%); initially hydrogen inventory level: 60.5 Nm3 (85% of the maximum capacity of the pressurized tanks). The various limits of the hysteresis band that were used during the simulation are: Case a (PMS2a) SOCmax_charge=SOCmax + 2%⋅ SOCnom SOCfc=SOCmin + 2%⋅ SOCnom

Case b (PMS2b) SOCmax_charge=SOCmax + 1%⋅ SOCnom SOCfc=SOCmin + 1%⋅ SOCnom

The operation cycle of an accumulator is defined as the process where a discharging (or charging) mode is followed by a charging (or discharging) mode. Table 2 presents the percentage of time that each subsystem operated during the simulated time period. As it can be seen, the use of the hysteresis band in PMS2a and PMS2b resulted, in most of the cases, in the reduction of the total operation time (charging and discharging time) for the accumulator. However, the most important outcome from the use of the hysteresis band is that the operation cycles of the accumulator are significantly reduced for all the case studies and by providing a more flexible zone for the SOC limits, the accumulator is protected from excessive usage. In general, it is concluded that case-b, which exhibited a more narrow zone for the SOC of the accumulator, led to a better operating performance for the accumulator than case-a, as far as the total operation time and operation cycles of the accumulator are of concern. One would expect that the electrolyzer and the fuel cell would operate more time during the four-month time period because of the use of the hysteresis band. As it can be seen from Table 2 though, the use of the hysteresis band does not affect much the total operation time of the electrolyzer or the fuel cell.

54

Therefore, the electrolyzer and the fuel cell were not subject to heavy utilization during the simulated time period and this behaviour favors the overall performance of the autonomous system. From the analysis of Table 3 it is concluded that the use of the hysteresis band in PMS2a resulted in very little difference for the hydrogen inventory (average values) as compared to PMS1, while PMS2b exhibited less hydrogen for all case studies. Since the main purpose of the stand-alone power system is the reliable energy supply to the load and not the hydrogen production, it shouldn’t be of prime concern the fact that the hydrogen inventory is less with the use of the hysteresis band as long as there isn’t any time period where lack of hydrogen to exist. A remarkable also result, is that in PMS2b there is extra energy (to the auxilliary units) which was not reported at such high levels in PMS2a and PMS1.

Table 2: Operation time and operation variables for the subsystems for two case studies at the end of the four-month time period [9].

%Tcharge %Tdischarge %Telec %TFC Cycles PMS1 40.92 52.07 7.08 5.96 99

PMS2a 40.79 52.20 7.08 5.83 96 SOCmin=84%

& PFC=1kW PMS2b 39.60 52.07 6.88 5.96 93

PMS1 39.36 54.88 5.76 3.15 112 PMS2a 39.36 54.88 5.76 3.15 106

SOCmin=80% &

PFC=1kW PMS2b 38.35 54.78 5.59 3.25 107 Table 3: Hydrogen production and consumption and energy data for the operation of

the subsystems for two case studies at the end of the four-month time period [9]

H2 prod., m3 H2cons., m3 Avg. Net H2, m3

Eextra, kWh

PMS1 92.06 102.80 45.84 0 PMS2a 91.11 99.76 45.74 0

SOCmin=84% &

PFC=1kW PMS2b 87.95 102.8 42.73 20.2 PMS1 75.22 53.94 61.26 0

PMS2a 75 53.94 61.19 0.85 SOCmin=80%

& PFC=1kW PMS2b 72.18 55.68 59.34 18.6

The effect of the hysteresis band on the operation of the fuel cell is more obvious in the case where the output power of the fuel cell is set at 2kW. Figure 5 shows that the fuel cell operated continuously at PMS2a and PMS2b, while there was an interruptible operation at PMS1 with frequent start-ups and shut-downs at a typical time interval of the simulated time-period. The same can be concluded for the electrolyzer [9].

55

2832 2834 2836 2838 2840 2842

4

PMS2a

PMS2b

Ope

ratio

n Po

ints

for

the

PEM

Fue

l Cel

l

Time, hr

PMS1

Figure 5. Operation of the fuel cell during a typical time period (case study for

SOCmin=84% and PFC=2kW), [9]

Conclusions The present study introduced the concept of a power management strategy using a hysteresis band in the SOC level and especially the modifications that can be made in a simple algorithm in order to achieve a reliable power supply to a constant load for a SAPS that exploits renewable energy systems. A major conclusion is that it is necessary to develop an efficient power management strategy in order not only to meet the power demands but also to protect the various subsystems from heavy utilization. A hysteresis band in the SOC level for the operation of the fuel cell and the electrolyzer protects the accumulator from over-utilization and properly regulates the operating pattern of the fuel cell and the electrolyzer. Accumulator operation cycles are significantly reduced and the charging and discharging time decreased in most of the simulated cases. Furthermore, the hysteresis band reduced the number of start-ups and shut-downs for the electrolyzer and fuel cell and therefore eliminated irregular operation of these two subsystems. Operation regularization (decrease of start-ups and shut-downs) protects the fuel cell and electrolyzer from mechanical or electrical failures that may eventually result in their early replacement. Suitable selection of the critical levels for the decision variables in the PMS can provide an optimal operating performance. The PMS will form the basis for the development of a model based control system for the online monitoring and control of the dynamic features of the stand-alone power system. Acknowledgements The financial support of the European Fund of Regional Growth and the Region of Eastern Macedonia and Thrace with final beneficiary the General Secretariat of research and technology under project contract (PEP/AΜTh 9) in the operating project Eastern Macedonia and Thrace is gratefully acknowledged.

56

References [1] Nelson DB, Nehrir MH, Wang C. Unit sizing and cost analysis of stand-alone hybrid wind/PV/fuel cell power generation systems. Renewable Energy 2006;31:1641-1656. [2] Ulleberg Ø. Stand-Alone Power Systems for the Future: Optimal Design, Operation & Control of Solar-Hydrogen Energy Systems. PhD thesis, Norwegian University of Science and Technology, Trondheim, 1998 [3] Ipsakis D, Ouzounidou M, Seferlis P, Voutetakis S, Papadopoulou S. Study of an Integrated System for the Production of Hydrogen by Autothermal Reforming of Methanol, Computer Aided Process Engineering Forum-CAPE-FORUM, 7-8 February 2008, Thessaloniki, Greece [4] Ipsakis D, Kechagiopoulos P, Martavaltzi C, Voutetakis S, Seferlis P, Daoutidis P, Stergiopoulos F. Study of an integrated system for the production of hydrogen by autothermal reforming of methanol, Computer Aided Chemical Engineering, Volume 24, 2007, Pages 913-918 [5] Kolhe M, Agbossou K, Hamelin J Bose TK., Analytical model for predicting the performance of photovoltaic array coupled with a wind turbine in a stand-alone renewable energy system based on hydrogen. Renewable Energy 2003; 28:727–742. [6] Görgün H., Dynamic modelling of a proton exchange membrane (PEM) electrolyzer. International Journal of Hydrogen Energy 2006; 31:29-38. [7] Busquet A, Hubert CE, Labbé J. , Mayer D, Metkeemeijer R, A new approach to empirical electrical modelling of a fuel cell, an electrolyser or a regenerative fuel cell. Journal of Power Sources 2004; 134:41-48. [8] Santrelli M, Cali M, Macagno S. Design and analysis of stand-alone hydrogen energy systems with different renewable sources. International Journal of Hydrogen Energy 2004; 29:1571–1586. [9] Ipsakis D, Voutetakis S, Seferlis P, Stergiopoulos F, Papadopoulou S, Elmasides C. The Effect of the Hysteresis Band on Power Management Strategies in a Stand-Alone Power System, (submitted for publication to Energy). [10] Ghosh PC. Cost optimization of a self-sufficient hydrogen based energy supply system. PhD thesis, Forschungszentrum Julich in der Helmhotltz-Gemeinschaft, Institut fur Werkstoffe und Verfahren der Energietechnik Institut 3:Energieverfahrenstechnik, (Diss. Aachen, RTWH, 2003). [11] Siegfried H., Grid Integration of Wind Energy Conversion Systems. John Wiley & Sons Ltd, 1998.


57

Flow Distribution Effects on the Thermal Efficiency of a Brazed Plate Heat Exchanger

David J. Kukulka, Anthony DeStefano1 State University of New York College at Buffalo, 1300 Elmwood Avenue

Buffalo, New York 14222 USA, [email protected] 1-716-878-4418; FAX:1-716-878-3033

1Cameron Compression Systems; Buffalo,New York 14225 USA

Abstract Compact heat exchangers have recently experienced an increase in popularity due to their effectiveness, relatively low cost and compactness. In many industries, brazed plate heat exchangers or plate/frame heat exchangers are being used to replace shell and tube heat exchangers. In order to utilize brazed plate heat exchangers in critical applications, performance sizing must be done accurately. Traditional methods for predicting heat exchanger performance assumes uniform heat transfer coefficients and uniform fluid flow across all flow channels. However many brazed plate heat exchangers yield unpredictable pressure drops and poorer thermal performance than predicted. This is the result of non-uniform surface flow known as flow maldistribution. Maldistribution can be characterized into, gross maldistribution, passage-to-passage maldistribution, and manifold induced maldistribution. A number of factors (i.e. plate design/geometry, fouling, fluid flow rates, surface finish and fluid properties) can contribute to flow maldistribution. An experimental investigation of pressure drop and flow visualization in a brazed plate heat exchanger is presented. These results show maldistribution present within the tested brazed plate heat exchanger and also indicate higher velocities in the upper channels of the heat exchanger. Flow visualization results reveal non-uniform flow and dead spots in the flow. These problems influence the thermal performance of the unit and lead to undersized heat exchangers. Insight into improving the performance of these brazed units is discussed.

Introduction Brazed plate heat exchangers are in theory capable of transferring the same amount of heat load as a shell and tube unit at a fraction of the cost. This is made possible through the use of chevron patterns that are stamped into the plates. Bassiouny and Martin [1] performed an analytical study which takes into account maldistribution. They present a general flow maldistribution characteristic parameter (m²) that is said to be valid for all plate heat exchangers including brazed units. Rao, Sunden and Das [2] present a theoretical and experimental investigation showing that maldistribution will increase pressure drops across a plate heat exchanger. Rao and Das [3] investigated the effect of port to channel flow maldistribution on the pressure drop across a plate heat exchanger. Experimental results show that pressure drop varies with port size and flow rate. Rao and Das [3] show that channel resistance plays a major factor in flow maldistribution. The aim of the present study is to extend

58

the analysis of flow maldistribution performed for a plate and frame style heat exchanger (PHE) to a brazed plate heat exchanger (BPHE). A number of parameters (i.e. number of plates, port diameter, flow rate, surface finish and plate type) can effect the flow distribution and pressure drop in a brazed plate heat exchanger. In plate heat exchangers, pressure profiles in the inlet and outlet ports change due to flow dispersing either into or out of flow channels. Bassiouny and Martin [1] describe three pressure curves encountered in plate heat exchangers (see Figure 1). The pressure profile in the inlet port is represented by (P) and the pressure profile in the outlet port is represented by (P*). Figure 1 (a) shows parallel pressure curves which indicate a uniform flow distribution in flow channels; Figure 1 (b) shows converging pressure curves which indicates maldistribution and flow decreasing in the direction of the intake stream; Figure 1 (c) shows diverging pressure curves; this also indicates maldistribution.

Figure 1- Pressure Distribution in a PHE (Bassiouny and Martin, 1984) Rao, Kumar and Das [4] present an experimental study showing how pressure drop across a plate heat exchanger is effected by maldistribution. They show that several factors influence pressure drop including channel resistance, flow rate, number of channels, and port size. Rao, Sunden and Das [2] present a study on how the number of plates in a plate heat exchanger effects flow distribution and pressure drop. Results include experimental testing on a 21 and 81-plate heat exchanger showing that pressure distribution (from the first to the last channel) decreases more in a plate heat exchanger consisting of a larger number of plates. They point out that adding additional plates can actually increase maldistribution within the heat exchanger. Fluid velocity decreases in the inlet port at such a rapid rate that plates at the bottom of the PHE may have little or no flow through them. Traditionally plate heat exchangers have been modeled with uniform flow in all channels. However, with maldistribution occurring, the assumption of a uniform heat-transfer coefficient is in question. A theoretical evaluation of how variable flow rates effect the heat-transfer coefficient has been conducted by Rao, Kumar and Das [4]. The study shows the importance of considering the use of a local heat-transfer coefficient in performance sizing. In a similar study, Rao, Sunden and Das [2] experimentally studied the effects of port-to-channel flow maldistribution on thermal performance. An increase in maldistribution can reduce the effectiveness of the heat exchanger by up to 75 percent. An investigation of the complex flow and heat transfer characteristics associated with chevron embossed plates has been completed by Kanaris, Mouza and Paras [5]. An experimental analysis and flow visualization of a 21 plate, oil/water PHE was conducted by Lozano, Barreras, Fueyob and Santadomigo [6]. This study added

59

valuable information about the flow variations over a chevron plate. Results for the U-type flow arrangement tested indicate that flow velocities were greater on the port side of the heat exchanger (see Figure 2). Finally, another visual investigation of fluid flow in narrow channels formed by corrugated walls was performed by Focke and Knibble [7].

Figure 2- Streamlines for oil flow (rate 0.91 l/min) Over a Chevron Plate. (Lozano et

al., 2007)

Experimental Setup The BPHE of the present study consisted of 60 plates with a chevron angle of 60 degrees, in U-type flow. Pressure readings were taken at various locations within the inlet and outlet ports. Probes consisted of 1/8 inch copper tubing with pressure transducers attached (see Figure 3). Pressure differences were taken for the range of flows from 20 GPM to 60 GPM. In order to visualize the flow is a BPHE, a set of clear plastic chevron embossed plates were produced and combined to create a transparent BPHE. Top and bottom plates, made from .25-inch Plexiglas, were used to compress and hold the clear chevron plates together. Flow rates tested ranged from 5 GPM to 15 GPM. Visualization was accomplished by injecting dye into the inlet and capturing the results on video. The results show the combined flow patterns through the series of plates.

60

Figure 3- Experimental Set Up to Measure Pressure Drop in a BPHE.

Results Channel pressure drops in the tested BPHE have been recorded in nine locations for a range of flow rates (20 GPM to 60 GPM). Results show variable pressure differences suggesting the occurrence of maldistribution. Pressure differences across upper flow channels were greater than the readings at the bottom of the BPHE. Results are given in Figures 4 thru 8.

Figure 4- Pressure Difference for 20 GPM Flow Rate in an ITT 415-60 BPHE

TOP

BOTTOM

61




TOP

BOTTOM

TOP

BOTTOM

TOP

BOTTOM

62


Pressure differences near the top of the heat exchanger are approximately 45 percent higher than at the bottom. Table 1 shows the comparison of results to a similar study conduced by Rao, Sunden and Das [2]. These results suggest flow maldistribution within the tested BPHE. Inlet and outlet pressure curves are presented in Figures 9-13. All these figures show diverging pressure curves (as Bassiouny and Martin [1] describe) indicating maldistribution in the tested BPHE.

Table 1- Pressure Differential for (A) Rao et al. (2) (B) Present Study 50 GPM

Number of Plates

Pressure Differential Across Upper Flow channels (psi)

Pressure Differential Across Lower Flow Channels (psi)

21A 14 10.8 60B 15 10 81A 9.96 1.13

20 GPM Pressure Curves

6

6.5

7

7.5

8

8.5

9

9.5

10

10.5

11

0 1 2 3 4 5 6 7

Location From Bottom of Heat Exchanger (in)

Figure 9- BPHE Inlet (P) and Outlet (P*) Pressure (psi) Curves for 20 GPM

P

P*

TOP

BOTTOM

63


16

17

18

19

20

21

22

23

24

25

0 1 2 3 4 5 6 7


Figure 10- BPHE Inlet (P) and Outlet (P*) Pressure (psi) Curves for 30 GPM


32

34

36

38

40

42

44

46

48

50

0 1 2 3 4 5 6 7


Figure 11- BPHE Inlet (P) and Outlet (P*) Pressure (psi) Curves for 40 GPM.


30

35

40

45

50

55

0 1 2 3 4 5 6 7



P

P*

P

P*

P

P*

64


17

22

27

32

37

42

47

52

0 1 2 3 4 5 6 7



Flow visualization of the flow distribution, for U-type flow, within the clear BPHE show that fluid flow is non-uniform across the plate’s surface suggesting maldistribution in the tested BPHE. From the visualization study it can be seen that the fluid moves directly across the port side of the heat exchanger at a higher velocity than the opposite side. Figure 14 illustrates the streamlines observed in the tested BPHE. This suggests that the heat-transfer coefficient across a plate’s surface is non-uniform and leads to inaccurate performance calculations if a constant heat transfer coefficient was used. Each of the flow visualizations were videotaped and analyzed to arrive at this conclusion. Observations also reveal several flow dead spots (see Figure 14). In general these areas remained consistent for the range of flows considered.

(a) (b)

Figure 14- (a) Typical Observed Dead Spots (b) Streamlines, in the Tested BPHE.

Summary and Conclusions An experimental study has been presented to investigate flow distribution

within a 60 plate BPHE. An overall pressure drop along with individual channel pressure differences have been collected and analyzed to develop a profile for flow distribution within a 60 plate BPHE. Visual inspection of a transparent BPHE revealed several areas of maldistribution and areas of dead flow. Data shows that maldistribution increases as the operating flow rate increases. Results show

P

P*

65

similarities to the results presented by Rao, Sunden and Das [1] for a PHE. Results from investigating the pressure drop across flow channels in various locations along both inlet and outlet ports revealed that non-uniform flow exists across individual channels. It was shown that pressure differences across the channels closest to the inlet ports (top) were highest, suggesting a higher fluid velocity in these channels. Entrance effects cause the fluid to disperse into to first few channels at a greater rate than in the bottom channel. Variation of flow leads to a non-uniform heat-transfer coefficient. This variance leads to accelerated fouling in areas exposed to the lower flows. Flow visualization studies also indicate that velocities across the entire plate’s surface varied. In addition, dead spots on the surface were visibly noticeable during the experiment. Overall, results indicate that maldistribution is present within the tested BPHE. Further testing on other flow configurations is currently being conducted. There is considerable difference in chevron design when comparing a typical plate used in a BPHE to one used in a PHE. Figure 15 shows the difference in plate design around the distribution area near the ports. This distribution pattern (shown in Figure 15a) can increase overall efficiency by dispersing fluid evenly across the plate’s surface. This design also reduces the occurrence of surface flow dead spots. New manifold distribution designs are currently being evaluated. Utilizing such features in a brazed unit could lower cost, decrease package size and increase efficiency of the unit. Operational costs and maldistribution can be minimized by optimizing the flow rate. Tests should be conducted to determine optimal flow rates. This testing has the potential to eliminate maldistribution and reduce fouling.

(a) (b)

Figure 15- Chevron Plates Characteristics (a) BPHE (b)PHE

66

References [1] Bassiouny, MK, and Martin H, Flow Distribution and Pressure Drop in Plate Heat Exchanger-II, U-Type Arrangement, Chem. Eng. Sci. 1984;39(4): 693-700. [2] Rao BP, Sunden B, Das SK, An experimental and theoretical study on the effect of flow maldistribution on the thermal performance of plate heat exchangers, ASME Journal of Heat Transfer 2005; 127: 334-342. [3] Rao BP, Das SK, An experimental study on the influence of flow maldistribution on the pressure drop across a plate heat exchanger, ASME Journal of Fluids Engineering 2004; 126: 680-691. [4] Rao BP, Kumar PK, Das SK, Effect of flow distribution to the channels on the thermal performance of a plate heat exchanger, Chemical Engineering Process 2002; 41 :49-58. [5] Kanaris AG, Mouza KA and Paras SV, Designing novel compact heat exchangers for the improved efficiency using CFD code. 1st International Conference, Athens, 8-10 September, 2004. [6] Lozano A , Barrerasa F ,Fueyob N ,Santodomingo S , The flow in an oil/water plate heat exchanger for the Automotive industry, Applied Thermal Engineering ; 2007; doi:10.1016/ j.applthermaleng. 2007.08.015. [7] Focke WW, Knibble PG, Flow visualization in parallel-plate ducts with corrugated walls, J. Fluid Mech. 1986; 165:73-77.


67

P-graph Methodology for Cost-Effective Reduction of Carbon Emissions Involving Fuel Cell Combined Cycles

Ferenc Friedler1, Petar Varbanov2, L. T. Fan3

1 DCS2, FIT, University of Pannonia, Egyetem u. 10, H-8200, Veszprém, Hungary 2 EC MC ERG ESCHAINS, FIT, University of Pannonia, Egyetem u. 10, H-8200, Veszprém, Hungary, e-mail:[email protected], [email protected]

3 Department of Chemical Engineering, Kansas State University, Institute of Systems Design and Optimization, 2045 Durland Hall, Manhattan, KS 66506, USA

Abstract Fuel cells are under extensive investigation for building combined energy cycles due to the higher efficiency potential they offer. Two kinds of High-Temperature Fuel Cells (HTFC) have been identified as best candidates for Fuel Cell Combined Cycles (FCCC) – Molten Carbonate Fuel Cells (MCFC) and Solid Oxide Fuel Cells (SOFC). The paper evaluates energy conversion systems involving FCCC subsystems, utilising biomass and/or fossil fuels, for reducing CO2 emissions. The significant combinatorial complexity is efficiently handled by using P-graph framework and algorithms. methodology for the synthesis of cost-optimal FCCC configurations is developed, accounting for the system carbon footprint. The results show that such systems using renewable fuels can be economically viable for wide range of conditions, due to the high energy efficiency of the FC-based systems.

Introduction The continuously increasing energy demands result in Greenhouse Gas Emissions (GHG) escalation. The current state-of-the-art covers mainly the traditional combined cycles (GTCC, IGCC) with efficiencies up to 55-60 %, using heat engines (GT and ST). HTFC are good options for higher efficiency. Present results on integrating HTFC with ST and GT indicate possibility to achieve both high efficiencies [1] and economic viability [2]. The use of biomass-derived fuels offers reduction of the CO2 emissions. Biomass can be utilised in two main ways – oxygen-deficient gasification and biogas digestion. FCCC are expensive to develop and resources should be economised. The presented tool for optimising the performance and economy of FCCC systems is a step in this direction. Systems for FCCC-based CHP and biomass processing present a large number of alternative routes, featuring combinatorial complexity. One approach to solving such problems employed Mathematical Programming (MP), where the selection of the operating units is represented by integer variables. For large problems applying MP becomes increasingly difficult – the solver needs to examine clearly infeasible combinations of integer variable values. In addition, building the problem superstructures heuristically is slow and error-prone.

68

For efficiently handling process synthesis problems of practical complexity the Process Network Synthesis (PNS) methodology based on the P-graph could be applied. P-graph is a rigorous mathematical tool for unambiguous representation of processing networks. The associated combinatorial instruments – axioms ensuring representation unambiguity [3] and algorithms generating the maximal network structure [4] and for generation of all possible solution structures [5], have several important properties making the approach superior to MP in solving network/process synthesis problems: • Automatic generation of the problem superstructure; • Optimising the generated superstructures avoids the examination of infeasible

combinations of binary variables representing the process units and this drastically reduces the combinatorial search space, making P-graph orders of magnitude more efficient than pure mathematical programming [5], [6].

Another important issue is the realistic evaluation of the CO2 minimisation potential. This issue has been studied in detail by Klemeš et al. [7, 8, 9]. Although biomass is nominally carbon-neutral, its processing contributes to certain small carbon footprint [9, 10].

Context definition: FCCC systems and biomass resources

Processing steps This study concentrates on evaluating the viability of using biomass as a primary resource. The processing architecture shown in Figure 1 is considered.

Fossil fuels

Biomass Processing:Gasificationor Digestion

Biofuel Energy Conversion:FCCCsBoilers

…

Power

Heat Figure 1. FCCC system boundary and processing steps

Efficiency of FC and combined cycles HTFCs are net sources of waste heat at temperatures above 700 ºC [2, 11]. Both MCFC and SOFC, define threshold heat integration problems with excess waste heat The FCs should be topping cycles with GT and ST as bottoming cycle options. FCCC system efficiencies vary with the FC operating temperature, the type of the bottoming cycle and with the degree of cycle integration [11]. HTFCs can be combined with different turbines - FC+GT and FC+ST or both: FC+GT+ST. The last combination results in only marginal improvements. A GT can be directly integrated with an FC (cheaper, less flexibility) or indirectly heated (more flexible, high-cost). There are two aspects how the fuel cell operating temperature affects the efficiency. From the diagram in Figure 2 [12] it is clear that the FC standalone efficiencies are strongly correlated with the operating temperature. Secondly, higher temperatures

69

favour higher potential for further power generation form the FC exhausts. Any drop in the temperature drastically decreases this potential.

FC efficiency vs temperature

0

10

20

30

40

50

60

80ºC(PEFC)

200ºC(PAFC)

700 ºC(MCFC)

1000 ºC(SOFC)

η MA

X, (%

)

FC efficiency vs temperature

0

10

20

30

40

50

60

80ºC(PEFC)

200ºC(PAFC)

700 ºC(MCFC)

1000 ºC(SOFC)

η MA

X, (%

)

Figure 2. Variation of FC efficiency with operating temperature

Process representation with P-graph P-graph is a directed bipartite graph, having two types of vertices – one for operating units and another for the objects representing material or energy flows/quantities, which are connected by directed arcs [3], [13]. Operating units and process streams are modelled by separate sets (O and M) and the arcs are expressed as ordered pairs. E.g., if an operation o1 ∈ O consumes material m1 ∈ M, then the arc representing this relationship is (m1, o1). Figure 3 illustrates the FCCC system representation using a conventional block-style diagram and a P-graph fragment. For this fragment: with the following sets: Vertices Arcs

FCCCO

CO,Q,W,FM 2

==

2CO,FCCC,Q,FCCC,W,FCCCOutletsFCCC,FInlets

==

FCCC

F

W Q

CO2FCCC

F

W Q

CO2

Q

FCCC

W CO2

F

Block-style flowsheet P-graph Legend F: Fuel;FCCC: Fuel Cell Combined Cycle; Q: Heat; W: Power

Figure 3 FCCC representations

Modelling procedures

General synthesis procedure In order to apply the P-graph approach, certain types of information need to be obtained, evaluated and supplied to the synthesis algorithms. This includes identification of the raw materials, products and intermediates; identification of the

70

candidate operating units; specification of the units’ performance; identification of upper and lower bounds on the capacities of the operating units.

Representation of the operating and capital costs The system operating costs and/or profits are estimated including: (i) Direct costs for fuels and raw materials (e.g. biomass) (ii) Specifically the biomass cost may vary widely and eventually cost nothing to the

CHP plant under investigation and the farmers may even need to pay to the plant. The biomass cost is defined as a factor in a sensitivity analysis.

(iii) The heat and power are sold at market prices, generating revenues. (iv) There are waste by-products – e.g. particulates and other residues, which are

impossible to process. These are associated with corresponding disposal costs. The capital costs of all operating units have been assumed to change linearly adhering to the form given below:

UCapBACC CCCC ⋅+= (1)

where the operating unit capacity is rated by a key inlet stream flowrate. Table 2 lists the capital cost coefficients used in the case study. More detailed evaluation can be performed using the data published by Taal, et all [14].

Optimisation objective The synthesis of the system requires defines an optimisation task. Several objectives are possible. The most obvious are the system profit to be maximised (cost to be minimised) and the amount of CO2 emissions to be minimised. Using profitability is most practical, since it drives the behaviour of the companies.

Sensitivity analysis procedure The current work aims at analysing the economic viability and the potential for reducing the CFP of energy conversion systems. The FCCC components are experimental technology having little or no market penetration. Many of their parameters are uncertain. The real issue is – what is the range of conditions for which FCCC-based systems can minimise the corresponding CO2 emissions while featuring maximum profit? The following parameters have been varied to evaluate the sensitivity of the systems economic and environmental performance: • Price of the biomass (from -10 to 30 €/MWh). • CO2 tax levels: 0 and 40 €/t. • Payback period for the process capital costs (10 and 20 years).

Applying P-graph: heat and power generation using FCCC

Case study description The problem considered requires CHP generation from waste biomass and/or natural gas, using a number of FCCC options. It is assumed that the biomass is suitable for both gasification and anaerobic digestion. Power and heat demands are 10 MW and 15 MW. The energy prices are: 100 €/MWh for power, 30 €/MWh for heat and 30

71

€/MWh for natural gas. The price of the fertiliser by-product is 50 €/t. The carbon tax and the biomass price have been chosen as factors for the sensitivity analysis. The plant payback time is initially 10 years. The carbon footprint for biomass is 0.025 t/MWh (t CO2 per MWh of biomass) [9] and that of natural gas is 0.2063 t/MWh. The fertiliser yield in the biogas digester is 0.0768 t/MWh.

Materials and streams The materials/streams are listed in Table 1. The prices given in Table 1 follow a strict convention about the sign. Inputs are assigned positive prices if the plant has to pay for them and negative ones if it receives payment. Similarly, all outputs generating revenues are assigned positive prices and those generating costs – negative prices.

Table 1.Materials and streams

Stream Classification Description Price BM Raw material Agricultural residues Varied BG Intermediate Biogas - BR Product Biomass residue (solid) -10 €/t CO2 Product CO2 emissions Varied FRT Product Fertiliser by-product from the digester 50 €/t NG Raw material Natural gas 36.8 €/MWhPR Product Particulates left from cleaning the syngas -10 €/t Q40 Intermediate Steam at P = 40 bar(a) - Q5 Product Steam at P = 5 bar(a) 30 €/MWh RSG Intermediate Raw synthesis gas - SG Intermediate Clean synthesis gas - W Product Electrical power 100 €/MWh The performance and economic data for the operating units are specified in Table 2.

Candidate operating units Figure 4 and Figure 5 show the indentified options.

BM

BMG

RSGBR CO2

RSG

SGF

SGPR

BM

BG

BGD

FRTCO2 Biomass gasifier Syngas filter Biogas digester

Legend BM: Biomass; BR: Biomass residues; RSG: Raw synthesis gas; PR: Particulates; SG: Synthesis gas; BG: Biogas; FRT: Fertiliser

Figure 4. Fuel preparation (biomass processing) options

72

Q

FCCC

W

CO2

F BG

Q40

BLR_BG

NG

BLR_NG

CO2

Q40 F: Fuels FCCC Q: steam Steam detailsNG: Natural gas MCFC-GT Q1 P = 1 barBG: Biogas MCFC-ST Q2 P = 2 barSG: Syngas SOFC-GT Q5 P = 5 bar

SOFC-ST Q10 P = 10 barQ20 P = 20 barQ40 P = 40 bar

Figure 5. Energy conversion options

Results and discussion CHP networks have been synthesised for the defined options using the P-graph algorithms developed gradually by Friedler et al. [1-4]. This has been performed for the entire range of conditions described above. The initial conditions include 0 €/t CO2 tax, biomass price variation, (-10÷30) €/MWh, and 10 years payback time

Table 2. Capital cost coefficients and performance data used in the case study

Unit Key stream

ACC BCC Min cap.

Max cap.

Performance

- MW € €/MW MW MW - BGD BM 4939 0.327 0 100 BG: 0.58 MW/MW

FRT: 0.0768 t/MW CO2: 0.025 t/MW

BLR_BG BG 1646 0.109 0 100 Q40: 0.85 MW/MW BLR_NG NG 1646 0.109 0 100 Q40: 0.88 MW/MW

CO2: 0.2063 t/MW BMG BM 42000 0.080 0 100 RSG: 0.65 MW/MW

BR: 0.0811 t/MW CO2: 0.025 t/MW

MCFC-GT F 7.231•106 0.251 0 100 W: 0.580÷ 0.672 MW/MW CO2: 0.000 ÷ 0.2063 t/MW Q: 0 ÷ 0.250 MW/MW

MCFC-ST F 4.6 •106 0.051 0 100 W: 0.590÷ 0.670 MW/MW CO2: 0.000 ÷ 0.2063 t/MW Q: 0.000 ÷ 0.250 MW/MW

SOFC-GT F 9.131•106 0.270 0 100 W: 0.630÷ 0.695 MW/MW CO2: 0.000 ÷ 0.2063 t/MW Q: 0.000 ÷ 0.241 MW/MW

SOFC-ST F 6.5 •106 0.070 0 100 W: 0.600÷ 0.695 MW/MW CO2: 0.000 ÷ 0.2063 t/MW Q: 0.000 ÷ 0.240 MW/MW

SGF RSG 6500 0.015 0 100 PR: 5 •10-4 t/MWh SG: 0.99 MW/MW

73

.

BM

BG

Q40

W 10.0 MW

BGD

55.2 MW

32.0MW

FRT

4.2 t/h

Q5

LD_40_5

12.8 MW

BLR_BG

15.0 MW

CO2

1.4 t/h

FCCC_36(MCFC+ST)

16.9 MW

2.2 MW

15.0 MW

15.1 MW0.17t/h

Figure 6. Network 1

BM

BMG

RSGBR

25.9 MW

2.1 t/h

SGF

SGPR

8·10-3 t/h

BG

Q40

FCCC_60(MCFC+ST)

W 10.0 MW

BGD

FRT

2.0 t/h

Q5

2.2 MW

LD_40_5

BLR_BG

15.0 MW

CO2

0.6 t/h0.7 t/h

16.8MW

16.7 MW 0.17

t/h

26.0 MW

12.8 MW

15.1MW

Figure 7. Network 2

BM

BMG

RSGBR

25.9 MW

2.1 t/h

SGF

SGPR

BG

Q40

FCCC_57(MCFC+ST)

W10.0 MW

BGD

FRT

1.8 t/h

Q53.3 MW

LD_40_5

BLR_BG

15.0 MW

0.6 t/h 0.6 t/h

16.8MW

16.7 MW

0.17t/h

23.7 MW

11.7 MW

11.7 MW

13.7MW

8·10-3 t/h

CO2

Figure 8. Network 3

BM

BMG

RSGBR

24.3 MW

2.0 t/h

SGF

SGPR

BG

Q40

FCCC_69(SOFC+ST)

W 10.0 MW

BGD

FRT

1.8 t/h

Q5

3.4 MW

LD_40_5

BLR_BG

15.0 MW

0.6 t/h 0.6 t/h

15.8MW

15.6 MW

0.16t/h

23.5 MW

11.6 MW

11.6 MW

13.6MW

CO2

8·10-3 t/h

Figure 9. Network 4

The networks resulting for these conditions are presented in Figures 6 through 11. The corresponding annual profit and CO2 emissions are given in the first two curves of Figures 12 and 13. It can be noticed that for the cheapest biomass price, biomass is the only primary energy source. Also the main route for power generation is via biogas using lower-efficiency FCCC blocks. For higher biomass prices, gasification and increasingly more efficient FCCC blocks are used. At biomass price of 20.35 €/MWh, (Figure 10, Network 5) the auxiliary heat production switches from biogas to natural gas, while due to the high efficiency of the FCCC subsystems, the main CHP generation is still based on biomass gasification. At biomass price 23.57 €/MWh, using natural gas becomes more economic completely, which is reflected by switching the FCCC CHP blocks to using this fuel. When subsequently the payback period is increased from 10 to 20 years, three more networks are generated.

74

BM

BMG

RSGBR

24.3 MW

2.0 t/h

SGF

SGPR

NG

Q40

FCCC_69(SOFC+ST)

W10.0 MW

13.1MW

Q53.4 MW

LD_40_5

BLR_NG

15.0 MW

CO2

0.6 t/h 2.7 t/h

15.8MW

15.6 MW

11.6 MW

11.6 MW0.16t/h8·10-3 t/h

Figure 10. Network 5

NG

Q40

FCCC_09(MCFC+ST)

W 10.0 MW

13.0MW

Q5

3.5 MW

LD_40_5

BLR_NG

15.0 MW

2.7 t/h

11.5 MW

11.5 MW

3.3 t/h

16.1MW

CO2

Figure 11. Network 6 The energy network topology changes happen in steps (Figure 13). This shows that the topologies are relatively resilient to the variations in the biomass price. The profit changes essentially linearly.

-5

0

5

10

15

20

-10 0 10 20 30

Mill

ions

Price, EUR/MWh

Prof

it, E

UR

/y

CO2 tax 00 10 yr payback CO2 tax 40-10 yr paybackCO2 tax 00 - 20 yr payback CO2 tax 40 - 20 yr payback

10

12

14

16

-10 -8 -6 -4 -2 0

Figure 12. Profits vs. biomass price

The sensitivity analysis (Figures 12 and 13) reveals that the main factor determining the network structures is the competition between natural gas and biomass prices. The sensitivity towards the other two factors – CO2 tax level and payback period is not as significant. Moreover, imposing a CO2 tax does not significantly reduce the corresponding emissions, but slightly widens the range of profitability of biomass utilisation. Also, even using biomass produces a certain CO2 footprint, so the tax notably reduces also the profitability of the biomass-based systems.

75

10

15

20

25

30

35

40

45

50

-10 0 10 20 30 40 50

Thou

sand

s

Price, EUR/MWh

CO

2, t/

y

CO2 tax 00 - 10 yr PB CO2 tax 40 - 10 yr PBCO2 tax 00 - 20 yr PB Co2 tax - 40 - 20 yr PB

10.5

11.0

11.5

12.0

12.5

13.0

-10 0 10 20

Figure 13. CO2 emission levels vs. biomass price

Conclusions and future work This contribution provides a tool based on a procedure for efficient evaluation of early-stage energy technologies, following the approach set by the EMINENT2 project [15], [16] specifying a set of market conditions and then testing the resilience of the design against variations of key parameters. The task of designing a complete energy system involves significant combinatorial complexity. This cannot be efficiently handled by Integer Programming procedures. The P-graph framework and its associated algorithms are capable of efficiently handling exactly this type of complexity, inherent to network optimisation and appear to be some of the best tools for solving this task. The presented process synthesis procedure can be readily used for evaluating technologies in their early stages of development, such as FC / FCCC. The case study shows that FCCC systems can be economical over a wide range of economic conditions. From the presented material it can be concluded that biomass can be a viable energy supply option, where the possible high efficiencies also mean smaller resource demands. The future work should concentrate on improving the integration of the unit process models with the network synthesis procedure, as well as evaluation of the dynamic and variability aspects of the concerned energy technologies and the associated biomass and fuel resources. With regard to the scope of the studies, considering complete supply chains for energy and value-added products as well as CO2 transport, storage and sequestration is necessary.

Acknowledgements The financial support from the EC projects and EMINENT2 – TREN/05/FP6EN/ S07.56209/019886 and ESCHAINS – MERG-CT-2007/46579 are gratefully acknowledged.

76

References [1] Massardo, A.L., Bosio, B., Assessment of Molten Carbonate Fuel Cell Models and Integration with Gas and Steam Cycles, Journal of Engineering for Gas Turbines and Power, 124 (2002), 103-109. [2] Varbanov P., Klemeš, J., Shah, R.K., Shihn, H., Power Cycle Integration and Effic. Increase of Molten Carbonate Fuel Cell Systems, J. of Fuel Cell Science and Technology, 3(4) (2006), 375-83. [3] Friedler, F., Tarjan, K., Huang, Y.W., Fan, L.T., Graph-Theor. Approach to Proc. Synthesis: Axioms and Theorems, Chem. Eng. Sci., 47(8) (1992), 1972-1988. [4] Friedler, F., Tarjan, K., Huang, Y.W., Fan, L.T., Graph-Theoretic Approach to Process Synthesis: Polynomial Algorithm for Maximal Structure Generation, Comput. Chem. Eng., 17(9) (1993), 929-942. [5] Friedler, F., Varga, J.B., Fan, L.T., Decision-Mapping: A Tool for Consistent and Complete Decisions in Proc. Synthesis, Chem. Eng. Sci., 50(11) (1995), 1755-1768. [6] Friedler, F., Varga, J.B., Fehér, E., Fan, L.T., Combinatorially Accelerated Branch-and-Bound Method for Solving the MIP Model of Process Network Synthesis, In State of the Art in Global Optimization, Ed. Floudas, C.A. and Pardalos, P.M., Kluwer Academic Publishers, 1996, Boston, Massachusets, pages 609-626. [7] Klemeš J., Cockerill, T, Bulatov I., Shackely, S., Gough, C., Engineering Feasibility of carbon dioxide capture and storage, Chapter in: C. Cough, S. Shackley (ed) Carbon Capture and its Storage: An Integrated Assessment, Ashgate Publishing Ltd, pp 43-82, 2006. [8] Klemeš J.; Bulatov I., Cockeril T, Techno-Economic Modelling and Cost Functions of CO2 Capture Processes, Computers & Chemical Engineering, 31(5-6) (2006), 445-455. [9] Klemeš J., Perry S., Bulatov I., The Feasibility of Micro Renewable Energies in Reducing the Carbon Footprint of Energy Use in Buildings, Proceedings of WellBeing Indoors Clima 2007 Conference, 10-14 June Helsinki, Finland, 2007, p. 303. [10] Perry S., Klemeš J., Bulatov I ., Integrating renewable energy sources into energy systems for the reduction of carbon footprints of buildings and building complexes, PRES’07, Ischia, ed. Jiri Klemes, Chemical Engineering Transactions, 12 (2007), 593-598 [11] Varbanov, P., Klemeš, J., Friedler, F., Critical Analysis of Fuel Cell Combined Cycles for Development of Low-Carbon Energy Technologies, Chem. Eng. Trans., Ed. J.Klemes, 12 (2007), p.739, AIDIC, PRES’07, June 24-27, Naples, Italy. [12] Yamamoto, O., Solid Oxide Fuel Cells: Fundamental Aspects and Prospects, Electrochimica Acta, 45 (2000), 2423-2435. [13] Nagy, A.B., R. Adonyi, L., Halasz, Friedler, F. Fan, L.T., Integrated Synthesis of Process and Heat Exchanger Networks: Algorithmic Approach, Applied Thermal Engineering, 21 (2001), 1407-1427. [14] Taal, M., Bulatov, I., Klemeš, J., Stehlik, P., Cost Estimation and Energy Price Forecast for Economic Evaluation of Retrofit Projects, Applied Thermal Engineering 23 (2003), 1819 – 1835. [15] Klemeš, J., Bulatov, I., Koppejan, J., Friedler, F., Hetland, J., Novel Energy Saving Technologies Evaluation Tool, T5-142, 17th European Symposium on Computer Aided Process Engineering – ESCAPE17, Bucharest, ed. V. Plesu and P.S. Agachi, 2007 Elsevier B.V./Ltd. p. 1035-1040. [16] Klemeš, J., Zhang, N., Bulatov, I., Jansen, P., Koppejan, J., Novel Energy Saving Technologies Assessment by EMINENT Evaluation Tool, PRES’05, Giardini Naxos, ed Jiri Klemeš, Chemical Engineering Transactions, 7 (2005), 163 – 167.


77

Reactive and non-reactive distillation sequences: Energy saving by process integration

Ivo Müller, Eugeny Y. Kenig

Chair of Fluid Separations, Dortmund University of Technology, 44227, Dortmund, Germany, email: [email protected]; fax: +49 231 755-3035

Abstract In this work, the energy consumption for distillation sequences with different integration degree is investigated with the aim to identify the optimal configuration. For this investigation, a non-reactive system (ternary alcohol mixture) and a reactive system (transesterification of dimethylcarbonate) are chosen, and, for each system, several promising configurations with increasing integration degree are identified. Their analysis shows that the unit integration leads to significant savings in energy consumption and production costs for both non-reactive and reactive systems.

Introduction Distillation represents a widely used, yet most energy intensive step in chemical and process industries. Since energy costs have permanently been increasing, an essential improvement of distillation performance has become more and more important. Therefore, in the last decades, significant effort has been undertaken towards optimisation of distillation units. Integration of different operations within a single unit represents an interesting approach to achieve significant energy savings. Compared to serial sequences of conventional distillation columns, the well-known dividing wall column (DWC) offers better thermodynamic efficiency, reduced energy consumption and costs [1]. Recently, this principle has also been applied to reactive distillation processes resulting in the co-called reactive dividing wall column (RDWC, see [2]). Both DWC and RDWC represent processes with a very high degree of integration. However, integration can sometimes be disadvantageous, as it may reduce the operating window. Therefore, less integrated processes still remain an alternative and should be considered for process design. In this work, the energy consumption for distillation sequences with different integration degree is investigated with the aim to identify the optimal configuration.

Estimation of feasible column set-ups In order to investigate different configurations, their set-up and operating parameters have to be determined. For the DWC and RDWC, this is especially difficult regarding the large number of required design parameters. In addition to conventional column variables (e.g., reflux ratio, distillate stream), further parameters, e.g., location and height of the dividing wall, distribution of liquid above the dividing wall (also called liquid ratio), have to be determined. To address this issue, Triantafyllou and Smith [3] presented the so-called decomposition method for non-reactive dividing wall columns.

78

The decomposition simplifies the design problem by replacing a complex integrated configuration by a (disintegrated) sequence of conventional single non-reactive and reactive columns. The latter can be designed by applying already existing approaches. This decomposition method has been extended to cover reactive systems by Mueller et al. [4]. In accordance with the approach of Triantafyllou and Smith [3], a suitable design of single non-reactive columns is determined using already existing short-cut methods, whereas for the single reactive columns, a more rigorous rate-based approach is applied. This procedure allows a quick design estimation under consideration of required product specifications (e.g., product concentration). Decomposition and short-cut methods are based on many simplifying assumptions; therefore, a reasonable testing of the decomposition results by a more accurate method is required. In our work, we use a fully rate-based description of both decomposed and integrated configurations. This approach developed earlier [5,6] directly considers actual rates of multicomponent mass and heat transport between liquid and vapour phases.

Software implementation All models are implemented into the simulation environment ASPEN Custom ModelerTM (ACM). ACM is an equation-oriented simulator which simultaneously solves the whole set of equations. In addition, ACM allows an access to the software ASPEN PropertiesTM to calculate pure component and mixture properties (e.g., vapour-liquid equilibrium). The rate-based description of distillation columns, especially for the RDWC configuration, often results in large and highly non-linear algebraic systems of equations. Therefore, a proper implementation and solution of the model equations is important. In addition, availability of good starting values is usually a deciding factor. To generate the starting values, a simpler model is used. The latter represents an extended equilibrium stage model, in which chemical reaction rates are taken into account.

Evaluation criteria Beside energy consumption, costs represent an important criterion for the evaluation of process alternatives. Therefore, in this work, the rate-based models are extended by the capital and operational cost calculation according to Mulet et al. [7] and Hirschberg [8]. The total investment cost is considered to be the expenses required for the manufacturing of the equipment. The total column cost consists of shell cost and packing cost. The cost correlation for the column shell is taken from Mulet et al. [7] and is updated to year 2002 using cost indices [9]. For the packing, a value of 2000 $/m3 is assumed as a first approximation. Total investment costs of condenser and reboiler are calculated as a function of heat exchange area [8] and are also updated to year 2002. For an easier comparison, the investment costs are annualised to account for a linear amortisation over 10 years with 10% interest rate. The total operating cost is generally considered to be expenses needed for the operation and maintenance of the equipment. As a first approximation, the total operating cost for the columns is assumed to be similar to the total cost required for

79

replacements of the column packing over a period of 10 years. For condenser and reboiler, the total operating cost is assumed to be equal to the annual cost of the utilities required. The operating cost for cooling water (15 °C) and steam (7 bar, 162°C) are taken to be 0.06 $/ton and 7.5 $/ton respectively [8]. The investment and operating costs of the pre-reactor are not considered, as these costs are the same for all configurations studied. The sum of annualised capital and operating costs is called annual total costs.

Results for non-reactive dividing wall column For the present study, five different distillation sequences are considered, namely, the direct configuration, the indirect configurations, the direct/indirect combined sequence, the prefractionator configuration, and the dividing wall column. The first two configurations represent conventional distillation sequences in which the components are separated in correspondence with their boiling temperatures. Such a separation often requires high energy input. (Fig 1).

Figure 1. Conventional column arrangements for separation of ternary mixtures: direct sequence (left); indirect sequence (right)

Figure 2. Column arrangements for separation of ternary mixtures with increasing integration degree: direct/indirect sequence (left); prefractionator configuration

(middle); dividing wall column (right) From the standpoint of thermodynamics (and, accordingly, efficient energy use), an optimal distillation arrangement for the separation of a ternary mixture requires three columns (Fig. 2, left) [10]. In the first column, the lightest (A) and the heaviest (C) components are separated very efficiently, due to their high relative volatilities. The

80

intermediate boiling component (B) appears in both top and bottom streams of column C1 and is separated in the second and third columns (C2 and C3). Although the direct/indirect sequence shown in Figure 2 (left) is thermodynamically more attractive than the direct or indirect arrangements, its set-up requires an additional column, a reboiler and a condenser. This increases the total energy demand (as the mixture has to be evaporated and condensed in each single column). The reboiler of the column C3 and the condenser of column C2 can be avoided by thermally coupling of both columns; thereby a reduction of equipment units can be achieved (Fig. 2, middle). This sequence is called prefractionator configuration. The highest degree of integration is realised by the dividing wall column (Fig. 2, right), when a vertical partition (wall) is introduced into a distillation column. This unit enables the separation of a ternary mixture within one single column shell. The feed used for the simulation is a ternary mixture methanol-isopropanol-butanol. The boiling points of the pure components at atmospheric pressure are 64°C (methanol), 82°C (isopropanol) and 117°C (butanol). The feed flow rate is 10 m3/hr, the feed temperature is 78.5°C and column pressure is atmospheric. Purities around 95 mol% for all components are required. The simulation results for a feed composition of 20 mol% butanol, 50 mol% isopropanol and 30 mol% methanol are presented in Figure 3 and 4. Figure 3 demonstrates that the energy consumptions of the direct and indirect configuration are almost equal. The application of the direct/indirect configuration offers only small savings, whereas significant energy savings around 40% can be realised by the integration step towards the prefractionator configuration. The unit with the highest integration, the dividing wall column, provides almost similar savings.

Figure 3. Reboiler and condenser heat duties for different distillation sequences Figure 4 shows the annualized capital and operating costs for each distillation sequence. The costs are related to the maximum value for all sequences, which is the total capital cost of the indirect configuration. It can be seen that the capital cost of the dividing wall column is around 40% lower than that of the cheapest conventional configuration (direct configuration). Savings can also be realized with the prefractionator configuration, with costs reduction about 32%. The annualized

0

1

2

3

DirectConfiguration

IndirectConfiguration

Direct/IndirectConfiguration

PrefractionatorConfiguration

Dividing WallColumn

Hea

t dut

y (1

06 ·W)

ReboilerCondenser

81

capital costs are low compared to the operating costs (< 15%), which demonstrates that significant cost savings can only be realized by more energy efficient processes.

Figure 4. Related capital and operating costs for different distillation sequences

Results for reactive dividing wall column As a test system, the transesterification of dimethyl carbonate (DMC) with ethanol (EtOH) to diethyl carbonate (DEC) and methanol (MeOH) is chosen. This system is industrially important, as the products are essential reactants and solvents for a number of chemical and pharmaceutical processes, e.g., production of polycarbonates and antibiotics [11, 12]. The reaction system consists of two consecutive reactions with the intermediate product ethyl methyl carbonate (EMC):

⇔⇔

DMC+EtOH EMC+MeOHEMC+EtOH DEC+MeOH

Richter et al. [13] found that potassium carbonate is an appropriate heterogeneous catalyst, which allows the application of heterogeneously catalysed RD, and determined a suitable pseudo-homogenous kinetics. Different configurations have been investigated for the DEC synthesis by Richter et al. [13] for RD and by Mueller and Kenig [2] for RDWC. These investigations showed that RD and RDWC lead to significantly increased conversion and selectivity as compared to conventional reactors. At atmospheric pressure, the system consisting of DMC, EMC, DEC, MeOH and EtOH does not have any miscibility gap; however, three binary homogenous azeotropes appear. The boiling temperatures of the pure components and azeotropes are presented in Table 1. For the description of the mixture thermodynamics, the UNIQUAC approach is chosen. The applied UNIQUAC parameters have been validated using VLE data from literature (see [2]). For the investigation, atmospheric pressure in the columns is assumed and the total feed molar stream is adjusted to 30 mol/s. A parameter study for a single RD column is performed to ascertain feasible operating and set-up conditions required for high conversion and selectivity. The obtained results are applied later to adjust the design parameters of the integrated configurations in order to achieve corresponding liquid and vapour streams in their reactive zone. The usage of a pre-reactor, which increases the total catalyst mass within the process, is found to be advantageous for

0

20

40

60

80

100

DirectConfiguration

IndirectConfiguration

Direct/IndirectConfiguration

PrefractionatorConfiguration

Dividing WallColumn

Rel

ated

cos

ts p

er y

ear

(max

imum

val

ue =

100

%) Annualized capital cost

Total operating cost

82

this test system, as it improves the process performance. The pre-reactor is described as an ideal reactor with equilibrium composition in the outlet stream which is fed to the column. Following variables are varied: reflux ratio, distillate stream, feed position and pre-reactor feed ratio of the reactants EtOH/DMC.

Table 1. Boiling temperatures at 1.013 bar for the reactive system Component/Azeotrope Boiling temperature Composition

MeOH/DMC 63.8 °C 87:13 mol/mol MeOH 64.7 °C -

EtOH/EMC 74.9 °C 95:5 mol/mol EtOH/DMC 77.8 °C 69:31 mol/mol

EtOH 78.4 °C - DMC 90.4 °C - EMC 107.9 °C - DEC 126.5 °C -

In Table 2, promising design parameters are shown for two selected pre-reactor feed ratios, namely, a stoichiometric feed ratio (EtOH/DMC = 2:1) and an excess feed ratio (EtOH/DMC = 4:1). The column set-up, packing height and diameter are the same for both feed ratios. The reflux ratio is set equal to 5 as it is optimal for high conversion and selectivity, while the distillate-to-feed ratio is adjusted to achieve a separation cut between EtOH and EMC. This cut is advantageous, because, in this case, the reactants EtOH and DMC are largely distributed within the reactive zone and, thus, the reaction is enhanced. It is obvious, that EtOH excess shifts equilibrium of both reactions to their product side resulting in a higher conversion and selectivity of DEC. It is found, that further increase of the feed ratio does not improve conversion and selectivity.

Table 2. Design parameters for single RD column with pre-reactor

Stoichiometric

pre-reactor feed (EtOH/DMC = 2:1)

Excess pre-reactor feed

(EtOH/DMC = 4:1) Packing height 4.8 m 4.8 m

Packing diameter 3.4 m 3.4 m Reflux-ratio 5 5

D/F-ratio 0.75 mol/mol 0.87 mol/mol Conversion

DMC 85.79 % 92.39 %

Selectivity DMC/EMC 66.53 % 74.92 %

The single reactive distillation column permits high conversion and selectivity, but it does not allow a complete separation of the reaction mixture consisting of products and non-converted reactants. The latter can be realised with additional non-reactive distillation columns. This combination of reactive distillation and subsequent separation results in a low integrated three-column configuration shown in Figure 5 (left) consisting of a pre-reactor, a reactive distillation column and two non-reactive distillation columns. The condenser of column C2 and reboiler of column C3 can be avoided by thermally coupling both columns to column C2* (Fig. 5, middle). A further

83

integration is achieved by combining column C1 and C2* to the highly integrated RDWC (Fig. 5, right).

Figure 5. Configurations for the DEC-synthesis with increasing integration degree: three-column configuration (left); prefractionator configuration (middle); reactive dividing wall column (right); (grey area: reactive zone)

Table 3. Analysis of configurations with different integration degrees

Three-column

configuration

Prefractionator configuration

Reactive dividing

wall columnConversion DMC 85.9 % 85.9 % 80.2 %

Selectivity DMC/DEC 66.8 % 66.8 % 72.3 % DEC purity 95.0 mol% 94.8 mol% 99.7 mol%

DEC production rate 1.80·107 kg / yr

1.87·107

kg / yr 1.97·107 Kg / yr E

TOH

/DM

C

= 2:

1

Specific reboiler duty 1.75·107 J / kg DEC

1.62·107 J / kg DEC

6.36·106 J / kg DEC

Conversion DMC 93.6 % 93.6 % 95.5 % Selectivity DMC/DEC 77.7 % 77.7 % 87.0 % DEC purity 94.8 mol% 96.5 mol% 94.9 mol%

DEC production rate 1.20·107 kg / yr

1.26·107 kg / yr

1.62·107 kg / yr

ETO

H/D

MC

=

4:1

Specific reboiler duty 2.08·107

J / kg DEC 1.91·107

J / kg DEC 8.32·106

J / kg DEC The simulation results in Table 3 demonstrate that, for all investigated configurations, high DEC conversion and selectivity is attained. Besides, the product streams reach sufficient purity (DEC > 94 mol%). This demonstrates that, for the DEC synthesis, the unit operations reaction and separation can be integrated without any loss in performance. In addition, the positive effect of ethanol excess in the feed, previously found for single RD, can be maintained for integrated processes.

84

Another important effect of the ethanol excess is that high DMC conversion enables higher purities (>90 mol%) of methanol in the distillate streams to be achieved (Table 3). This purities are usually limited by the binary minimum azeotrope formed by methanol and non-reacted DMC. As the produced amount of DEC in the bottom stream changes (this effect is especially significant when the feed ratio varies), the total reboiler duties are referred to one kilogram of produced DEC in the bottom stream. From the energy point of view, the first integration step towards the prefractionator configuration offers only small energy savings around 9% (Table 3, EtOH/DMC = 2:1). For the excess feed ratio (EtOH/DMC = 4:1), similar savings can be found. The configuration with the highest degree of integration, the RDWC, allows appreciable energy savings of about 65% for the stoichiometric and about 55% for the excess feed ratio compared to the corresponding three-column configuration. In addition, it can be shown, that the positive effect of ethanol excess in the feed (higher conversion and selectivity) can be reached at the expense of higher specific energy consumption (about 20% more compared with the stoichiometric feed ratio).

0%

20%

40%

60%

80%

100%

three-columnconfiguration

prefractionatorconfiguration

reactivedividing wall

column

rela

ted

cost

s pe

r kilo

gram

m D

EC(th

ree-

colu

mn

conf

ig. w

ith e

xces

s fe

ed ra

tio =

100

%)

stoichiometric feed ratioexcess feed ratio

Figure 6. Comparison of the specific production costs of DEC

The specific production costs for the different configurations are presented in Figure 6. These costs include the total operating and the annualised capital cost related to one kilogram of DEC produced. These data confirm the conclusion already drawn for specific reboiler duty: whereas the integration step towards the prefractionator configuration offers only small savings, the application of the RDWC reduces the cost by more than a half.

Conclusions The present study deals with different distillation sequences including those based on thermal coupling and integration principles, with the aim to reduce energy consumption and total costs. A combination of short-cut and rate-based models is applied for a rapid determination of the required set-up and operating conditions and for obtaining detailed information about the process behaviour and costs. The

85

developed design and analysis approach is of a general nature and can be applied to any non-reactive and reactive distillation system. For the non-reactive test case, the ternary alcohol mixture methanol-isoporpanol-butanol is chosen. For this ideal system, the product purities predicted by the short-cut models are confirmed by the rate-based simulations. In addition, the investigations show that the capital and operation costs can be significantly lowered using the prefractionator configuration as compared to conventional column arrangements. The application of the high-integrated dividing wall column leads to a further decrease of the costs. The conventional direct and indirect configurations are generally less efficient. For the reactive test case, the DEC synthesis is investigated for three feasible configurations. The rate-based analysis confirms the applicability of all configurations due to their high conversion, selectivity and product purity. On the other hand, the comparison of the specific reboiler duties and specific total cost shows that the process with the highest integration degree, the reactive dividing wall column, provides the best energy and cost efficiency.

References [1] Schultz M A, Stewart D G, Harris J M, Rosenblum S P, Shakur M S, O’Brien D E. Reduce costs with dividing-wall columns. Chemical Engineering Progress 2002; 98 (5): 64-71. [2] Mueller I, Kenig E Y. Reactive distillation in a dividing wall column: Rate-based modeling and simulation. Industrial Engineering and Chemistry Research 2007; 46: 3709-19. [3] Triantafyllou C, Smith R. The design and optimization of fully thermally coupled distillation columns. Transactions of Institution of Chemical Engineers - Part A 1992; 70: 118-32. [4] Mueller I, Pech C, Bhatia D, Kenig E Y. Rate-based analysis of reactive distillation sequences with different degrees of integration. Chemical Engineering Science 2007; 62: 7327-35. [5] Kenig E Y, Górak A. A film model based approach for simulation of multicomponent reactive separation. Chemical Engineering and Processing 1995; 34: 97-103. [6] Noeres C, Kenig E Y, Górak A. Modelling of reactive separation processes: Reactive absorption and reactive distillation. Chemical Engineering and Processing 2003; 42: 157-78. [7] Mulet A, Corripio A B, Evans L B. Estimate costs of distillation and absorption towers via correlations. Chemical Engineering 1981; 28: 180-83. [8] Hirschberg H G. Handbuch Verfahrenstechnik und Anlagenbau. Berlin Heidelberg, New-York: Springer Verlag, 1999. [9] Peters M S, Timmerhaus K D, West R E. Plant Design and Economics for Chemical Engineers. Boston: McGraw-Hill, 2003. [10] Shah, P. B.: Squeeze more out of complex columns. Chem. Eng. Progress 2002; 98 (7): 46-55. [11] Parrish J P, Salvatore R N, Jung K W. Perspectives on alkyl carbonates in organic synthesis. Tetrahedron 2000; 56: 8207–37.

86

[12] Luo H-P, Xiao W-D. A reactive distillation process for a cascade and azeotropic reaction system: Carbonylation of ethanol with dimethyl carbonate. Chemical Engineering Science 2001; 56: 403–10. [13] Richter J, Zielinska-Nadolska I, Górak A. Transesterification of dimethyl carbonate via reactive distillation, Proceedings of PRES’2004, Prague (Czech Republic), 2004: 1328–29.


87

Evaluation of Heat-integrated Distillation Schemes

Mansour Emtir Libyan Petroleum Institute, P.O. Box 6431 Tripoli, Libya

e-mail: [email protected], fax: +218 21 4836820

Abstract Conventional and non-conventional distillation schemes for the separation of ternary mixture are investigated and optimized based on shortcut calculations, pinch analysis and rigorous simulation in order to achieve the highest saving values in energy and total annual cost (TAC) as the optimization objective function. The studied chemical system is benzene, toluene and m-xylene with feed composition of (25/50/25) and 99.9 mol % product purity. The optimization parameters are feed compositions. The state of feed condition is considered as the optimization parameters; feed at 20 °C, liquid at bubble point and vapor at dew point. The results from shortcut and pinch analysis are found to be very close to rigorous simulation results regarding number of stages, reflux ratios and utilizing trim reboiler or trim condenser in studied distillation schemes. Both shortcut and rigorous optimization results indicate that heat-integrated schemes are consuming less energy compared to non-integrated distillation schemes, consequently TACs of heat-integrated schemes are attractive compared to non-integrated schemes. The state of feed conditions is playing imprtant role on the percentage of saving and direction of integration.

Introduction Distillation units are the most widely used technique for the separation of fluid mixtures in chemical and petrochemical industry. It is known that distillation is used for the separation of about 95% of all fluid separations in the chemical industry, and that around 3% of the total energy consumption in the world is used in distillation units (Hewitt et al [1]). The main disadvantage of the distillation is its high-energy requirement. As a result, new distillation sequences are emerging in order to reduce or improve the use of energy so that, there are several techniques which used to overcome this problem like integration of the distillation column with the overall processes which can give significant energy saving, e.g. Smith and Linnhoff [2], Mizsey and Fonyo [3], but these kinds of improvements can be limited. There are different configurations or distillation schemes that can be applied to get more energy saving, like integration of distillation columns with forward or backward heat integration, side-stripper, side-rectifier, fully thermally coupled distillation column (Petlyuk column) or dividing-wall column, heat-integrated of sloppy sequence, and double heat integration sequence. Energy-integrated distillation schemes give a great promise of energy savings up to about 70%. In addition to saving energy, which are accompanied by reduced environmental impact and site utility costs; there is also a possibility for reduction in capital costs.

88

Theoretical studies, e.g. Petlyuk et al. [4], Stupin and Lockhart [5], Fonyo et al. [6], Stichlmair and Stemmer [7], Annakou and Mizsey [8], Dunnebier and Pantelides [9], Emtir et al. [10], Kolbe and Wenzel [11] have shown that the column coupling configurations are capable of achieving typically 28-33 % of energy savings compared with the best conventional scheme. In addition, the coupling configuration can also be achieved with the so-called dividing-wall column (Kaibel 1987). By this arrangement, reduction in capital cost can be expected through the elimination of the prefractionator column shell (but not the column internals). Emtir et al. [10] compared five different energy integrated schemes, among them the forward and backward integrated prefractionator arrangement, with a non-integrated direct split sequence. The study compared the total annual costs (TAC) and the controllability of the different schemes. In terms of TAC they found that the backward integrated direct split configuration has the maximum savings of 37%. The integrated prefractionator arrangements have similar savings of 34% for the forward-integrated case and 33% for the backward-integrated case. Skogestad et al [12] had studied four pressure-staged distillation columns to see if multi-effect integration can be applied to any two columns in the sequence based on shortcut equations and Vmin-diagrams have been used for screening purposes to find the columns with the highest potential for energy savings. The results showed that when considering the existing number of stages available the ISF arrangement was the best, however when considering infinite number of stages the PF arrangement was the best.

The scope of this work This study is devoted towards separation of benzene/toluene/m-xylene (BTX) ternary mixture by continuous distillation at high product purity of 99.9 mol %. The feed composition to be separated consists of (25/50/25) mixture at atmospheric pressure and total flow rate of 100 kgmol/h. The aim of this study is to investigate the effect of changing feed conditions on the saving potential of the heat-integrated distillation schemes. The tools which are implemented on the study to investigate the feasibility of energy and TAC savings are shortcut methods based on minimum vapour flow rate at infinite number of stages using shortcut equations, pinch analysis and rigorous simulation by utilizing HYSYS simulation for rigorous modelling with the following assumptions: UNIQUAC thermodynamic model is used, pressure drop across distillation columns is taken 5 Kpa, pumping is not considered in cost calculations, maximum internal flows are at 75% -80% of the flooding, and exchange minimum approach temperature (EMAT) = 10°C. In this study the total cost for each configuration is assumed to be the sum of utility costs (steam and cooling water) and equipment costs (purchase and installation). Detailed utility cost data are extracted from Emtir et al. [13].

Studied distillation schemes Throughout this work A, B, and C denote the light, intermediate, and heavy components, respectively. The impurities are symmetrically distributed in the middle product stream. The studied distillation configuration are showing below: Figure 1; direct sequence without heat integration (DQ), direct sequence with forward heat

89

integration (DQF), direct with backward heat integration (DQB), Figure 2; indirect sequence without heat integration (IQ), indirect sequence with forward heat integration (IQF), indirect with backward heat integration (IQB).

DQ

DQF DQB

Figure 1. direct sequence with possible heat integration

IQ

IQF IQB

Figure 2. Indirect sequences with possible heat integration

Rigorous simulation and optimization The investigated schemes are simulated rigorously by HYSYS and design parameters are exported to Microsoft Excel where the final cost calculations for optimization are executed. Moreover, detailed column and heat exchanger costing are calculated using the default column and heat exchanger sizing. The sizing of distillation columns and heat transfer equipments requires the determination of flow rates, temperatures, pressures, and heat duties from the flow sheet of mass and energy balance, and these quantities can then be used to determine the capacities needed for the cost correlation. In addition, the concept of material pressure factor (MPF) is used to evaluate particular instances of equipment beyond a basic configuration. This concept is an empirical factor developed by Biegler et al [14] as part of the costing process. For each column system, the pressure, number of trays and feed location are considered as the optimization variables and they are manipulated until the optimal design is found. Optimization variables can be more in case of additional feeds, draws or recycle streams are present. In every run, design parameters (optimization variables) are changed, specifications and optimality are checked. The process simulations are stopped when the global optimal system design is achieved.

90

Shortcut analysis In this work, minimum vapour loads are estimated based on shortcut method for conventional and heat-integrated distillation schemes at two different feed conditions, saturated liquid and saturated vapor. Underwood equations are used to find out minimum reflux ratio where reflux ratio was as set 1.1 times minimum reflux ratio, Fenske equations for minimum number of stages, Gilliland equqtion for theoretical number of stages and Kirkbride equation for feed stage location. FORTRAN program is written to solve all shortcut equations where the input data are feed, top and bottom flow rates and their compositions. The effect of pressure is taken into account for calculating average relative volatility of top and bottom streams for integrated and non-integrated configurations. The minimum vapor flow rate in every separation section can be calculated using (1)

For saturated liquid feed, the vapor flow at the top is equal the vapour flow in the bottom but in saturated vapour feed vapour flow in the top is given as (2) Where q is zero for saturated vapour. The formulas listed in Table 1 have been used to determine the total minimum vapor flow rates of integrated and non-integrated distillation schemes.

Table 1. Total minimum vapor rates for integrated and non-integrated schemes

Direct and indirect sequences without heat integration, Prefractionator without heat integration, Direct with forward and backward, Prefractionator with forward and backward,

Results and discussion The shortcut results are summarized in Tables 2 & 3 for feeds at saturated liquid and saturated vapor conditions.

91

Table 2. Shortcut result of conventional and heat-integrated schemes at saturated liquid feed

SequanceColumn1 Column2 Column1 Column2 Column1 Column2 Column1 Column2

Nmin 16 18 18 15 16 20 21 18Rmin 2.13 1.24 0.816 1.91 2.13 1.38 2.944 1.24R 2.34 1.37 0.89 2.1 2.34 1.525 3.23 1.37N 40 48 48 39 40 51 52 48Feed Stage 19 24 26 19 19 26 23 24Vmin kgmol/hr 78.25 112 136.2 72.75 78.25 119 98.6 112Vmin total kgmol/hrVmin svaving % 41.13

190.25 208.95 119 112-9.830.00 37.45

IQ DQFDQ DQB

For conventional direct and indirect sequence, its clear that IQ scheme has been improved from -9.83 % to 17.81 % with respect to the base case (DQ), where as in case of heat-integrated cases the maximum savings is achieved in case of DQF with 41.13 % at saturated liquid feed. The higher savings in minimum vapor load for heat-integrated cases at satured liquid feed is attributed to the energy added inside the system are utilized effecintly by recycling the heat between the integrated columns.

Table 3: Shortcut results for conventional and heat-integrated schemes at saturated vapour

SequanceColumn1 Column2 Column1 Column2 Column1 Column2 Column1 Column2

Nmin 16 18 18 15 16 20 21 18Rmin 5.0633 1.28 1.178 1.91 5.06 1.38 6.236 1.24R 5.56 1.37 1.296 2.1 5.56 1.525 6.859 1.37N 38 48 46 39 38 51 51 48Feed Stage 18 24 26 19 18 26 22 24Vmin kgmol/hr 51.5825 114 63.35 72.75 51.5 119 80.9 112Vmin total kgmol/hrVmin svaving %

DQ IQ DQB DQF

165.5825 136.1 119 1120.00 17.81 28.13 32.32

Pinch analysis has been conducted for the heat-integrated cases by utilizing SPRINT software, results are showing in Figures 3 & 4 for case of DQB. It is obvious from pinch analysis that trim condenser is needed in case of DQB, steam temertaure level can be decided also.

92

Figure 3. Composite curve for DQB case

Rrigorous optimization results are showing similar trend in case of savings in both energy and TAC. DQB shows maximum saving values of 44.44 % in energy with 25.4 % of TAC saving. IQ, IQF and IQB integrated schemes indicates inferior savings in both of energy and TAC which is attributed to the wide gap in temperatures between heat source and heat sink and use of high pressure steam in distillation

system.

1 N:1

78.8

2 N:2

129.95

3N:3

117.79

4N:4

160.62

1 N:7

78.76

DT:-0.04DH:-810.9

1

1

N:5

N:6

129.93

119.76

*Q:846.9 A:76.0327 S:0

3 N:9

129.93

DT: -0.0044DH:-237.1

2N:8

160.66

DT:0.04DH:1175

condenser c1

condenser c2

reboiler c1

reboiler c2

Figure 4. Matching between hot and cold streams

93

Considering the effect of different feed conditions on DQB savings in energy and TAC are shown in Figure 5 and the results can be summarized as follows:

• Direct sequence with backward heat integration gives maximum saving values 44.44 % in energy, and in addition to 25.4 % of TAC saving with respect to direct sequence without heat integration.

• IQB distillation scheme are showing the highest energy saving of 51.42 % but lower values in TAC saving (not shown).

25.4

44.44

17.84

37.2

10.97

28.88

liquid statesat.liquidsat.vapor

Energy saving%

TAC saving%

Figure 5. Effect of feed conditions on DQB distillation scheme\

Conclusions The state of feed conditions plays important role on the ranking and evaluation of heat-integrated distillation schemes from heat integration point of view. DQB heat-integrated scheme is showing higher energy saving for feed entering at liquid state, this due to the recycling of heat added in the system to change the phase of the feed inside the distillation scheme, where as in case of feed entering at vapor phase the recycled heat between the columns is reduced which will effect the saving of the heat-integrated scheme directly. In some case although the saving in energy is high, but due to increasing in preussures of the integrated columns which will lead to utilization of high pressure steam causing its TAC saving to drop down. Shortcut and pinch analysis methods gives priliminary indication regarding energy saving in distillation schemes, rigourous simulation and TAC saving comparison is showing the most rigourous method of selection between heat-integrated schemes.

References [1] Hewitt G, Quarini J, Morell M. Chem. Eng, 1999:10:21. [2] Smith R., Linnhoff B. Trans. IChemE., 1988:66:195. [3] Mizsey P and Fonyo Z. Computers and Chemical Engineering, 1990:14:1303. [4] Petlyuk FB, Platonov VM and Slavinskii DM. Int. Chm. Engng., 1965:5:561. [5] Stupin WJ and Lockhart FJ. Chemical Engineering Progress, 1972:68:71. [6] Fonyo Z, Szabo J, Foldes P. Acta Chimica Academia, 1974:82:235. [7] Stichlmair J, Stemmer A. Chemical Engineering Technology, 1989:12:163. [8] Annakou O, Mizsey P. Ind. Engng. Chem., 1996:35:1877.

94

[9] Dunnebier G, Pantelides CC. Ind. Eng. Chem. Res., 1999:38:162. [10] Emtir M, Mizsey P, Rev E, Fonyo Z. Chem. Biochem. Eng. Q., 2003:17(1):31-42. [11] Kolbe B, Wenzel S. Chemical Engineering and Processing, 2004:43:339. [12] Engelien HK, Skogestad S. Chemical Engineering and Processing 2005:44:819–826 [13] Emtir M, Rev E, Fonyo Z. Applied Therm. Eng., 2001:21:1299-1317. [14] Biegler LT, Grossmann IE, Westerberg AW. Eds., Systematic Methods for Chemical Process Design, Prentice-Hall, NJ. 1997.


95

Using Captured Carbon Dioxide for Enhancing Growth of Algae Ponds

Attilio Converti, Patrizia Perego, Luigi Maga, Vincenzo Dovì

Dipartimento di Ingegneria Chimica e di Processo“G.B.Bonino”,Università di Genova, Via Opera Pia 15, 16145 Genova, Italy,

email:[email protected]; Tel.: +39-010-3532593;.

Abstract The amount of carbon dioxide that could be fed to algae ponds is bound to increase considerably if and when carbon dioxide capture technologies are fully developed, making its capture from large power station becomes a feasible option. The most widespread option for its storage is to inject it into deep geological strata. This will require its transport by pipelines or in containers for short distances. These are both costly options. Thus, any reduction of the amount of carbon dioxide to be transported can play a significant role in the overall financial balance. To this purpose one of the most promising technologies is the use of concentrated carbon dioxide for algae farming. In addition to reducing the CO2 flow-rate, algae can be harvested for further use in the production of biodiesel. The economics is strongly affected by a number of parameters that have to be optimised: type of algae selected, reaction environment, time of harvesting, light intensity, temperature, saturation and inhibition effects.

Introduction Selecting the right type of alga, the reaction environment (reactors or open ponds), the operating conditions (if reactors are employed) and optimising the carbon dioxide flowrate are the sequential steps to be examined in the overall economic optimisation procedure. The general physical, chemical and biological properties of each strain can be investigated in a laboratory environment, the basic equipment being shown in Fig 1. The resulting information can be utilised in open ponds, as well as in industrial reactors.

Air

MagneticStirrer

CO2

Fig. 1. Basic equipment required

Light

96

An integrated approach STEPS 1-6 ARE REPEATED FOR EACH ALGA STRAIN CONSIDERED

1. Automated data acquisition and parameter estimation of bioreactor fed by thecarbon dioxide

2. Computation of biodiesel yield from algae harvest

3. Evaluation of optimal CO2 absorption rate by algae

4. Optimisation of the reaction environment (open ponds or bioreactors)

5. Design and optimisation of the CO2 capture process

6. Evaluation of the overall economics

97

The goal of the project is precisely the identification of the most effective alga strain and the optimal process configuration for the overall minimisation of the costs of carbon dioxide sequestration thanks to absorption from algae. It consists of an automated laboratory reactor system interfaced to a sequence of software tools, which provide for every strain the corresponding optimal configuration for cost minimisation. A graphical presentation of a simplified diagram of the structure showing the logical flow of its components is provided in Fig 2.

Application to Spirulina Platensis The cyanobacterium Spirulina platensis has already proved to be a valuable algal strain for the production of biomass, using the CO2 generated from the industrial processes as carbon source (Binaghi et al. [1]). While light is undoubtedly the most important factor influencing its growth, the range of light intensity which maximises it (an optimum irradiance in the range 30–50 Wm-2 has been suggested by Chojnacka and Noworyta, [2]) as well the range of photoinhibition have already been examined and documented in the literature (Zeng and Vonshak, [3]). Similarly, optimal temperature conditions had already been identified and reported in the literature (Therefore the experimental investigation was carried out in a bench-scale helical photoreactor in the optimal irradiance interval under different conditions of CO2 feeding rate, which was the main object of the experimentation. Nowadays, the commercial production of S. platensis is mainly performed in open ponds (Tredici et al., [4]), which are cheap and easy to operate, as they use the solar irradiance as a free source of energy. However, they do not allow reaching high biomass productivity owing to the difficulty to keep the optimum temperature, thus they are restricted to the tropical and sub-tropical regions (Jiménez et al., [5]). The use of a closed bioreactor like tubular reactors (Borowitzka, [6]) can partially avoid these problems (Torzillo, [7]): a controlled environment does in fact allow controlling temperature, reducing risks of contamination and increasing the surface hit by the light radiation. Previous studies, developed in the near past, showed that the factors that most influence the growth in a tubular reactor are the size of the tubes, the mixing and the circulation of the culture (Molina et al., [8]) that in turn depend on the reactor configuration. The strain Spirulina platensis UTEX 1926, classified as Arthrospira (Spirulina) platensis (Nordstedt) Gomont (Silva et al., [9]), was obtained from the University of Texas Culture Collection. The cells were maintained in the culture medium of Schlösser [10], whereas fed-batch cultivations were performed in this medium lacking NaHCO3 and Na2CO3, the necessay carbon source being supplied by bubbling carbon dioxide. All the cultivations were carried out using a working volume of 4l and an initial cell concentration of 0.4 g l-1 by dry weight (DW) (Soletto et al., [11]). The starting pH was initially adjusted to 9.5 ± 0.2 by means of NaOH 6 M solution, daily controlled and, if necessary, increased to 9.5 by addition of the same solution. The continuous light intensity was regulated by means of a variable number of 40 W-fluorescent

98

lamps. Temperature was maintained at the optimum value for this biosystem (30°C) (Danesi et al., [12]; Sánchez-Luna et al., [13]) in a thermostated chamber. CO2 additions were daily performed by bubbling the selected amount of pure gas directly into the medium from a cylinder with a flowmeter. In the photobioreactor of Fig 1 circulation was granted by an airlift mechanism. To avoid any atmospheric CO2 accumulation in the photobioreactor due to the air used in the airlift device, the ambient air was firstly passed through two 4 l-closed vessels containing 2 M NaOH and distilled water, respectively, and then through a drier containing CaCl2 before entering the system. The NaOH solution, distilled water and CaCl2 were renewed weekly. The glass helical cylinder had a diameter of 16 cm and was 48 cm long, the internal diameter of the photostage’s tube being 1.5 cm, while the PVC tube had the same diameter and was 1.0 m long. The overall starting volume was 4 l, while the photostage volume was about 0.6 l. The airlift pump, model AC-1000 (Resun, Shenzhen, China), granted a flow of 1.0 l min-1, so ensuring a recycling residence time of 4 min. The following parameters were determined during the experiments: pH, temperature and concentration of biomass. Dry cell mass concentration was determined by optical density (OD) measurements at 560 nm (LeDuy and Therien, [14]) using a calibration curve and a Spectronic 21 UV-VIS Spectrophotometer (Milton Roy Company, Rochester, NY). Temperature and pH of the cultures were measured with a pH sensor Hanna 211. Light intensity was measured with an illuminance meter, model TL-1 (Minolta, Osaka, Japan), as photosynthetic photon flux density (PPFD), expressed as μmol photons m-2 s-1. Measurements were performed at different points of the photobioreactor to ensure average illumination intensity. To convert PPFD to potosyntetically active radiation (PAR), expressed as kJ m-2 d-1, the conversion factor 18.78 kJ s d-1 for cool white fluorescent lamps was used (Hall and Scurlock, [15]). The input of PAR (IPAR) into the reactor was calculated by multiplying PAR to the illuminated surface (m2) (Watanabe and Hall, [16]). The photosynthetic efficiency (PE) (%) was calculated as:

100⋅=IPAR

HrPE GG (1)

where rG is the maximum daily biomass growth (gDW d-1) and HG the enthalpy of dry biomass (21.01 kJ gDW-1) (Watanabe and Hall, [17]). Assuming for S. platensis dry biomass the elemental composition reported by Cornet et al. [18] (CH1.650O0.531 N0.120S0.007P0.006), i.e. a carbon content of dry biomass of 48.2%, the carbon utilization efficiency (CUE) (%) was calculated according to the equation:

100⋅=F

DW

CCCUE (2)

where CDW is the dry biomass carbon content (g) and CF the carbon supplied with CO2 addition (g).

99

Experiments were designed using a sequential planning approach [19] after a first set of fed-batch pulse-feeding runs (A) was carried out by keeping constant the CO2 feeding rate (F = 0.44 g d-1) and varying the photosynthetic photon flux intensity (PPFD) from 80 to 210 μmol photons m-2 s-1 (Table 1). The main advantages of such a bioreactor configuration are the low stress conditions, because of the low relative velocities between the liquid and the bubbles in current flow, the low energy requirements and the simple design (Sajc et al., [20]; Sánchez et al., [21]).

Table 1- Experimental schedule followed for S. platensis cultivations, performed at variable photosynthetic photon flux density (PPFD) and CO2 feeding rate (F)

Test A1 A2 A3 A4 B1 B2 B3 C1 C2 C3

PPFD (μmol

photons m-2s-1) 80 125 166 210 166 210 250 166 210 250

F (g d-1) 0.44 0.44 0.44 0.44 0.74 0.74 0.74 1.03 1.03 1.03

0

5

10

15

20

25

0 10 20 30 40 50 60

Time (d)

Bio

mas

s co

ncen

trat

ion

(g l-1

)

0

20

40

60

80

100

120

CU

E (%

)

Fig. 2. Biomass concentration (closed symbols) and carbon utilization efficiency (CUE) (open symbols) during fed-batch pulse-feeding cultivations of S. platensis

carried out in the helical photobioreactor. F = 0.44 g d-1. Photosynthetic photon flux density (μmol photons m-2 s-1): 80 ( , ); 125 ( , ); 166 ( , ); 210

The results in Fig 2 show that, a progressive increase in PPFD from 80 to 125 and then to 166 μmol photons m-2 s-1 led to an increase in maximum biomass concentration (Xmax) after 50 days from 2.90, 3.84 and 12.8 g l-1. This maximum value was much higher than those previously obtained in raceway ponds (Xmax = 0.8 g l-1) (Materassi et al., [22]), elevated panels (Xmax = 5.0 g l-1) (Tredici et al., [4]), and two-plane tubular reactor (Xmax = 4.2 g l-1) (Torzillo et al., [23]). In a recent study, optimization of PPFD and minimization of mechanical stress in a horizontal tubular air-lift reactor allowed achieving Xmax = 10.6 g l-1 after only 15-19 days (Converti et al., [24]). These results suggest that the longer recycling residence time in the helical configuration, although delaying the achievement of stationary conditions with respect to the horizontal one, ensured higher pseudo-steady state biomass level.

100

An increased PPFD also improved cell productivity: after 15 days the cell productivity at 80, 125 and 166 μmol photons m-2 s-1 was 0.12, 0.14 and 0.20 g l-1d-1, respectively. A further increase in PPFD did not lead to any further productivity enhancement, as a likely result of variations in the carbon utilization efficiency (CUE). At PPFD = 80 and 125 μmol photons m-2 s-1, CUE reached after 21 days maximum values of 53 and 74%, respectively; after 50 days it decreased, likewise biomass concentration, to 21 and 29%, respectively, hence highlighting a photolimitation effect. At higher PPFD (166 μmol photons m-2 s-1), CUE was 96% after 21 days and even exceeded its theoretical value after 47 days. This result would be consistent with the complete consumption of the carbon source, and, what is more, with a consequent reduction of S. platensis C content (Lacaz-Ruiz, [25]). This change, induced by the lack of carbon, could have caused biomass to stop growing, as it was evident during the last 10 days of the cultivation. The above CUE values are substantially higher than the maximum ones reported either for batch (40%) or fed-batch (20%) S. platensis cultivations in Erlenmeyer flasks and demonstrate the good performance of the proposed photobioreactor configuration. A further increase in PPFD to 210 μmol photons m-2 s-1 confirmed the above trend: after 24 days biomass grew similarly to PPFD = 166 μmol photons m-2 s-1, thus suggesting that the higher light intensity accelerated the effect of carbon shortage by completely stopping cell growth. Resuming, both PPFD = 80 and 125 μmol photons m-2 s-1 led to photolimited cultures, while PPFD = 166 and 210 μmol photons m-2 s-1 to a growth limitation owing to carbon source starvation.

02

4

6

8

10

1214

16

18

20

0 10 20 30 40 50 60

Time (d)

Bio

mas

s co

ncen

trat

ion

(g l-1

)

0

10

20

30

40

50

60

70

CU

E (%

)


carried out in the helical photobioreactor. F = 0.74 g d-1. Photosynthetic photon flux density (μmol photons m-2 s-1): 166 ( , ); 210 ( , ); 250 ( , )

In order to prevent carbon limited culture, a new set of tests (B) was performed using higher CO2 feeding rate (F = 0.74 g d-1) (Fig 3). As expected, CUE at PPFD = 166 μmol photons m-2 s-1 was lower (never reached 50%), and the cell concentration always exhibited lower values than at the previous feeding rate. Under these conditions, the cultivation ceased to be carbon limited and became light limited; therefore, PPFD was raised to accelerate the metabolic activities of S. platensis as well as to better CO2 fixation. At PPFD = 210 μmol photons m-2 s-1, biomass

101

concentration became higher than that at 166 μmol photons m-2 s-1 after 13 days and achieved a maximum value of 9.22 g l-1after 54 days. However, a further increase to 250 μmol photons m-2 s-1 resulted in the worst cell growth (Xmax = 9.03 g l-1 after 58 days), and CUE was lower than at PPFD = 210 μmol photons m-2 s-1, hence highlighting a clear photoinhibitory effect. On the basis of these results, focusing the attention only on Xmax and the respective CUE, the test A3 performed at F = 0.44 g d-1 and PPFD = 166 μmol photons m-2 s-1 was undoubtedly the best one, providing Xmax = 12.8 g l-1 and totally removing CO2; on the other hand, larger amounts of the carbon source resulted in lower values of these parameters, as a consequence of excess salinity occurrence owing to NaOH addition for pH adjustment. In addition, as far as the cultures at PPFD = 250 μmol photons m-2 s-1 are concerned, photoinhibition certainly added to this effect. Although the cultivation A3 (F = 0.44 g d-1 and PPFD = 166 μmol photons m-2 s-1) seemed to be the most suitable for a long-term cultivation, ensuring the highest biomass level, it is worth of note that, for a short-term cultivation, C3 (F = 1.03 g d-1 and PPFD = 250 μmol photons m-2 s-1) would be better. Table 2 - Main results of S. platensis cultivations performed at different CO2 feeding

rate and light intensities

Test Fa PPFDb IPARc Xmaxd μmax

e CUEmaxf PEg

(g d-1) (μmol m-2 s-1) (kJ d-1) (g l-1) (d-1) (%) (%)

A1 0.44 80 298 3.25 0.22 53 6.20

A2 0.44 125 466 4.26 0.37 74 9.37

A3 0.44 166 619 12.8 0.20 108 7.47

A4 0.44 210 783 5.76 0.20 96 5.04

B1 0.74 166 619 5.64 0.41 61 6.65

B2 0.74 210 783 9.22 0.26 47 5.79

B3 0.74 250 932 9.03 0.26 45 4.42

C1 1.03 166 619 1.60 0.19 25 5.37

C2 1.03 210 783 3.93 0.51 40 4.92

C3 1.03 250 932 2.72 0.59 56 4.69 aF = CO2 feeding rate bPPFD = Photosynthetic photon flux density. cIPAR = Input of photosynthetic active radiation. dXmax = Maximum cell concentration.

emax= Maximum specific growth rate. fCUEmax = Maximum carbon utilization efficiency. gPE = Photosynthetic efficiency.

Light intensity is another crucial factor for the performance of a tubular photobioreactor, and, combined with CO2 feeding rate, strongly influenced biomass growth. The results of tests performed at different PPFD, summarized in Table 2, suggest that a higher light intensity increased the photosynthetic activities of the cell,

102

but it must be supported by a suitable CO2 feeding rate to provide satisfactory results. In the last set of tests (C), the CO2 feeding rate was raised again, from 0.74 to 1.03 g d-1, to ensure a larger amount of carbon in the attempt to let the biomass stand a higher PPFD. However, giving a look at the curves of Fig 4, it is possible to notice the impossibility to perform a long-term cultivation with such a feeding rate. According to the observations of Binaghi et al. [1], the presence of excess inorganic carbon in the medium was in fact responsible for poor growth. Nevertheless, at PPFD = 210 and 250 μmol photons m-2 s-1, after 10 days, the productivity was the highest (0.23 g l-1 d-1), but, later on, due to carbon quick accumulation in the medium, biomass stopped growing and its concentration began to decrease. At the beginning of all the cultivations performed at high PPFD, biomass was able to grow very quickly without any photoinhibition, if supplied with sufficient carbon source.

0

3

6

9

12

15

0 5 10 15 20 25

Time (d)

Bio

mas

s co

ncen

tratio

n (g

l-

1 )

0

10

20

30

40

50

60

CUE

(%)


carried out in the helical photobioreactor. F = 1.03 g d-1. Photosynthetic photon flux density (μmol photons m-2 s-1): 166 ( , ); 210 ( , ); 250 ( , )

The type of experimentation used cannot be completely generalized to all algae strains, but can provide enough guidance for a partial automation of experimental campaigns for other specimens. The following steps (process and overall economic optimisation) can be carried out sequentially using suitable simulation tools and can provide, if the necessary parameters are available, the optimal choice of the alga strain, of the reaction environment and of the generated CO2 flowrate that can be used to feed algae optimally.

Conclusions The procedure developed has to be applied to a large number of algae through a trial-and-error procedure, since the complex relations among light, CO2 takeup, biomass growth and lipid content is presently not reliably predictable from the morphogenic properties of the alga strain alone. Eventually GM algae will have to be considered to maximise carbon dioxide uptake and harvest yield.

103

Economic data are volatile. Discarded strains should be reconsidered as financial values change. The way ahead for the project is an automated link among data collection and the various pieces of software (including simplified models for preliminary analyses), as well as the setup of a structured data base for the retrieval and update of available data for further elaboration.

References [1] Binaghi L, Del Borghi A , Lodi A , Converti A , Del Borghi M. Batch and fed-batch uptake of carbon dioxide by Spirulina platensis. Proc. Biochem. 2003; 38: 1341-1346. [2] Chojnacka K, Noworyta A. Evaluation of Spirulina sp. growth in photoautotrophic, heterotrophic and mixotrophic cultures. Enzyme Microbial Technol. 2004; 34: 461-465. [3] Zeng M-T, Vonshak A. Adaptation of Spirulina platensis to salinity-stress Comp Biochem Phys A, 1998;120: 113-118 [4] Tredici M, Chini Zitelli G, Biagiolini S, Materassi R. Novel photobioreactors for the mass cultivation of Spirulina spp., Bull. Inst. Oceanogr., 1993;12: 89–96. [5] Jiménez C, Cossío B, Niell F. Relationship between physicochemical variables and productivity in open ponds for the production of Spirulina: a predictive model of algal yield, Aquaculture, 2003; 221:331-345. [6] Borowitzka M. Commercial production of microalgae: ponds, tanks, tubes and fermenters. J. Biotechnol., 1999; 70: 313-321. [7] Torzillo G. Tubular bioreactors. In: Vonshak A, editor. Spirulina platensis (Arthrospira): Physiology, Cell-Biology and Biotechnology. Taylor & Francis, London, 1997: 101-115. [8] Molina E, Fernández J, Acién F, Chisti Y. Tubular photobioreactor design for algal cultures. J. Biotechnol. , 2001; 92: 113-131. [9] Silva PC, Basson PW, Moe RL. Catalogue of the Benthic Marine Algae of the Indian Ocean, University of California Publications in Botany, Berkeley, CA, 1996; 79. [10] Schlösser UG. Sammlung von Algenkulturen, Ber. Deutsch. Bot. Ges. 1982; 95:181-276. [11] Soletto D, Binaghi L, Lodi A, Carvalho JCM, Converti A. Batch and fed-batch cultivations of Spirulina platensis using ammonium sulphate and urea as nitrogen sources, Aquaculture, 2005; 243: 217-224. [12] Danesi EDG, Rangel-Yagui CO, Carvalho JCM, Sato S. An investigation of the effect of replacing nitrate by urea in the growth and production of chlorophyll by Spirulina platensis, Biomass Bioenergy, 2002; 23: 261-269 [13] Sánchez-Luna LD, Converti A, Tonini GC, Sato S, Carvalho JCM. Continuous and pulse feedings of urea as a nitrogen source in fed-batch cultivation of Spirulina platensis, Aquacult. Eng., 2004; 31: 237-245 [14] LeDuy A, Therien N. An improved method for optical density measurement of the semimicroscopic blue algae Spirulina maxima. Biotechnol. Bioeng., 1977; 19:1219-1224. [15] Hall DO, Scurlock JMO. Biomass production and data. In: Hall DO, Scurlock JMO, Bolhar-Nordenkampf HR, Leegood RC , Long SP, editors, Photosynthesis and Production in a Changing Environment. Chapman & Hall, London, 1993; 425-444.

104

[16] Watanabe Y, Hall DO, Photosynthetic CO2 fixation technologies using a helical tubular bioreactor incorporating the filamentous cyanobacterium Spirulina platensis. Energy Convers. Manag. 1995; 36: 721-724. [17] Watanabe Y, Hall DO. Photosynthetic CO2 conversion technologies using a photobiobioreactor incorporating microalgae. Energy and material balances. Energy Convers. Manag. 1996; 37: 1321-1326. [18] Cornet JF, Dussap CG, Cluzel P, Dubertret G. A structured model for simulation of cultures of the cyanobacterium Spirulina platensis in photobioreactors: II. Identification of kinetic parameters under light and mineral limitations. Biotechnol. Bioeng. 1992; 40: 826-834. [19] Bard Y. Nonlinear Parameter Estimation, Academic Press, 1974 [20] Sajc L, Grubisic D, Vunjak-Novakovic G. Bioreactors for plant engineering: an outlook for further research. Biochem. Eng. J. 2000; 4: 89–99. [21] Sánchez MJ, Jiménez-Aparicio A, López GG, Tapia GT, Rodríguez-Monroy M. Broth rheology of Beta vulgaris cultures growing in an air lift bioreactor, Biochem. Eng. J. , 2002.;12: 37-41. [22] Materassi R, Balloni W, Pushparaj B, Pelosi E, Sili C. Coltura massiva di Spirulina platensis in sistemi colturali aperti. In: R. Materassi (Ed.), Prospettive della Coltura di Spirulina in Italia, Consiglio Nazionale delle Ricerche, Rome, 1980:241. [23] Torzillo, G., Carlozzi, P., Pushparaj, B., Montaini, E., Materassi, R. A two-plane tubular photobioreactor for outdoor culture of Spirulina. Biotechnol. Bioeng. 1993:42:891-898. [24] Converti A, Lodi A, Solisio C, Del Borghi M. Cultivation of Spirulina platensis in a combined airlift-tubular reactor system. Biochem. Eng. J., under revision 2006. [25] Lacaz-Ruiz R. Espirulina: Estudos e Trabalhos. ROCA, São Paulo, Brazil 2003.


105

Energy and resource saving at operating plants based on the analysis of historical process data

László Dobos, Balázs Balasko, Sándor Németh and János Abonyi* University of Pannonia, Egyetem u.10, 8200 Veszprém, Hungary,

email: [email protected]; fax: +36 88 624171

Abstract Chemical industry is a market-driven playground where continuous process improvement is needed to challenge competitors. As these developments get implemented and chemical processes get more and more automated, lots of historical process data gets available, which can be used not only for monitoring of the control performance, but also for the monitoring of plant efficiency. Unfortunately, till nowadays small portion of data is analyzed and energy-efficiency improvement of the operating processes is an unnoticed research area. The paper briefly reviews recent advances on historical process data applications in plant performance optimisation, and as a main contribution applies exploratory data analysis techniques for monitoring energy and resource consumption related to the operation of a polymerization technology.

Introduction Continuous development of chemical processes is an inevitable must for companies that want to reach competitive advantage or - due to the high competition level - stay in business. Process improvement is a widely known abstraction, a common part of sustainable development studies, a valuable procedure among system optimization and system renewal, which means improving the efficiency and effectiveness, i.e. the performance of the process in some manner that through revision, retrofit design and reorganization eco-efficiency is improved [1]. As such, it has a strong connection with optimization of operating processes, and the corresponding engineering tasks as well: process monitoring and fault detection [2,3,4], process identification for control [5,6] or simulation [7,8]. These tasks are common tasks in chemical engineering practice with the same aim: minimizing waste or losses [9], maximizing throughput [10] and improving quality of products [11] while operating in a safe and efficient way. As chemical industry is a highly automated industry, lots of historical process data is available, which carries underlying information about the process thus has the potential to guide the above mentioned tasks if this information can be extracted. Solution for this purpose can be the procedure of Knowledge Discovery in Databases (KDD) or Exploratory Data Analysis (EDA) [12,13]. Based on a constructed database of the system, the iterative process of KDD interactively takes the user along the path from data preparation and data mining to extracted knowledge. EDA techniques can be considered as a special case of KDD where the visual understanding has the major importance and thus it is widely applied for monitoring of the process [14,15] or control [16].

106

It is obvious from the above applications that historical process data is applicable for monitoring of control performance, but it can be also useful for the monitoring of plant efficiency improvement. E.g. by identification and comparison of different time-segments, knowledge of best practice can be extracted in terms of the detection of the optimal operating regimes. In [17], this type of process improvement is applied as a methodology for a hot stove system; dominant variables are selected by principle component analysis (PCA) technique from the collected data and operating regimes are detected by hierarchical clustering. The weakness of their methodology stands in the lack of possibilities for visually follow up the technology, but in these approaches visualization plays a large role in the knowledge discovery process: not only PCA, quantile-quantile and box plots can be applied and combined with e.g. energy consumption related data to detect the bottlenecks of the technology [18]. Moreover, historical process data provides a more detailed understanding to the process. This ability was realized by many researchers who applied historical data for plant performance improvement, e.g. Puigjaner et al. applied historical process data of a batch plant to build an artificial neural network for fault detection and with the help of HAZOP analysis to generate an if-then rule base for optimal operation [19]. Predictive models can be identified to detect the hidden relations among the process variables, like in [20] historical data reconciliation and PCA is combined to improve plant efficiency of an ammonia synthesis by early fault detection. Numerous regression and classification models have been used to build soft sensors giving useful measures related to plant efficiency. Software (or soft) sensors are estimators of non- or heavily measurable process variables derived from measurable variables of the system [21]. With such tools the optimal operating regimes are not detected, they are designed with the use of the combination of first principles and statistical models. A systematic approach for soft sensor design is presented in [22], which contributes outlier detection into PCA for dynamic partial least squares regression and tests this approach on free lime and NOx emission of a cement kiln process. Choi and Park applied hybrid neural network to build an online waste water quality estimator where PCA decreased the data noise sensibility of the model by decreasing the number of variables [23]. The contribution of this paper is that it deals with an integrated approach of using historical process data to build soft sensors for estimating plant efficiency and to visualize performance level, which task is unfortunately set aside in the above mentioned works for plant and process optimization based on historical process data. The paper is organized as follows: first, a short description of the applied EDA techniques is presented then their applicability is shown for economical analysis of polymerization reactors in a Hungarian polymerization plant. Finally conclusions for efficiency improvement for plant operation in given.

Statistical process data analysis using Exploratory Data Analysis method

Box plots of technological variables Process data collected during the operation of complex production processes can be used for system identification, process monitoring and optimization. To analyze the

107

enormous amount of information in process data warehouse, statistical data analysis tools are needed: trend diagrams, box plots, quantile-quantile plots, etc [13]. Suppose that X is a process variable mesasured whie the plant was operaing. In our work, the analysis of the process variables is considered to invetigate the relationshop between the energy consumption, energy cost and the measurable process variables, to become able to develepoa soft sensor. Hence, the variables are X ∈ z k , yk . The (cumulative) distribution function of X is the function F given by F(x) = P(X ≤ x) for x, which is a function giving the probability that the random variable X is less than or equal to x, for every x. For a discrete variable, the cumulative distribution function is found by summing up the probabilities. For a continuous random variable, the cumulative distribution function is the integral of its probability density function. Suppose that p∈ [0, 1]. A value of x such that F(x−) = P(X < x) ≤ p and F(x) = P(X ≤ x) ≥ p is called a quantile of order p for the distribution. Roughly speaking, a quantile of order p is a value where the cumulative distribution crosses p. Hence, by a quantile, we mean the fraction (or percent) of points below the given value. That is, the 0.25 (or 25%) quantile is the point at which 25% percent of the data fall below and 75% fall above that value. Note that there is an inverse relation of sorts between the quantiles and the cumulative distribution values. A quantile of order 1/2 is called a median of the distribution. When there is only one median, it is frequently used as a measure of the center of the distribution. A quantile of order 1/4 is called a first quartile and the quantile of order 3/4 is called a third quartile. A median is a second quartile. Assuming uniqueness, let q0.25, q0.5, and q0.75 denote the first (lower), second, and third (upper) quartiles of X. Note that the interval from q0.25 to q0.75 gives the middle half of the distribution, and thus the interquartile range is defined to be IQR = q0.75 − q0.25, and is sometimes used as a measure of the spread of the distribution with respect to the median. Let q0 and q1 denote the minimum and maximum values of X, respectively (assuming that these are finite). The five parameters q0, q0.25, q0.5, q0.75, q1 are often referred to as the five−number summary. Together, these parameters give a great deal of information about the distribution in terms of the center, spread, and skewness. Tukey’s five number summary is often displayed as a boxplot. Box plots are an excellent tool for conveying location and variation information in data sets, particularly for detecting and illustrating location and variation changes between different groups of data [10]. A box plot consists of a line extending from the minimum value q0 to the maximum value q1, with a rectangular box from q0.25 to q0.75, and tick marks at the median q0.5. Hence, the lower and upper lines of the "box" are the 25th and 75th percentiles of the sample. The distance between the top and bottom of the box is the interquartile range. The line in the middle of the box is the sample median. If the median is not centered in the box that is an indication of skewness. Thus the box represents the body (middle 50%) of the data. A single box plot can be drawn for one batch of data with no distinct groups. Alternatively, multiple box plots can be drawn together to compare multiple data sets or to compare groups in a single data set.

Quantile-quantile plots The quantile-quantile plots compose the other maon part of the Exploratory Data Analysis method. A q-q plot is a plot of the quantiles of the first data set against the

108

quantiles of the second data set. Both axes are in units of their respective data sets. That is, the actual quantile level is not plotted. For a given point on the q-q plot, we know that the quantile level is the same for both points, but not what that quantile level actually is. If the data sets have the same size, the q-q plot is essentially a plot of sorted data set A against sorted data set B. If the data sets are not of equal size, the quantiles are usually picked to correspond to the sorted values from the smaller data set and then the quantiles for the larger data set are interpolated. A diagonal reference line is also plotted. If the two sets come from a population with the same distribution, the points should fall approximately along this reference line. The greater the departure from this reference line, the greater the evidence for the conclusion that the two data sets have come from populations with different distributions. If the two data sets come from populations whose distributions differ only by a shift in location, the points should lie along a straight line that is displaced either up or down from the diagonal reference line. The q-q plot is similar to a probability plot, where the quantiles for one of the data samples are replaced with the quantiles of a theoretical distribution. The q-q plot can be used to answer the following questions: Do two data sets come from populations with a common distribution? Do two data sets have common location and scale? Do two data sets have similar distributional shapes? Do two data sets have similar tail behavior? It is very important to explore wheter connections between process variables exist, because the first step in soft sensor developing is detecting the connections between the studied variables. The graphical exploring methods such as quantile-quantile plots can make these detection easy and this way a huge amount of data sets can be compared to each other. When the existing of relationships between the variables are found we are able to elaborate these connection through different kind of models. After that developing soft sensors are possible, beacuse the necessary sets of data are given and the knowledge of connections are accessible. In this paper connection between operating costs, energy consumption and process variables of an operating polymer process are investigated to be able to find the bottleneck of the technology, and this way it is possible to enhance the economical efficiency.

Exploratory Data Analysis for Resource and Energy Saving

Polypropylene polymerization technology The PolyPropylene-4 plant (PP4) of TVK Ltd. (www.tvk.hu) is based on the licensed Spheripol® Process (see the scheme on Figure 1.), which produces spherical polypropylene polymer particles. Two kind of polymer products are produced: homopolymer (propylene polymer) and copolymer (propylene-ethylene polymer). A three-component-catalyst (Ti-catalyst on MgCl2 base - CAT-, Aluminium-triethyl – TEAL - and a Silane compound – Donor) is used for polymerization. In this catalysts system it is very important to maintain Al/Ti and Donor/Ti ratios within the preset range because of the product quality and productivity. The catalyst components are mixed in pre contacting pot it is injected into a specific portion of the propylene stream in the pre-polymerization reactor.

109

Figure 1. Scheme of Spheripol® polypropylene technology

After pre-polymerization, the blend is incorporated into the circulated slurry of the first loop reactor. The polymerization reaction for homopolymer production takes place in two loop reactors in series. All the loop reactors operate completely full, a combination of an expansion drum and an evaporator is exploited to keep them so and prevent any pressure fluctuation. The hydrogen feed is used to control the intrinsic viscosity of the polymer acting on the length of the polymer chains. To maximize productivity and catalyst yield, residence time and reactor slurry density are operated at the maximum. The large quantity of monomers (ca. 50 wt%) in the outlet of the second loop reactor must be recovered. After a complete evaporation and flash separation, the condensed polymer particles are and discharged to the gas-phase reactor, if impact or specially impact copolymer is produced from homopolymer. Otherwise the loop reactor outlet is transferred to the purification and extrusion section and into the storage silos. The recovered monomers are fed back to the propylene feed tank after they are purified from the powder entertainment and aluminium-alkyl. High impact copolymers are produced in heterophased, fluid-bed reactor, but our measured data sets are from homopolymer productions, so detailing the copolymer production process is not necessary in this paper. A detailed process analysis can be answered many technological questions, discovered relevant information and relationships of process variables at the important “events” of the process, as product change, catalyst change, catalyst productivity, product quality, etc. The discovered knowledge can be used to optimize the production process, improve the productivity and product quality, and reduce the operating (energy and raw materials) costs to stay compatible or enhance the profit of the process.

Pre-contacting

Pot

1st Loop Reactor

Prepoly-merization

Reactor

2nd Loop Reactor

Evaporator & Expansion drumCWS CWR CWS CWR

Gas Phase Reactor

Propylene Feed Tank Dynamic

Separator

Extrusion & Storage

Recirc. Compressor

Unit Recycle Gas Filter

Scrubber & Stripper

Unit

Steamer & Dryer Unit Steam, N2

Propylene

Donor TEAL CAT

Hydrogen

Ethylene

110

Process Datawarehouse to handle Historical Process Data In the plant the safety operation is assured by a distributed control system (DCS). The DCS collects the most important measured process variables and calculates different kinds of special process variables, which can provide more information about the process. Most DCS have functions also to store historical process data. These data definitely have the potential to provide information for product and process design, monitoring and control. Unfortunately data stored by the DCS has limited accessible in time on the process control computers, they are archived retrospective about two or three months. If there is a need to store technology data more length of time interval, it is expedient to store data in a separate, analysis driven database, in a so-called process data warehouse (DW). A process data warehouse is a data analysis-decision support and information process unit, which is operated separately from the databases of the DCS. It is an information environment in contrast the data transfer-oriented environment, which contains trusted, processed and collected data in order to historic analyzes, so it makes possible to get relevant information from the technology. In this study process data were collected through the Honeywell PHD (Process History Database) module of the control system. To store data for historical analysis, a process data warehouse designed and implemented in MySQL database system has been applied. Historical process data related to twelve periods of the productions were selected to be analyzed. In these productions four different kind of products were produced. These grades were selected to be comparable to each other.

Using Exploratory Data Analysis for Resource and Energy Saving This study concentrates on the reactors and examines the energy consumption of the serving units of the reactors such as pumps and the cooling system. The energy consumption of circulating pumps are measured during the process, the energy consumption of the raw material pumps and the cooling system have been calculated based on historical process data. When these data sets are available, it is possible to calculate the operational cost of these units. It is useful to compare these energy costs at different operations of a given technology, e.g. while producing different kind of grades, or at different periods of the operations. In Figure 2 the energy cost can be seen while producing same grade (the periods of the productions are denoted as gy3,gy4,gy5,gy6,gy7). Although some similarity can be allocated differences can also be recognized most. These differences highlight that it is necessary to find process variables which can help to estimate the energy consumptions, and find the sources of differences, and investigate what kind of operating variables really have influence on energy consumptions. The slurry density in the reactors seems to be appropriate to this demand, because it may have an effect on the energy demand of the circulating pumps, and it can also have connection to the rate of the production.

111

Figure 2. Boxplots of energy expenses producing the same grade in different

productions (The scales are normalized due to the request of the industrial partner).

Figure 3. Comparison of the distributions of the slurry density

during different productions As Figure 3 shows that the distribution of the slurry density has different type of distributions during different period of operations. Hence, this variable could be useful to explain why the energy costs of different productions varies. The next step of the investigation is to find the right relationship between the correct, and needed process variables, such as raw material inlet flow and slurry density. This is very important, because if it is possible to find a right process variable that has influence on the energy consumption, it can be used to estimate and reduce the operational cost. As can be seen on Figure 4, at the first glance it is important to analyze the relationship between the slurry density and the operational cost.

112

Figure 4. Analysis of the effect of the slurry density to the distributions of the energy

costs of the cooling system and circulating pump energy consumption. Analyzing these figures the relationship between the determined variables can be found. This means if the distributions of some variables are similar, the source of their variations could be common. This also means the possibility of the estimation.

Figure 5. Comparison of the distributions of slurry density and propylene inlet flow

The third key element of the energy consumption is the energy demand of the raw material inlet pumps. Since propylene is not just used as a reagent, but also serves as solvent, it's inlet flowrate is frequently used to control the slurry density. So correlation between the slurry density and the propylene inlet is assumed, which can be very useful in this case, because the slurry density can be the common process variable what the energy cost depends on. This connection can be seen on Figure 5. On these plots it is proven that the most important variable that relates to the energy consumption is the slurry density, so the whole energy demand, and energy cost can be estimated from this important, frequently measured process variable. Proving this statement the quantile- quantile plot presented on Figure 5. can be created. These figures prove the existing of the relationship between slurry density, as a measured process variable, and the operational cost. The knowledge of this fact can be useful

113

to develop a soft-sensor for monitoring the actual operation costs, and can support the economical decisions to enhance the efficiency and profit (Figure 6).

Figure 6. The relationship between the energy cost, and propylene inlet flow

Conclusions Because of the increasing demand of plastics in the world it is necessary to find the chances where the costs can be reduced, which can cause the increasing of profit. In a case of an operational plant it is obvious to try to reduce the operating costs. To realize this target the connection between to operational costs and the process variables is needed to explore. This paper proposed a novel technique based on the analysis of historical process data to explore the appropriate sets of process variables that are important to estimate the operational cost of an operating technology.

Acknowledgement Janos Abonyi is grateful for the support of the Janos Bolyai Research Found of the Hungarian Academy of Sciences, the Oveges Fellowship of the Hungarian Ministry of Education, the support of OTKA 49534 and TEMPUS-TACIS JEP_26045_2005, ECORSE "Ecological and Resource Saving Engineering”.

References [1] Jansen L, The challenge of sustainable development. Journal of Cleaner Production 2003:11:231–245. [2] Zogg D, Shafai E, Geering HP. Fault diagnosis for heat pumps with parameter identification and clustering. Control Engineering Practice, 2006:14:1435 – 1444. [3] Piciarelli C, Foresti GL. On-line trajectory clustering for anomalous events detection. Pattern Recognition Letters 2006:27:1835–1842. [4] Venkatasubramanian V. A Syntactic Pattern-recognition Approach for Process Monitoring and Fault Diagnosis. Engineering Applications of Artificial Intelligence 1995:8(1):35-51. [5] Zhu Y. Multivariable System Identification For Process Control. Applications of Identification in Process Control 2001:251-291.

114

[6] Dayal BS and MacGregor JF., Multi-output process identification. Journal of Process Control 1997:7(4): 269-282. [7] Balasko B, Nemeth S, Janecska A, Nagy T, Nagy G, Abonyi J. Process modeling and simulation for optimization of operating processes. Computer Aided Chemical Engineering 2007:24 : 895-900. [8] Rodríguez-Toral MA, Morton W, Mitchell DR. Using new packages for modelling, equation oriented simulation and optimization of a cogeneration plant. Computers & Chemical Engineering 2000:24(12) : 2667-2685. [9] O’ Reilly AJ. Batch Reactor Optimization, Profitability vs. Waste minimization. Chemical Engineering Research and Design 2002:80(6) : 587-596. [10] Chen J, Askin RG. Throughput maximization in serial production lines with worksharing. Int. J. Production Economics 2006:99 : 88–101. [11] Subawalla H, Zehnder AJ, Turcotte M, Bonin MJ. Multivariable optimization of a multiproduct continuous chemicals facility. ISA Transactions 2004:43 :153–168. [12] Fayyad U, Piatestku-Shapio G, Smyth P. Knowledge discovery and data mining: Towards a unifying framework. Advances in Knowledge Discovery and Data Mining, AAAI/MIT Press, 1994 [13] Mosteller F and Tukey J. Data Analysis and Regression. Addison-Wesley, 1977 [14] Wang D and Romagnoli JA, Robust multi-scale principal components analysis with applications to process monitoring. J. of Proc. Control 2005:15(8):869–882. [15] Uraikul V, Chan CW, Tontiwach-Wuthikul P. Artificial intelligence for monitoring and supervisory control of process systems. Engineering Applications of Artificial Intelligence 2007:20 : 115–131. [16] MacGregor JF and Kourti T, Statistical process control of multivariate processes. Control Eng. Practice 1995:3(3) : 403–414. [17] Lee YH, Min KG, Han C, Chang KS, Choi TH. Process improvement methodology based on multivariate statistical analysis methods. Control Engineering Practice 2004:12 :945–961. [18] Pach FP, Balasko B, Nemeth S, Arva P, Abonyi J. Black-Box and First-Principle Model Based Optimization of Operating Technologies. In Proc. of 5th MATHMOD, Vienna, 2006. [19] Ruiz D, Nougués JM, Calderón Z, Espufia A, Puigjaner L. Neural network based framework for fault diagnosis in batch chemical plants. Computers and Chemical Engineering 2000:24 : 777-784. [20] Amand T, Heyen G, Kalitventzeff B. Plant monitoring and fault detection Synergy between data reconciliation and principal component analysis. Computers and Chemical Engineering 2001:25 : 501–507. [21] Feil B, Abonyi J, Pach PF, Nemeth S, Arva P, Nemeth G and Nagy G. Semi-mechanistic Models for State-Estimation - Soft Sensor for Polymer Melt Index Prediction. Lecture Notes in Computer Science 3070 2004:1111-1117. [22] Lin B, Recke B, Knudsen JKH, Jørgensen SB. A systematic approach for soft sensor development. Computers and Chemical Engineering 2007:31 : 419–425. [23] Choi DJ and Park HY. A hybrid artificial neural network as a software sensor for optimal control of a wastewater treatment process. Wat. Res. 2001:35(16) : 3959-3967.


115

The Use of Plate Heat Exchangers for Energy Saving in Phosphoric Acid Production

Petro Kapustenko, Stanislav Boldyryev, Olga Arsenyeva, Gennadiy Khavin

AO “Sodrugestvo–T”, Krasnoznamenniy per. 2, k. 19, Kharkiv 61002, Ukraine; Tel.: +380577202278 e-mail: [email protected]

Abstract The principles, algorithms and software for energy saving retrofit of phosphoric acid production process are discussed. The essential feature is replacement of tubular heat exchangers with plate ones and installation of units on the new placements to increase heat recovery. The main problem in product behavior during heating or cooling is high intensity of heat transfer surface fouling due to precipitation of gypsum and fluorides from a solution. Intensity of fouling, besides other factors, such as heat exchanger type, its design and parameters of work, in many ways depends on content and structure of impurities in a solution of a product, quantity and the sizes of particles. Despite of the specified difficulties and heavy operation conditions, plate heat exchangers of a various design find the application in these processes. For such corrosively active environment as a phosphoric acid the Hastelloy G30 alloy is used as a material of plates. The minimal thickness of plates from Hastelloy G30 is 0.6 mm. It is necessary to apply synthetic rubber EPDMCT as a material for inter plate gaskets for work with a phosphoric acid. The direction of flow of heat-carriers is counter-current. Calculations were made in view of a 10% margin. It is possible to install М10-BFM units with diameter of collectors 100 mm. The analysis of data shows, that using mixed grouping of plates in the unit allows to satisfy heat exchange requirements fully. Thus, the number of plates (the heat transfer area) is minimal and the condition of pressure drops on heating and heated up heat-carriers is completely satisfied. The software was developed for calculations of units working both with liquid, and with gaseous streams: liquid – liquid; steam/liquid – liquid (condensation); steam/liquid – liquid/steam (condensation – boiling).

Introduction The main aim of this research is development of principles, algorithms and software for energy saving retrofit of phosphoric acid production process. The essential feature of work is replacement of the old equipment of factories with more effective new one. In particular, the replacement of tubular heat exchangers with plate ones and installation of units on the new positions which allow to increase heat recovery. The chemistry of phosphoric acid production from apatite concentrate by wet method is sufficiently developed and rather well described in literature (see e.g. Evenchik et al. [1]). The typical process flowsheet of phosphoric acid production by dehydrate method is presented on Figure 1. This process enables to obtain weak acid with

116

concentration 26-32% of P2O5 and should be followed by evaporation unit to obtain acid with higher P2O5 concentration (see Figure 2). The plate heat exchangers are one of the most efficient types of heat transfer equipment. The principles of their construction and design methods are sufficiently well described elsewhere (see e.g. the books by Hesselgreaves [2], Shah and Seculic[3], Tovazshnyansky et al [4]). This equipment is much more compact and require much less material for heat transfer surface production than conventional shell and tubes units. It makes feasible and economically efficient to produce plates from expensive sophisticated alloys and metals, which can work with highly corrosive substances, such as most of the process streams in phosphoric acid production.

Figure 1. Typical process flowsheet of phosphoric acid production by dehydrate

method

Placement of heat exchangers in wet process of phosphoric acid production

The main placements for application of heat exchangers in manufacturing of phosphoric acid, proceeding from presented process flowsheet, are:

• Cooling of sulfuric acid solution 78 - 98% 2 4H SO – plate heat exchanger (position 1 on Figure 1);

• Cooling of weakly concentrated phosphoric acid (3%) after washing of the filter sediments – plate heat exchanger (position 2 on Figure 1);

• Heating of 30 % phosphoric acid before sulfate sedimentation – spiral or plate heat exchanger (position 3 on Figure 1);

117

• Cooling of phosphoric acid (final product) with 40 – 42% or 50 – 54% 2 5P O concentration after evaporation – plate or spiral heat exchanger (position 4 on Figure 2);

• Cooling of scrubber acid (8 – 11% 2 6H SiF ) (position 5 on Figure 1);

• Evaporation of a phosphoric acid (position 6 on Figure 2); • Cooling the water, which irrigates barometric condensers on an evaporating

station (position 7 on Figure 2). For calculation of heat exchanging units on these positions under the set conditions of operation it is necessary to develop software based on modern algorithms for design of plate heat exchangers.

Figure 2. Flowsheet of concentration unit

Calculation of plate heat exchangers with channels of different geometry in wet process of phosphoric acid production It is possible to identify following processes of cooling and heating of streams with various concentration of phosphoric acid: • Heating of stream which contains 30% 2 5P O from 20ºС up to 40ºС; • Cooling of stream with 54% 2 5P O from 85ºС to 55ºС; • Cooling of stream with 54% 2 5P O from 55ºС to 25ºС. The important feature of stream heating or cooling is high intensity of fouling on heat transfer surfaces due to gypsum and fluorides precipitation from a solution. Intensity of fouling, besides such factors, as heat exchanger type, its design and parameters of work on a rated duty, in many ways depends on structure of impurity in a solution of a product, quantity and the sizes of particles. Despite of the specified difficulties and heavy operation conditions, plate heat exchangers of a various design find the application on these positions. The accounting of fouling was discussed by Gogenko et al. [5].

118

There are different types of units that can be used for heating or cooling of the concentrated phosphoric acid: plate-and-frame heat exchangers and spiral ones. The choice is determined by the structure and quantity of gypsum in a product. Besides, presence of other impurities substantially influences the choice of heat transfer area material. It should be selected depending on acid concentration, temperature range and presence of chlorine, iron and other impurities. It is recommended by Alfa-Laval the use of Hastelloy G30 or graphite for concentration 50% P2O5 at temperature lower than 85 °С and in presence of:

2 3 2 3, 2 41% , , 4% ,600HF Fe O Al O H SO ppm HCl . The synthetic rubber EPDM can be used as gaskets material. The alloys AISI 316L, 904L, 254 SMO, Sanicro 28, C-276, G30 are also used as a heat transfer surface material of spiral heat exchangers depending on specific conditions, see table 1.

Table 1. Composition of alloys for heat transfer surface manufacturing

Metals, % Alloy Cr Nickel Mo Cu Others

AISI 316 17.0 12.0 2.0 - - Avesta 254 SMO 20.0 18.0 6.1 1.7 N 0,2

Alloy C276 15.5 58.0 16.0 - W Hastelloy C22 21.0 44.0 17.5 - W 3, Fe 2-6 Hastelloy G30 29.5 40.0 5.0 1.7 W 2,5, Fe 18-21 Hastelloy D205 20.0 64.5 2.5 2.0 Si 5, Fe 6

Cooling of phosphoric acid with concentration about 3% P2O5 by water after washing the filter deposit should be made from 50°С down to 25-30°С and it require heating of water from 20°С up to 30-35°С. For such conditions the plate heat exchanger with plates from AISI 316 or SMO alloy and gaskets from synthetic rubber EPDM can be selected. The margin of 10% should be accepted in the design to account for possibility of fouling and recommendation to maintain wall shear stress not less than 50 Pа must be followed. The cooling of a finished product of 54 % P2O5 concentration by a phosphoric acid of 30 % P2O5 was considered for calculation of plate heat exchangers with different plates geometry. The cooling is usually made in two stages. At first the concentrated acid is cooled from temperature 85°С down to 55°С and the diluted acid with concentration of 30% heats up from temperature 20°С up to 40°С. The content of components in both streams is presented in Table 2. Thermo physical properties for phosphoric acid of different concentrations at various temperatures has been taken from literature data and approximated. Intermediate values were calculated by means of linear interpolation of adjacent values. Data for calculation of the first stage are presented in Table 3. Further there is a cooling of the concentrated phosphoric acid from temperature 55°С up to 25°С at the second stage, thus the diluted acid with concentration of 30 % heats up from temperature 20°С up to 40°С. Physical properties of streams for 54 % and 30 %

119

concentration for a considered temperature range are presented in Table 4. Data for calculation of the second stage heat exchanger are presented in Table 5.

Table 2 – Composition of phosphoric acid aqueous solutions

Main elements, % 30% P2O5 54% P2O5 P2O5 29.8 53.0

H2SO4 1.9 2.0 SO3 1.6 2.5

Fe2O3 0.15 0.24 F 0.9 0.4

SiO2 0.3 0.03 Cl, ppm 755 165

Table 3 - Data for calculation of the first stage heat exchanger

Heat load Q = 866.8 kW Flow Cooling Heating

Medium Phosphoric acid (54% P2O5)

Phosphoric acid (30% P2O5)

Working pressure, MPа P1 = 0.5 P2 = 0.5 Flow rate, kg/h G1 = 40 000 G2 = 40 000 - 60 000

Inlet temperature, °С T1 = 85 T3 = 20 Outlet temperature, °С T2 = 55 T4 <= 40

Pressure drop, МPа ΔP1 ≤ 0.1 ΔP2 ≤ 0.1

Table 4 - Physical properties of phosphoric acid for 54 % and 30 % concentration

Medium 1 (cooling): physical properties for three temperatures – phosphoric acid 54% P2O5

Temperature, 0С Т1 = 85 Т2 = 70 Т3 = 55 Density, kg/m3 1335 1346 1357 Specific heat, кJ/(kg·0K) 2.659 2.639 2.612 Heat conductivity, W/(м0С) 0.546 0.535 0.520 Dynamic viscosity , centipoises 1.122 1.506 2.753

Medium 2 (heating): physical properties for three temperatures - phosphoric acid 30% P2O5

Temperature, 0С Т1 = 20 Т2 = 30 Т3 = 40 Density, kg/m3 1181 1176 1171 Specific heat, кJ/(kg·0K) 3.237 3.262 3.284 Heat conductivity, W/(м·0С) 0.533 0.47 0.560 Dynamic viscosity, centipoises 2.996 2.310 1.738

For such corrosively active media as a phosphoric acid the Hastelloy G30 alloy is used as a plate material. Its main feature is the highest content of nickel and chrome, high enough presence molybdenum. The minimal thickness of plates from Hastelloy G30 is 0.6 mm. It is necessary to apply synthetic rubber EPDMCT as a material for plate gaskets. The direction of flow of heat carriers is counter-current. Calculations were made in view of a 10% margin on a heat transfer coefficient.

120

Table 5 - Data for calculation of the second stage heat exchanger

Heat load Q = 861.2 kW Flow Cooling Heating

Medium Phosphoric acid (54% P2O5)

Phosphoric acid (30% P2O5)

Working pressure, МPа P1 = 0.5 P2 = 0.5 Flow rate, kg/h G1 = 40 000 G2 = 40 000-60 000 Inlet temperature, °С T1 = 55 T3 = 20 Outlet temperature, °С T2 = 25 T4 <= 40 Pressure drop, МPа ΔP1 ≤ 0.05 ΔP2 ≤ 0.05

It is possible to install М10-BFM units with diameter of collectors 100 mm on both stages. Results of calculations for the first stage with various combinations of plates in the unit are presented in table 6. As one can see the use of plates with different geometry of corrugation in one heat exchanger allows the reduction of heat transfer surface area in comparison with of only one plate type in a heat exchanger. The plates have corrugations with different angle to vertical axis and to main flow direction. Plates of H type have corrugations with bigger angle (about 600) and form the channels H with higher intensity of heat transfer and hydraulic resistance. Plates of L type have lower angle (about 300) and form the channels L with lower heat transfer and hydraulic resistance. Combined together these plates form channels MH or ML with intermediate characteristics (see Figure 3). From results presented in Table 6 it can be recommended to install one pass unit (8.4 м2, 37 plates, grouping 1*(5*MH+13*L) / 1*(15*ML+13*L)) for cooling of the concentrated phosphoric acid at the first stage.

Channel L formed by L-

plates Channel ML-MH formed

by L- and H- plates Channel L formed by H-

plates

Figure 3. Channels formed by combining plates of different geometry Results of calculations for the second stage with various combinations of plates in one unit are presented in table 7. Advantages of using the various plate types in one unit on the second stage are seen even more clearly, in spite of the fact that two-passes heat exchanger is needed due to operating conditions. From the presented results of calculations it is possible to recommend for installation two-passes unit with the area 50.8 м2 (213 plates) and grouping 2*(42*H+11*MH)/2*(42*H+11*ML).

121

Table 6 - Results of calculation for heat exchangers with different plate types combinations for the first stage

Plate combina-

tion

Grouping

Number of plates

Area, m2

Pressure drop

(hot side), kPa

Pressure drop

(cold side), kPa

MH/L–ML/L 1×(3MH+9L)/1× (3ML+9L)

25 6.0 61.3 85.8

MH - ML 1×15MH/1×15ML 31 7.44 61.8 87.1 H - H 1×25H/1×25H 51 12.24 59.1 95.0 L - L 1×15L/1×15L 31 7.44 36.0 58.3

Table 7 - Results of calculation of heat exchangers with various plates combinations

for the second stage

Plate combina-

tion

Grouping

Plates number

Area, m2

Pressure drop

(hot side), kPa

Pressure drop

(cold side), kPa

H/MH–H/ML

2×(24H+18MH)/2× (24H+18ML)

169 40.56 33.5 47.7

MH - ML 3×39MH/3×39ML 235 56.4 37.3 48.2 H - H 2×52H/2×52H 209 50.16 34.2 49.0

The analysis of data shows, that the use of mixed grouping of plates in one heat exchanger (type H/MH - H/ML and MH/L-ML/L) allows to satisfy heat exchange requirements fully. Thus, the number of plates (the heat transfer area) is minimal and the condition of pressure drops on one of streams is completely satisfied. The application of the calculation algorithm for mixed-grouping plate heat exchangers allows implementing following advantages: • To intensify process of a heat transfer and to reduce the heat transfer area; • An opportunity to select one pass units; • Completely utilize the available pressure drop; • Eliminate pressure losses compared with multi-pass heat exchangers.

Software for plate heat exchangers calculation To design plate heat exchangers working at different conditions at phosphoric acid production processes the software was developed which enables to calculate heat transfer area, the number and grouping of the plates and to select an optimal heat exchanger for the specified duty. The software permits to calculate heat exchangers working with different conditions of the streams, namely: liquid – liquid, steam/liquid – liquid (condensation) and steam/liquid – liquid/steam (condensation – boiling). The basis for calculation algorithms development was described earlier in papers by Tovazshnyansky et al. [6], [7], [8]. The main feature of the processes with phase change, like condensation and boiling, is considerable variation of the process parameters, velocity of the stream and vapor content and others, along the channels length, which was accounted using one dimensional mathematical model of

122

combined heat and mass transfer. The software was developed at AO Sodrugestvo-T and is used for calculation of plate heat exchangers which are produced by this company in Ukraine from plates supplied by Alfa-Laval.

Conclusions In this paper the possibilities of plate heat exchangers application and their placement in phosphoric acid production process are discussed. The high corrosive activity of heat exchanging streams requires the use of highly corrosion resistant materials for plates manufacturing. The possible choice of alloys is presented. To minimize the surface area of plate heat exchangers the reliable methods and algorithms for their design are needed. It is shown that one of the possibilities to minimize heat transfer area is the principle of combination in one heat exchanger of plates with different geometry of corrugations on their heat transfer surface. In case of phase change in plate heat exchanger channels (condensation or boiling) for reliable calculations it is necessary to take into account the change of all process parameters along the channel length. The software which accounting all mentioned features of plate heat exchanger design and enables accurate calculation of heat transfer area for heat exchanger, the number and grouping of plates, has been developed. It facilitates the use of plate heat exchangers in phosphoric acid production to increase energy and material saving potential.

Acknowledgments The financial support of EC Project ECOPHOS (contract INCO-CT-2005-013359) is sincerely acknowledged.

References [1] Evenchik S et al.Technology of Phosphoric and Complex Fertilizers. Moscow, Chemistry, 1987. (In Russian). [2] Hesselgreaves JE. Compact Heat Exchangers. Selection, Design and Operation. Amsterdam, Elsevier, 2001. [3]Shah RK and Seculic DP. Fundamentals of Heat Exchanger Design. New York, Wiley and sons, 2003. [4] Tovazshnyansky LL, Kapustenko PO, Khavin GL, Arsenyeva OP. Plate Heat Exchangers in Industry. Kharkiv, NTU KhPI, 2004.(in Russian). [5] Gogenko AL, Anipko OB, Arsenyeva OP, Kapustenko PO. Accounting for fouling in plate heat exchanger design.In: Klemes J, editor. Proceedings of the 10th Conference on Process Integration, Modelling and Optimisation for Energy Saving and Pollution Reduction (PRES’07). Chemical Engineering Transactions. Published by AIDIC 2007, v.12: 207-213. [6] Tovazhnyansky LL, Kapustenko PO. Intensification of heat and mass transfer in channels of plate condensers. Chemical Eng.Comm. (USA) 1984; 31(6):351-366. [7] Tovazshnyansky LL, Kapustenko PO, Perevertaylenko OY. The investigation of flow boiling for flows in channels with cross-corrugated walls. Heat Transfer Engineering 2002; 23(6):62-69. [8] Tovazshnyansky LL, Kapustenko PO, Nagorna OG, Perevertaylenko OY. The simulation of multicomponent mixtures condensation in plate condensers. Heat Transfer Engineering 2004; 25(5):16-22.


123

The role of Process Integration in CCS Systems in Oil Refineries and Pulp Mills

Erik Hektor1, Thore Berntsson2

Heat and Power Technology, Chalmers University of Technology, SE-412 96 Göteborg, Sweden, fax: +46 31 82 19 28;

1telephone: +46 31 772 8536, e-mail: [email protected] 2telephone: +46 31 772 3009, e-mail: [email protected]

Abstract Capturing CO2 from flue gases in process industries like oil refineries and pulp mills can be an important part in reducing the CO2 emissions in the world. The advantage with process industries when capturing CO2 is the existence of excess heat. In this paper it is shown that process integration and the use of excess for regenerating the absorbent is a good way of reducing avoidance cost. A reduction of up to 10-15 EUR/t is achievable dependent on the system layout.

Introduction In a climate conscious society the CO2 emissions must be reduced. One way of reducing the emissions is to capture and store the CO2 from flue gases. One main drawback with this technology is the large amount of energy needed for regeneration of the absorbent. However, in large process industries like pulp mills and oil refineries where there are large energy flows there is often a large amount of excess heat at various temperature levels. This excess heat can be used to cover part of the energy demand of the regeneration and thereby reduce the cost for CO2 capture. In this paper it is presented how the avoidance cost is affected by different examples of process integration at a pulp mill. Identified possibilities for oil refineries are also presented.

Pulp mill The calculations in this work are based on the eco-cyclic reference pulp mill (KAM), which was a Swedish national research and development programme. The pulp mill is a fictitious market kraft pulp mill based on the best available technology in Sweden and Finland in the late 1990s [1]. A chemical absorption unit is added to the recovery boiler in order to capture the CO2 from the flue gases. Schematic charts of the studied configurations can be found in Figures 1, 2 and 3. Chemical absorption is a proven way to separate CO2 from flue gases and requires no rebuilding of the recovery boiler. In order to regenerate the absorbent, large quantities of heat are needed for the reboiler in the desorption unit. After the capture, the CO2 is compressed in two stages with inter-cooling to a pressure of 80 bar. In order to satisfy the need for heat for regeneration of the absorbent, a condensing turbine is removed and the rest of the heat demand can be satisfied in different ways, described in more detail in upcoming sections. Two different absorbents are considered, mono-ethanolamine (MEA) and ammonia (NH3).

124

Mono-ethanolamine Using MEA as an absorbent is a commercially available way of cleaning flue gases. The absorption takes place at a temperature of 50-60 °C and the desorption takes place at a temperature of 110 °C. The heat demand for regenerating MEA is 2880 kJ/kg CO2 [2].Different ways of supplying this heat demand are explored. First a reference case is considered by just producing more steam. Then three types of process integration alternatives are considered, using a heat pump, an energy combine and creating excess steam. Data for the configurations can be found in Table 1. For more in-depth description and analysis of the investigated configurations, see Hektor and Berntsson [3], [4], [5].

Table 1- Key specifications for the studied configuration

Orig. mill

MEA BIO

MEA HP

MEA H2

MEA MeOH

MEA ES-1

MEA ES-2

NH3 1

NH3 2

Black liquor (MW) 338 338 338 338 338 338 338 338 338 External fuel (MW) - 74 - 428 428 29 - - - CO2 formed (kg/s) 29 37 29 74 29 32 29 29 29 CO2 capture rate (kg/s) - 33 27 71 27 29 27 26 26 MP steam to mill (tonne/h) 101 101 101 101 101 86 86 101 101 LP steam to mill (tonne/h) 137 137 137 137 137 93 93 137 137 Heat consumption for CO2

separation (MW) - 96 77 77 77 84 77 39 39

Power consumption for heat pump (MW) - - 13 - - - - - -

Power consumption for CO2 capture (MW) - 16 13 35 13 14 13 9 17

Steam turbine output (MW) 53 62 53 53 53 60 51 49 57 Net electricity output from BIGCC (MW) - - 23 12 - - - -

Net electricity output (MW) 53 46 27 42 52 46 38 40 40 Mill electricity consumption (MW) 39 39 39 39 39 39 39 39 39

Electricity surplus (MW) 14 7 -12 3 13 7 -1 1 1 Hydrogen production (MW) - - - 148 - - - - - Methanol production (MW) - - - - 142 - - - - Boiler In the boiler configuration (BIO), a boiler is added in order to produce a sufficient amount of steam for the regeneration of the absorbent. The fuel is biofuel. The boiler produces high-pressure steam that is fed into the steam cycle of the mill and then expanded in the steam turbine to low pressure steam. The CO2 from the biofuel boiler is also captured. A simplified chart of the configuration is shown in Figure 1. Heat pump The heat pump (HP) uses the excess heat from the pulp mill to supply the heat needed for regeneration, see Figure 1. It is designed using the software IEA Heat Pumping Technologies Annex XXI [6]. The purpose of this software is to suggest suitable working fluids and compressors for a given application. The suggested heat pump is an electrically driven closed cycle using two three-stage turbo compressors with R114 as working fluid. It has a coefficient of performance of 4.2. R114 is being

125

substituted due to environmental reasons, but it is assumed that new working fluids with the same performance will be developed.

Figure 1. Simplified chart of the boiler (to the left) and heat pump (to the right)

configurations BIGCC The BIGCC is combined with either hydrogen (H2) or methanol (MeOH) production and the configurations are based on hydrogen concept 1 and methanol concept 1 presented in Hamelinck et al [7]. The excess heat generated in the hydrogen and methanol production is used in the CO2 separation equipment at the mill, i.e. the reboiler in the desorption unit for regeneration of MEA. A simplified chart can be found in Figure 2. The size of the BIGCC is set by the need for excess heat. Excess steam Another way to supply the heat needed for regeneration is to save steam in the pulping processes through thermal process integration and use this excess steam (ES) for the capture process. The steam surplus is 2 GJ/ADt, which corresponds to 36 MW [8]. This steam is not sufficient for the regeneration of the absorbent. Two alternatives are considered here.

In ES-alternative 1 a biofuel boiler is installed to supply the rest of the steam needed. The high pressure steam generated by the biofuel boiler is fed into the steam cycle of the mill and then expanded in the steam turbine to low pressure steam. The CO2 from the boiler is also captured.

126

Figure 2. Simplified chart for the BIGCC configurations

Figure 3. Simplified chart of the chilled ammonia process

In ES-alternative 2 the process integration is taken one step further. The CO2 absorption process is modified and the reboiler uses medium pressure steam. Heat from the absorption process is then used to produce low pressure steam that is needed at the mill. Through these measures there is no need for external fuel, but there is a loss of electricity production since the medium pressure steam is not allowed to expand to low pressure steam in the turbine. Since the temperature in the reboiler is increased, MEA cannot be used as an absorbent. In the calculations in this study a fictive absorbent working at higher temperatures is used. The fictive absorbent is assumed to have the same steam consumption for regeneration as

127

MEA and the size and type of equipment needed for absorption and desorption, and cost are also assumed to be the same as for MEA. If the equilibrium water vapour pressure is higher then it is in the incoming gases, a part of the heat of absorption will be used for the vaporization. Therefore, the absorbent is presumed to have a high boiling point elevation and maybe even a negative deviation from Raoult’s law so that the water vapour pressure in the absorption column is kept constant and the heat of CO2 absorption can be recovered for production of LP steam.

Ammonia The ammonia system used here is the chilled ammonia capture system described by EPRI [9]. The flue gases are cooled down to around 2 °C before the absorber. After the absorption the CO2-rich slurry is pumped through a heat exchanger to a high pressure desorber of 30 bar and the working temperature is 115 °C (NH3 1). The high pressure in the desorber reduces the need for compression work of the CO2. The heat demand for regeneration is 1520 kJ/kg CO2. The available steam from the removed condensing turbine is sufficient to cover the heat demand. A configuration (NH3 2) where the condensing turbine is still in place and the pressure in the desorber is lowered in order to allow the use of excess heat from the mill. The lowest temperature possible in the desorber at atmospheric pressure is 60 °C. Data for the configurations can be found in Table 1.

Economic evaluation In order to evaluate the performance of the suggested systems for capturing the CO2, the cost of each avoided tonne of CO2 is calculated:

annual

annualavoid m

Cc = (EUR/tonne CO2) (1)

where mannual Annually avoided CO2 (tonne) trFpprodeMOinvannual CpFpEpECCC +⋅Δ+⋅Δ+⋅Δ+Δ+Δ= .& (EUR) (2) where ΔCinv Annualised investment cost for new and changed equipment (EUR) ΔCO&M Annual change in operating and maintenance costs (EUR) ΔE Annual change in net electricity output (MWh) pe Electricity price (EUR/MWh) ΔEprod. Annual change in electricity production from renewable sources (MWh) pp Price or power certificates (EUR/MWh) ΔF Annual change in fuel demand (MWh) pF Fuel price (EUR/MWh) Ctr Annual cost for transportation and storage of the CO2 (EUR) The annual avoided amount of CO2 is different for the studied configurations. The avoided CO2 amounts are the produced CO2 emissions for the delivered services (pulp, electricity and fuel) when there is no capture minus the emitted CO2 with the capture. The energy market parameters are taken from the energy market scenarios described below.

128

Description of scenarios The energy market scenarios used here are created by Axelsson et al [10] and are based on an energy market model adapted for evaluation of investments in the process industry. The main input data required for the model are fossil fuel prices and policy instruments. The main output from the model include electric power prices, willingness to pay for biofuel and CO2 emissions associated with marginal use of electricity and biofuel. Four different scenarios for 2020 are used. Two levels of fossil fuel prices and CO2 charge were combined into four sets of input data for the scenarios, see Table 2. The scenarios are used here as an indication of how different futures affect the economic performance of the technology in each case, i.e. as a packaged sensitivity analysis.

Table 2- Energy market scenarios

Sce-nario

Biofuel prices (EUR/MWhfuel)

Electric power (EUR/MWhel)

CO2 emissionsa (kg/MWhfuel / el)

CO2 charge (EUR/tonne)

Pellets Chips Byprod Price Power cert Biofuel Electricity 1 25 17 14 54 16 329 426 27 2 34 23 20 59 5 329 136 43 3 27 18 15 57 16 329 723 27 4 49 33 21 62 5 159 136 43 a From marginal use, see Axelsson et al [10] Investment cost The investment costs included correspond to the incremental investment cost compared to the original mill. The pulp mill is regarded as new or a mill in need of improvement. The equipment needed in the original mill is not included unless it has been changed due to CO2 capture (steam turbine). When there is a change in equipment, it is the difference between a new unit designed as needed and a new unit of the original design that is used. The incremental investment cost compared to the original mill can be viewed in Table 3. In the calculations the investment costs have been annualised using the annuity method. An annuity factor of 0.1 has been used.

Table 3- Incremental investment cost compared to original mill.

MEA BIO

MEA HP

MEA H2

MEA MeOH

MEA ES-1

MEA ES-2

NH3 1

NH3 2

Investment cost [MEUR] 76 77 342 299 80 69 31 32 Maintenance cost The maintenance cost has been estimated as 4 percent of the cost of the components in the analysis. It is the increase compared to the original mill that is of interest. The maintenance cost includes the cost for the refill and handling of MEA or ammonia [11]. Increased fuel demand The CO2 capture demands large amounts of heat. To produce the heat needed, extra fuel is needed in most configurations. As described above, the increased fuel demand is supplied by biofuel. Electricity balance When the absorption unit and boiler, if any, are added the steam balance of the pulp mill is changed, and this affects the electricity production on site. The electricity

129

demand is also increased due to the CO2. The figure that is of interest for the evaluation is the net change in electricity surplus compared to the original pulp mill. Cost for transportation and storage of CO2 The cost for storage of the captured CO2 is based on estimations made for storage in an on-shore aquifer in southwest Skåne in the southern part of Sweden. The storage cost is estimated to be around 3.3 EUR/t CO2 according to Elforsk [12]. When it comes to the cost of transporting the CO2 from the pulp mill to the storage facility, the cost varies with transportation distance and way of transporting the CO2. The CO2 is transported in pipelines since that is the cheapest way of transporting the CO2 in these quantities and the cost can be calculated according to the following function, (0.1+8/1000*Distance) EUR/t CO2 [13].

Figure 4. Avoidance cost for the studied configurations. Transport distance 500 km

Results The avoidance cost calculated for the studied configurations, with a transportation distance of 500 km, are presented in Figure 4. It should be noted that the absolute levels of the avoidance costs calculated are only valid under the conditions specified in the scenarios. Nevertheless, the costs give a good understanding of how the performance of the configurations relate to one another. The investment cost data are based upon a situation where post combustion CO2 capture and transportation is commercially available and it is not valid for the first units being built. This means that the results cannot be compared with results presented for the first units being built. The annuity factor used in the calculations is 0.1, since it is being considered a strategic investment. A higher annuity factor would of course give a higher avoidance cost.

130

Figure 5. Energy balance for the studied configurations

The results show that a high CO2 charge is needed for CO2 capture to be profitable. The lowest avoidance cost is achieved when ammonia is used as the absorbent, this is due to the low energy demand. When looking at the results for the configurations using MEA, one can see that an increased degree of process integration leads to a reduced avoidance cost. A reduction of around 10-15 EUR/t can be achieved. This is due to the use of excess heat from the pulp mill, which leads to a decrease in the need for external fuel. The energy demand for the studied configurations can be viewed in Figure 5.

Oil refineries In a refinery one of the main sources of CO2 are the fired heaters. CO2 can be separated from these flue gases. The challenge is that they are distributed around the refinery. By having separate local absorber for each heater stack the size of the flue gas ducts can be minimized. The regeneration and CO2 handling is done at a convenient central location. The solvent can readily be circulated back-and-forth between the absorbers and the desorption unit. By separating these activities both the cost and space required are reduced. [14] The regeneration is as mentioned before a process with a considerable energy demand. The energy demand depends on which absorbent is used. In refineries where there often are large amounts of excess heat, there is also the possibility to use the excess heat for the regeneration. How the excess heat can be used depends on the temperature level of the excess heat and the temperature level of the regeneration, which varies from absorbent to absorbent. The results from pulp mills presented in this paper show that integrating the processes can reduce the avoidance cost significantly. In Figure 6, the available excess heat at different temperature levels for a refinery that is being studied by the authors can be viewed.

131

The refinery has a capacity of 5.4 Mt crude oil and emits 0.54 Mt CO2 per year. [15] The figure shows that there are large amounts of excess heat available at the temperature levels of interest for CO2 capture.

Figure 6. Cooling demand at a Swedish refinery

Discussion As can be seen from the results presented here, process integration can play an important role in reducing the cost for CO2 capture in process industries. However, the excess heat can also be used in other ways than for CO2 capture. In previous work by the authors a comparison was made with producing electricity from the excess steam and the result showed that with high CO2 charges it is more profitable to capture and store CO2 than to produce more electricity [5]. For refineries calculations are needed to see the performance for CO2 capture. From the results one can see that a tough CO2 reduction target and subsequently high CO2 charge are necessary for CO2 capture to be a reality. With the expected levels of the CO2 charge within the European Union, the economic performance is very good. For CO2 capture in pulp and paper mills it is also necessary for CO2 emission reductions from biofuel to be valued in the same way as for fossil fuels.

Conclusions Main conclusions from the study:

• Process industries with available excess have an interesting possibility of reducing the need for external fuel for the capture process and thereby reducing the avoidance cost.

• Initial studies in refineries indicate that there are attractive possibilities for process integration.

Acknowledgement The authors would like thank the European Community for the financial support through the EMINENT-2 project TREN/05/FP6EN/S07.56209/019886 fexecuted under the 6-th Framework Programme for Research and Technological Development.

132

References [1] STFI. Ecocyclic Pulp Mill – Final report KAM 1. KAM Report No. A31, 1999. [2] Möllersten K, Gao L, Yan J. Efficient energy systems with CO2 capture and storage from renewable biomass in pulp and paper mills. Renewable Energy 2004;29(9):1583-98. [3] Hektor E, Berntsson T. Carbon Dioxide Capture in the Pulp and Paper Industry. In: Proceedings of the 4th Nordic Minisymposium on CO2 Capture and Storage, Helsinki (Finland), September 2005. [4] Hektor E, Berntsson T. CO2 capture from recovery boiler flue gases with biomass energy or heat pump. In: Proceedings of the 8th International Conference on Greenhouse Gas Control Technologies, Trondheim (Norway), june 2006. [5] Hektor E, Berntsson T. Future CO2 removal from pulp mills – Process integration consequences. Energy Conversion and Management 2007;48(11):3025-3033. [6] IEA. Industrial Heat Pump Screening Program. IEA Heat Pump Centre, Borås (Sweden), 1997. [7] Hamelinck CN, Faaij APC. Future prospects for production of methanol and hydrogen from biomass. Universiteit Utrecht, Copernicus Institute, 2001. [8] Algehed J, Wising U, Berntsson T. Energy efficient evaporation in future pulp and paper mills. In:Proceedings of the 7th Conference on New Available Technologies, SPCI, 2002. [9] EPRI. Chilled Ammonia Post Combustion CO2 Capture System – Laboratory and Economic Evaluation Results. EPRI Report 1012797, November 2006. [10] Axelsson E, Harvey S, Berntsson T. A tool for creating energy market scenarios for evaluation of investments in energy intensive industry. In: Proceedings of ECOS 2007 Padova, (Italy), June 2007. [11] Chladná Z, Chladný M, Möllersten K, Obersteiner M. Investment under Multiple Uncertainties: The Case of Future Pulp and Paper Mills. IIASA Report No. IR-03-XXX, 2003. [12] Elforsk. Avskiljning och lagring av koldioxid i ett nordiskt systemperspektiv. Elforsk rapport 05:27, oktober 2005. [13] Svensson R, Odenberger M, Johnsson F, Strömberg L. Transportation systems for CO2 – application to CCS. Energy Conversion and Management 2004;45:2343-2533. [14] IEA. CO2 abatement in oil refineries: Fired heaters. Report no PH3/31, October 2000. [15] Chalmers EnergiCentrum. Towards a sustainable oil refinery – Pre-study for larger co-operation projects. Report – CEC 2008:x, 2008.


133

Energy Saving Processes in Biomass Refinery Endre Nagy

Research Institute of Chemical and Process Engineering, FIT, University of Pannonia, Veszprém, Hungary, email: [email protected]

Abstract Bioenergy from renewable resources is already today a viable alternative to fossil fuels; however, to meet the increasing need for bioenergy several raw materials have to be considered for the production of e.g. bioethanol and biogas. In this paper the bioethanol production has been studied. Bioethanol can be produced from e.g. sugars, starch and various lignocellulosic materials such as straw, wood and waste. The production of ethanol from sugars and starch is known and widely applied. The ethanol production from lignocellulose creates severe technical challenges, such as a need for efficient pretreatment. In lignocelluloses materials cellulose, a linear polymer of glucose is associated with hemicelluloses and surrounded by lignin seal. Lignin, a complex three-dimensional polyaromatic matrix prevents enzymes from accessing some regions of the cellulose polymers. Crystallinity of the cellulose further impedes enzymatic hydrolysis. The pretreatment of lignocellulosics is primarily employed to open up the lignocellulosic structure, to increase the accessible surface area of cellulose to enhance the conversion of cellulose to glucose. These treatments are still mostly thermal, thermo-mechanical or thermo-chemical and require a considerable input of energy.

The main milestones of biorefinery development Bioethanol can be produced from e.g. sugars, starch and various lignocellulosic materials such as straw, wood and waste (Solomon et al. [1]; Petersson et al. [2]). Biorefineries represent complex systems of sustainable, environment- and resource-friendly technologies for the comprehensive utilization and the exploitation of biological raw materials as plant biomass. Each biorefinery refines and converts its corresponding biological raw materials into a multitude of valuable products. The product palette of a biorefinery includes not only the products in a petroleum refinery, but also in particular products that are not accessible in petroleum refinery. The plant biomass always consists of the basic products carbohydrates (about 75 % of the biomass, mainly in the form of cellulose, starch and saccharose; from that it is about 35-50 % of dry weight cellulose, and 20-35 % hemicellulose), lignin (10-25 % of biomass), and other material (5% of biomass) as proteins and fats (oils), beside various substances, such as vitamins, dyes, flavors and aromatic essences, which have very different chemical structures. Three main classes of biorefinery can be distinguished (Kamm and Kamm, [3]): Phase I biorefinery: an example of phase I biorefinery is a dry-milling ethanol plant. It uses grain as a feedstock, has a fixed processing capability and produces a fixed amount of ethanol, feed co-product and carbon dioxide (Kamm and Kamm, [3], [4]). The increase of the grain ethanol production in the U.S is given, as an example in Fig. 1.

134

Fig. 1. The grain ethanol production in the U.S. (Solomon et al. [1])

The conventional dry mill consists of grinding, cooking, liquefaction, saccharification of the starch to sugars with enzymes, fermentation of the sugars to ethanol with yeast, followed by distillation and dehydration processes of ethanol (Fig. 2, Huang et al. [5]). Recently, another modified dry-grind processes were developed, which allows separation of non-fermentable corn components such as germ and fiber for further reduction of cost.

Fig. 2. A dry corn mill process (Huang et al. [5])

Phase II biorefinery is the current wet-milling technology uses grain feedstock, yet has the capability of producing various end-products, depending on product demand (starch, high fructose syrup, ethanol, corn oil and corn gluten feed and meal). In the wet mill process (Fig. 3) corn is cleaned, steeped, de-germed to obtain germ from which corn oil is extracted, defibered to obtain, and subjected to separation of gluten and starch (Huang et al. [5]). Phase III biorefinery (biomass feedstock-mix + process-mix→ product-mix) is illustrated in Fig. 4 (Kamm and Kamm, [3]). A phase III biorefinery is not only able to produce a variety of chemicals, fuels and intermediates or end-products, but can also use various types of feedstocks and processing methods to produce products for industrial market. The flexibility of its feedstock is the factor of first priority for

135

adaptability towards changes in demand and supply for feed, food and industrial commodities.

Fig. 3. Corn wet mill process (Huang et al. [5])

It is necessary to develop new biorefinery-basic technologies, such as: (1) the lignocellulosic feedstock biorefinery (LCF), including LCF pre-treatment and effective separation into lignin, cellulose and hemicelluloses, (2) further development of thermal, chemical and mechanical processes, such as extractive methods, gasification (syngas) and liquefaction of biomass, (3) further development of biological processes (biosynthesis, bacteria for the degradation of starch and cellulose, etc.), (4) combinations of substantial conversions, such as biotechnological and chemical processes, (5) corn biorefinery concepts, (6) promotion of research and development into phase III biorefinery (Fig. 4).

Lignocellulose feedstock (LCF) biorefinery While significant progress has been made in the conversion of lignocellulosic or cellulosic biomass -- such as agricultural residues (e.g. corn stover, crop straws, sugar cane bagasse), herbaceous crops (e.g. switchgrass) foresty wastes, wood (hardwoods, softwoods) wastepaper and other wastes such as municipal waste -- to fuel ethanol, it has not yet been commercialized due to existing technical, economic, and commercial barriers. However, cellulosic ethanol can be more effective and promising as an alternative renewable bio-fuel than corn ethanol in the long run because it could greatly reduce the net greenhouse gas emissions as well as higher net fossil fuel displacement potential (Huang et al. [5]) Conversion of lignocelluloses materials to higher value products requires fractionation of the materials into its components: cellulose, hemicelluloses, and lignin. Hemicellulose is a heterogeneous polymer comprising of pentose (xylose, arabinose), hexoses (mannose, glucose, galactose) and sugar acids. The removal and recovery of hemicelluloses is an essential feature of pretreatment processes for biological conversion to ethanol or other products.

136

Fig. 4. Basic principles of the phase III biorefinery (Kamm and Kamm, [3])

The general equations of the conversion of the compounds of the lignocellulosic feedstock are as follows (Kamm and Kamm, [3]):

Lignocellulose+ H2O → Lignin + Cellulose + Hemicellulose Hemicellulose + H2O → Xylose

Xylose (C5H10O5) + catalyst → Furfural (C5H4O2) + H2O Cellulose (C6H10O5) + H2O → Glucose (C6H12O6) The basic process for conversion of cellulosic biomass to fuel ethanol mainly consists of the following steps (Huang et al. [5]): feedstock handling, pretreatment and conditioning/detoxification, saccharification and co-fermentation, product separation and purification, wastewater treatment, product storage, lignin combustion for production of electricity and steam and all other utilities (Fig. 5). This overall process involves the following separation steps: removal of inhibitors in fermentor (aliphatic acids, furan derivatives, and phenolic compounds); ethanol recovers from the fermentation broth and its dehydration; pre-extraction of hemicelluloses and separation of hemicelluloses from other components in the extract when considering separate fermentation of pentose and hexoses.

Pretreatments methods for pre-extraction of hemicelluloses To date, a variety of effective pretreatment methods to hydrolyze and fractionate hemicelluloses components have been investigated (Moiser et al. [6], Huang et al. [5]):

137

Fig. 5 Overall process block diagram for a basic lignocellulose to ethanol biorefinery

(Huang et al. [5]) dilute acid pretreatment (Sassner et al. [7]), liquid hot water extraction (Mosier et al. [6]), steam explosion-based extraction Öhgren et al. [8]), dilute acid-steam explosion (Viola et al. [9]), alkaline extraction (Kaar and Holtzapple [10]), ammonia fiber/freeze explosion (Xu et al., [11]), alkaline-acid method (Spigno et al., [12]),

acid-alkaline/oxidative method (Spigno et al. [12]), wet-oxidation (Oleskowitz-Popiel et al. [13]), microwave pretreatment (Linde et al. [14]).

Deriving fermentable sugars from lignocellulosic biomass is one of the major R&D issues. The evaluation shows that the pretreatment technologies which have been studied in great detail over the last years, e.g. strong and weak acid hydrolysis and steam pretreatment, still suffer from drawbacks as formation of inhibitors, the need to regenerate acids, formation of inorganic waste streams, etc. A mild alkaline [Ca(OH)2, NaoH] or weak acid hydrolysis using CO2 dissolved in pressurized hot water seem to be more attractive for pretreatment. These pretreatment methods are needed for further investigation and development. The energy needs for the pretreatment processes can be very different. Generally it can be said that they can essentially increase the cost of the bioethanol production.

The overall process steps of lignocellulosic bioethanol production According to Fig. 5 the lignocellulose-to-ethanol technology will comprise the following unit operations (Reith et al. [15]): 1. Biomass treatment/handling (milling/chipping) for size reduction;

138

2. Pretreatment for mobilization of the (hemi)cellulose and the lignin biopolymers and breakage of structural components;

3. Liquefaction: hydrolysis of the highly viscous polysaccharide matrix to a liquid steam of sugar oligomers;

4. Saccharification: enzymatic hydrolysis of sugar oligomers; 5. Liquid/solid separation of the lignin fraction for use as a fuel of electricity; 6. Evaporation: concentration of sugar content to a sugar level sufficient for a final

ethanol concentration of at least 8-9% in the fermentation broth; 7. Fermentation of C6 (glucose) and C5 (xylose) sugars to ethanol; 8. Separation of yeast usually by centrifugation; 9. Distillation for separation and upgrading of ethanol to ca 45 vol%; 10. Rectification: concentration to approx. 96 vol%; 11. Dehydration of ethanol (> 99.9 vol%); 12. Dewatering and drying of “stillage”, the bottoms fraction for the distillation

process; 13. Thermal conversion of “non-fermentable organics”;

Ecomomics of ethanol production Existing ethanol plant have varied in size from 1500 m3 to 1hm3 (=106 m3) year-1 of production capacity (Solomon et al. [1]). About 80 percent of production, including at all recent plants, occurs in anhydrous (dry grind) mills, with the rest made from wet mills. The main cost components are capital and the feedstock supply. It is very difficult to precisely model the production technology. Solomon recommends a simple equation in order to predict the production cost: CA=CC+CK+CL+CE+CM+CO-CE where CA is the cost of ethyl alcohol production (Euro/L); CC the cost of amount of corn yielding a liter of ethanol (Euro/L); CK the cost of energy; CM the cost of raw materials; CO the other costs, including maintenance, water, overhead, insurance, taxes, etc.; CE the price of distiller’s dried grain to be sold. The same equation can be used for cellulosic ethanol, as well. As it was shown, several technological configurations are being actively researched and developed to produce ethanol from cellulosic biomass. The reference technology assumed by Solomon et al [1] for the production cost estimation is dilute acid pretreatment and enzymatic hydrolysis of cellulose, since it offers the best near-term potential for commercialization competitive with fuel ethanol from grain. The data are based to the plant to build in US with production capacity of 0.11hm3 ethanol. Table 1 lists the estimated cost of cellulosic ethanol production given by Solomon et al. [1].

Table 1. Estimated cost of cellulosic ethanol production (the data in US Dollar was recalculate to Euro, 1Euro=1.58 US Dollar)

Cost category eurocent/L Feedstock (wood or switchgrass) 11.5 Enzymes 3.35 Other raw materials (sulfuric acid, lime, glucose, nutrients) 1.65 Gypsum disposal 0.16

139

Electricity -1.4 Water 0.07 Labor 0.67 Maintenance 2.21 Direct overhead 0.4 General overhead 2.0 Insurance & property taxes 1.1 Total cost 21.7 Annualized capital charge 14.3 Total production cost 36.0 Several factors could further lower the production cost of cellulosic ethanol. These include the use of cheap residues for biomass feedstocks, integration into higher value chemical products, to increase the glucose content of the fermentation broth, to improve the cost and efficiency of the enzyme, energy demand reduction by energy integration, to optimize the energy demanding process steps directly, e.g. by employing more advanced distillation or evaporation systems or by replacing these steps with less energy-intensive unit operations, recirculation of process stream for water supply, applying pervaporation for ethanol dehydration, etc. Let us look at two examples which can reduce the production cost or energy. Fig 6. illustrates the energy demand of ethanol production as a function of the ethanol content of the fermentation broth given by Öhgren et al. [8]. The higher ethanol concentration reduces the energy demand of the further concentration of the ethanol solution. In lignocellulosic processes the sugar content of the solution to be fermented is generally low, it should be increased by evaporation before feeding it in the fermentation reactor. Alzate and Toro [16] analyzed the energy cost of distillation and evaporation by different process configurations. They obtained that this cost can be reduced about 18 % (from 52.5 MJ/LEtOH to 43 MJ/LEtOH) with the process configuration used simultaneous saccharification and cofermentation, conventional distillation pervaporation, stillage evaporation as well as water recycling, comparing to that of the basic case (the basic process contains the following process steps: enzymatic hydrolysis, hexose fermentation, pentose fermentation, conventional distillation, azeotropic distillation, evaporation; other processes applied in both cases are: dilute acid pretreatment, ion exchange detoxification).

140

Fig. 6 Energy demand in a single distillation unit for concentration

of dilute ethanol feed to 94.5 % (w/w) Wingren et al. [17] investigated the energy demand for SSF- based (simultaneous simultaneous saccharification and fermentation) softwood ethanol plant also applying different configurations. They obtained that in base case configuration, consisting of three thermally coupled distillation columns and multiple-effect evaporation units with 5 evaporators, the process steps with the highest net energy demand are distillation and evaporation. These two steps require 29.6 MW steam. Pretreatment and drying require high pressure steam but also produce significant amounts of secondary steam (16.3 MW), which can replace some of primary steam needed in the distillation and evaporation step. The total energy demand in the primary steam is thus 38.8 MW or 19.0 MJ/LEtOH and the ethanol production cost 0.37 Euro/LEtOH (the detailed cost for production of ethanol, in Euro/LEtOH, is as follows: raw material: 0.128; solid fuel:-0.045; capital: 0.136; electricity: 0.017; chemicals: 0.023; enzymes: 0.054; other: 0.06). Different alternatives were considered, such as integration of a stripper with the evaporation step, increasing the number of evaporation effects and the application of mechanical vapor recompression to the evaporation step. With these configurations the energy demand could be reduced about 8 %. Further energy saving can be reached using more membrane processes replacing partly the evaporation and the distillation steps for the concentration of the solutions, to my opinion. E. g. applying reverse osmosis instead of evaporation the ethanol production cost can be reduced about 25 % in batch fermentation (Zacchi and Axelsoon, [18] ). The pervaporation offers excellent opportunities for reduction of the distillation and rectification costs. Figs. 7A and 7B illustrate the excellent separation properties of the ceramic and zeolite membranes for dehydration of organic solutions. Because the energy demand of the pervaporation process can be essentially lower than that of distillation or rectification, its application to produce poor ethanol can reduce the cost of the bioethanol production.

141

A) B) Fig. 7. Graphical representation for separation of ethanol-water by ceramic (Fig. 7A)

and by zeolite membrane (Fig. 7B) (Chapman et al. [19])

Conclusion The lignocellulosic ethanol production consists of several process steps with different energy demands and production cost. The cost of the production should be reduced, for it primarily the enzyme cost, the pretreatment cost, the cost of ethanol concentration increase and glucose concentration increase could/should be lowered. This paper discussed some aspects of the energy/cost reduction.

References [1] Solomon BD, Barnes JR, Halvorsen KE. Grain and cellulosic ethanol: History, economics, and energy policy. Biomass and Bioenergy, 2007:31:416-425. [2] Petersson A, Thomsen MH, Hauggard-Nielsen H, Thomsen A-B. Potential bioethanol and biogas production using lignocellulosic biomass from winter rye, oilseed rape and faba bean. Biomass and Bioenergy 2007:31:812-819. [3] Kamm B, Kamm M. Principles of biorefineries. Appl. Microbiol. Biotechnol 2004:64:137-145. [4] Kamm B, Kamm M. Das Konzept der Bioraffinerie- Produktion von Plattformchemikalien und Finalprodukten. Chemie Ingeniur technik 2007:79:592-603. [5] Huang H-J, Ramaswamy S, Tschirner UW, Ramarao BV. A review of separation technologies in current and future biorefineries. Sep. Purif. Technol. (2008):62:1-21. [6] Mosier N, Wyman Ch, Dale B, Elander R, Lee YY, Holtzappe M, Ladish M. Features of promising technologies for pretreatment of lignocellulosic biomass. Bioresource Technology, 2005:96:673-686. [7] Sassner P, Galbe M, Zacchi G. Techno-economic evaluation of bioethanol production from three different lignocellulosic materials. Biomass and Bioenergy 2008 doi:10.1016/j.biombioe.2007.10.014. [8] Öhgren K, Rudolf A, Galbe M, Zacchi G. Fuel ethanol production from steam-pretreated corn stover using SSF at higher dry matter content. Biomass and Bioenergy 2006:30:863-869. [9] Viola E, Cardinale M, Santarcangelo R, Villone A, Zimbardi F. Biomass and Bioenergy 2008 doi: 10.1016/j.biombioe.2007.12.009. [10] Kaar WE, Holtzapple, MT. Using lime pretreatment to facilitate the enzymatic hydrolysis of corn stover. Biomass and Bioenergy 2000:18:189-199.

142

[11] Xu Z, Wang Q, Jiang ZH, Yang X-x, Ji Y. Enzymatic hydrolysis of pretreated soybean straw. Biomass and Bioenergy 2007:31:162-167. [12] Spigno G, Pizzorno T, De Faveri DM. Cellulose and hemicelluloses recovery from grape stalks. Bioresource Technology 2008:99:4329-4337. [13] Oleskowicz-Popiel P, Lisiecki P, Holm-Nielsen JB, Thomsen AB, Thomsen MH. Ethanol production from maize silage as lignocellulosic biomass in anaerobically digested and wet-oxidized manure. Bioresource Technology 2007, doi:10.1016/j.biortech.2007.11.029. [14] Linde M, Galbe M, Zacchi G. Bioethanol production from non-starch carbohydrate residues in process streams from a dry-mill ethanol plant. Bioresource Technology 2008, doi:10.1016/j.biortech.2007.11.032. [15] Reith JH, den Uil H, van Veen H, de Lat WATM, Niessen JJ, de Jong E, Elbersen, HW, Weusthuis R, van Dijken JP, Raamsdonk. Co-production of bio-ethanol, electricity and heat from biomass residues. Proc. of 12th European Conference and Technology Exhibition on Biomass for Energy, June 17-21, 2002, Amsterdam, The Nederlands. [16] Alzate CAA, Toro OJS. Energy consumption analysis of integrated flowsheets for production of fuel ethanol from lignocellulosic biomass. Energy, 2006:31:2447-2459. [17] Wingren A, Galbe M, Zacchi G. Energy consideration for a SSF-based softwood ethanol plant. Bioresource Technology 2008:99(7):2121-2131. [18] Zacchi G, Axelsoon A. Economic evaluation of preconcentration in production of ethanol from dilute sugar solutions. Biotechnology and Bioengenieering 1989:34:223-233. [19] Chapman PD, Oliviera T, Livingston AG, Li K. Membranes for the dehydration of solvents by pervaporation. J. Membrane Sci., (2008) doi:10.1016/j.memsci.200802. 061.


143

Energy Efficiency Improvement of Wastewater Treatment Processes Using Process Integration Techniques (Anaerobic Digestion)

J.Rojas-Hernandes, T. Zhelev

Charles Parsons Initiative, University of Limerick, Ireland Tel: +35361213474, Fax: +353-61-202568, E-mail: [email protected]

Abstract The focus of this paper is on an attempt to make a further step towards cost attractiveness of waste treatment processes when experiencing rapidly raising fuel prices. It includes two main ideas. The first one is exploring the potential possibility to introduce an anaerobic treatment component in the classical activated sludge treatment facility, thus decreasing the energy bill through bio-gas generation. The second option is the utilization of process integration advances, developed for large industrial operations for energy efficiency improvement of wastewater treatment plants. Two case studies are used to explore the concepts. The analysis of the sludge organic content for the case-study plant demonstrates a great potential for energy reuse. The study concludes: Anaerobic digestion can provide up to 100% of plant energy requirements. The combination of the two proposed efficiency improvement options demonstrates potential for practical implementation.

Introduction Ireland is one of the fastest growing economies in the EU for the past decade. With the fast developing industries, the demand for energy supplies is increasing. The raising concern about expected prospect of a bigger contribution to greenhouse gas emissions and climate change is a good reason for searching opportunities for energy conservation and energy-efficient technologies as the fastest and cheapest way towards energy independence. Another environmental problem facing all modern societies is the huge amount of waste generation and the problems of minimising its effect on the environment. Waste can be a precious source of energy, which can help solving the above problems and contribute to the global challenge - achieving a sustainable and secure delivery of energy, and reducing greenhouse gas emissions. Municipal Wastewater Treatment Plants (WWTP) are substantial contributor to the raising energy demands due to the growing Irish population. The forecasted growth in energy demand is 2-3% annually until 2020, with continued heavy dependence on imported fossil fuels and a need to invest in energy infrastructure. Energy efficiency is declared as a priority for Ireland. The objective is to deliver cumulative improvements of 20% in energy efficiency by 2020. To meet the future targets set in the governmental Green Paper, unprofitable WWTP technological systems must be redesigned. One of the most common energy efficiency targeting technologies in WWT is sludge stabilisation by Anaerobic Digestion (AD). Reused biogas can sufficiently cover the process heating demand and can power electricity generation for internal needs and

144

export. Although the AD technology seems to be fully developed and known, there are many opportunities for improvement. The need to sophisticate the system as a whole and integrate AD with other existing processes on site is calling for improved technologies and employment of new methodologies for energy management [1]. This paper attempts to develop the idea of energy efficient Wastewater Treatment Plants using process integration techniques.

Objective and Methodology The objective of the study is to create guidelines and assemble the scope for future work in energy conservation of wastewater treatment processes such as anaerobic digestion. The methodology discusses literature available studies and expands the potential for implementation of energy recovery techniques and best practice solutions making a further step towards future improvements based on a chosen case study. This study focuses on anaerobic digestion technologies for the treatment of the organic fraction of municipal wastewater. Anaerobic digestion produces stabilised biosolids, reduces biomass by the degradation of volatile solids, and produces biogas as a by-product. Process is the consequence of a series of metabolic interactions among various groups of micro-organisms and occurs in a series of stages: hydrolysis, acidogenesis, acetogenesis and methanogenesis [2]. Methane, the main focus of this study, is formed in the last stage, therefore, it is important to respect the sequence of operations and analyse the systems as a whole.

Operating Conditions The balance of the process requires consideration of the following parameters

• Waste composition • Temperature

0

5

10

15

0 20 40 60 80

Temperature (Celcius)

Gas

Pro

duct

ion

(Vol

ume/

time)

Figure 1. Effect of Temperature on Gas Production.

The rate of methane generation increases with temperature. It is possible to indicate the relation between biogas generation and temperature in the AD process (Fig 1). Below 10°C gas production decreases drastically, and above 30-35°C operation of the digester depends on the energy input for digester heating which can be uneconomical. This shows that mesophilic digesting provides an optimal range of temperature [2].

• pH value.

145

Typically, anaerobic digestion operates in the narrow pH range of 5.5-8.5. Once the production of methane is stabilised, the pH level stays between 7.2 & 8.2.

• Mixing. Slow mixing is preferred, preventing microbe degradation during excessive mixing [3]

• Metabolism Time. Hydraulic retention time (HRT) and solids retention time (SRT) are also important to the anaerobic digestion process. HRT must be long enough to achieve the required degree of volatile solids degradation [3]. To increase the performance of the process, the dispersed-growth digesters can recycle part of their slurry to the digesters to retain more active biomass and increase the solids retention time.

• Volatile Fatty Acids Digestion of fatty acids is important because it can yield to approx. 75% of methane. Propic acids are valuable indicators of process performance [3].

• Chemical Oxygen Demand Another helpful indicator of AD processes and its relation to methane production is Chemical Oxygen Demand. The quantity of methane generation can be estimated as methane produced per gram of COD removed.

• Total solids content/Organic loading rate The AD process can be classified according to the total solids (TS) content of the slurry in the digester reactor. Organic loading rate (ORL) is a measure of the biological conversion capacity of AD.

• Biogas composition Biogas as a by-product can be used as an alternative source of energy. Monitoring biogas quality can be regarded as measure of digestion efficiency [4]. The gas obtained during the AD process contains methane 55-56%, carbon dioxide 35-45%, Nitrogen 0-3%, Hydrogen 0-3%, and Hydrogen sulphide 0-1%. Usually 100 – 200 m3 of total gas are produced per ton of organic matter digested. Biogas composition depends mostly on the raw material, organic loading, time, alkalinity and temperature of decomposition [3]. One m3 of biogas with a methane content of 70% (20MJ/m3) is equivalent to: 1.70 kWh of electricity (assuming the efficiency of 30%), and 2.50 kWh of heat only (efficiency of 70%). Up to one third of biogas energy may be needed to heat the influent and maintain the digester temperature, although the average requirement is closer to 10%. The most common use of biogas is to direct it into the CHP unit and recover electricity and heat.

Energy Efficiency Biogas is also used as fuel in natural gas boilers, which are often part of the sludge drying system. The temperature of the sludge often exceeds 37°C because of the microbial activity and the obtained heat can used for sludge preheating which is also a significant option for energy reuse. Energy conservation within anaerobic digestion is influenced mostly by biogas production and its application issue. A focused attention towards obtaining better and greater volume of biogas and its efficient usage and energy conservation must be supported by detailed energy analysis of the entire anaerobic treatment system.

146

Energy Management and Ideas for Improvement The main energy demands in conventional digesters are associated with heating, pumping, recirculation and mixing. The analysis shows that there are a number of areas within the AD process where energy efficiency can be improved [5].

Figure 2. Typical biogas energy management in wastewater treatment [3]

Figure 2 supports the ideas about preferable areas of expected energy efficiency improvement in WWTP. Some ideas support sludge combustion for power generation. This technology is barely feasible in terms of energy gained. Burning 1 tDS (tonne dry solids) sludge can generate 3.360 MJ/tDS primary energy equivalent. However, the entire treatment process will require 3,150 MJ/tDS much of this is used for treating stack emissions to meet environmental requirements. In addition, 1.150 MJ/tDS will be consumed in mechanically dewatering the sludge prior to incineration. The analysis of all energy using operations shows that the biggest contributors to the energy bill are the pumping and aeration systems. It is estimated that wastewater pumping represents 70% of the total electricity requirement. Mixed digestion has been proposed to enhance operational efficiency. Some authors suggested separate stages of digestion; separate hydrolysis phases, or mix of mezophilic with thermophilic digestion [5]. This paper mainly focuses on so called homogenized anaerobic digestion. The idea is based on two principles: (a) Separation of primary and secondary sludge; (b) Increase the efficiency the biomass biodegradation. Those principles are based on different bio-degradability of different types of sludge. Sludge transferred to digesters mainly consists of secondary biomass containing some living bacteria forms (overflowed activated sludge) and some primary sludge. It is claimed that about 50% of the volatile solids in primary sludge can be converted into biogas, when for the secondary sludge this percent is between 10 to 15%. To provide greater biodegradable mass in the secondary sludge, intensive mixing (homogenisation) is proposed with the help of special sharp propellers. The homogenisation causes degradation of living microorganisms. This provides more biodegradable substrate for digesting and increases biogas production. Destroyed cells release captured water thus improving dewatering characteristics of sludge. This also leads to improved sludge density making it easier to pump with a less hydraulic resistance.

147

Figure 3. Redesign for gas generation efficiency improvement

Figure 3 shows conceptual redesign (highlighted) of anaerobic digestion including sludge homogenization applied for our case-study plant.

Case Study Chosen case study attempts to demonstrate the feasibility and viability of the idea to introduce an anaerobic treatment section in a conventional activated sludge facility in the chosen case study plant. The plant reports high energy bills caused mainly by pumping and thermal stabilization of the sludge using natural gas [6]. The idea is to implement anaerobic digestion and generate biogas, which can cover partially or fully the high energy demands of the treatment process.

Feasibility Tests and Calculations Potential of biogas production is evaluated according to the current standards based on organic dry solids in tested sludge [3]. Three types of sludge were tested: Imported (regarded as a secondary sludge), primary (after the first sedimentation tank) and Waste activated sludge (secondary sludge). Calculations related to homogenised anaerobic digestion option: Primary sludge: (50% biodegradability) – 757.8 kg VS/1000 kg of DS 0.5*757.8 = 378.9 kg of destroyed organic dry solids 0.8*378.9kg = 303.12 Nm3 biogas produced per 1000 kg dry solids. Caloric value: 5.5 *303.12=1667.16 kWh/1000 kg of sludge Primary sludge directed to digesting system daily: 7 590 kg DS /day 1667.16 kWh* 7.6 = 12670.4 kWh /day

148

Secondary and imported sludge (10-15% biodegradability): Secondary: 652.4 kg VS/1000 kg DS and imported 771.5 kg VS/1000 kg DS. 0.15*1523.5 = 228.525 kg of destroyed organic dry solids 0.8 * 228.525 = 182.82 Nm3 biogas produced per 1000kg dry solids. 303.12 + 182.82 = 485.94 Nm3 of biogas at caloric value of 5.5 kWh / m3 Biogas: 5.5*485.94 = 2672.67 kWh / 1000 kg of DS transported to digesters.

Daily amount of secondary and imported sludge: Secondary – 6888kg/day; Imported – 3365,6kg/day; Total:10 253.6 kg DS sludge loaded 2672.67 kWh * 10.253 = 27402 kWh/day Total production (split sludge degradation): 12670.4 kWh /day + 27402 kWh/day = 40072.4 kWh/day Mixed sludge was tested to calculate the potential of biogas generation with 65% organic reduction: (0.8 – m3 of biogas yield per kg of V.S destroyed and V.S. degradation – hydraulic retention time (HRT) 30 days – 65%) Mixed sludge daily = 16 409.35 kg/day at 3.24% DS = 683.72 kg / h DS 0.65*16 409.35 kg = 10666.07 kg VS/day 0.8*10666.07 kg = 8532.862 m3 of biogas yield/ day Caloric value of biogas: 8532.82 m3*5.5 = 46 930.73 kWh Summary of generated energy:

Energy Demand for Anaerobic Treatment: Energy demand for the entire process = heating the sludge + electricity (pumps for circulating + sludge mixing) + energy for dewatering,

Heat:

a) summer period: Q1=(37-14)*1*154.20644*1000/860 Q1=4124.12573 kWh/d heat losses are estimated at 15% so Q1= 4742.75kWh/d b) winter period: Q2=5557.56kWh/d Electricity: a) Pumps for recirculation and sludge mixing: 0.03-0.05 kWh/m3 sludge or 0.05*154.20644=7.71 kWh/d b) Energy for dewatering; Mechanical dewatering - 0.9 kWh/m3 sludge or 0.9*154.20644=138.78 kWh/d Qtot = 5557.56 kWh/d + 7.71 kWh/d + 138.78 kWh/d = 5704.05 kWh/d

Calculation method Energy [kWh/d] Based on split degradation of sludge 40072.4 65% degradation 46930.73

149

Key Observations/Summary The split sludge biodegradability method resulted in 40043 kWh/day of energy produced, which fully covers the process energy requirements. Conventional biogas calculations (65% reduction) gives a figure of 46 930.73 kWh energy obtained. The energy requirements for the treatment plant selected as a case study are in excess of 5.88 GWh per annum. Figures obtained, based on 65% organic solids reduction, gave 17.13 GWh per year. The feasibility study of anaerobic digestion in our case study plant indicates that biogas production would be a highly prospective system for energy generation, which would cover all energy requirements of the plant.

Process Integration for Efficiency Improvement In parallel with the bio-gas generation the hypothesis of this paper promotes the application of process integration technique for energy efficiency improvement of wastewater treatment plants. To prove or disapprove the applicability of the Pinch analysis for better energy management in anaerobic wastewater treatment facilities, the example of a newly proposed modern facility was taken as a case study. The Process and Instrumentation diagram for the proposed CAD plant (Figure 4) [6] indicates that there is a potential scope for heat integration. Heat source for internal process heating purposes is supplied from the biogas generated. It is used mainly to preheat the influent from reception storage, which should reach 38 +/- 1oC tanks and to assist pasteurisation in pasteurisation tanks, where temperature should be increased to 65-70°C. The heat content of digested sludge (product) transferred to storage can also be a source of low potential heat. The temperature in the hot water line (accepted as utility stream) is approximately 90oC throughout the system. The problem of heat integration can be summarised as design of a system of heat exchanger, which would require the least capital cost and consume minimum of home generated bio-gas, thus increasing the potential of energy (gas) export without jeopardising the specified process efficiency. Extracted heat exchanger network as proposed by the design project is shown in Figure 5.

Stream Data The waste accepted to the plant is a blend of three different sources: (a) Cattle manure, poultry slurry (both arriving to plant on tracks) and (b) industrial waste arriving through a pipeline. Table 1 summarises the stream data accounting for the highest predicted load during the year. Table 1: Stream data

Stream name Ts (°C) Tt (°C) Cp (J/ kg °C) m (kg/s) ΔH (J/s)

1 - From storage to pasteurising tank 10 70 4050 8.3 2024919

2 - From pasteurising tank to digester 70 38 4050 8.3 -1079957

3 - From digester to storage 36 20 4050 8.3 -539978 There are two parallel trains of the treatment facility as shown in Figure 4 and Figure 5. The optimal heat exchanger network based on the scenario when the two trains

150

are exposed to different loads is shown in Figures 6. The solution is obtained using the Pinch concept applied through the SuperTarget software.

1 2

Pasteurising tank A (50 m3)

70oC

Digester A (2000 m3) 36 +/- 1oC

Boiler biogas

3

10oC

20oC

58oC

70oC

38oC

36oC

26oC

1Feedstock storage B

2

Pasteurising tank B (50 m3)

70oC

Digester B (2000 m3) 36 +/- 1oC

3

10oC

20oC

58oC

70oC

38oC

36oC

26oC

Digested matter storage B

Feedstock storage A

Digested matter storage A

Figure 5. Heat exchanger network as per proposed design

Conclusion Presented study explored two options for energy efficiency improvement of wide-spread wastewater treatment facilities, based on activated sludge digestion. The first one was to study any eventual possibility to introduce anaerobic digestion section and study the viability of possible satisfaction of internal energy demand from internal energy resources, mainly based on bio-gas generation. This option shows definite potential even for the quite energy demanding case of dry pasteurised sludge. The second option for energy efficiency improvement – the application of the heat integration approach also showed a good impact to the drive for minimisation of heating utility powered by combustion of generated bio-gas.

151

A future step would be the combination of these two concepts, designing integrated anaerobic digestion plants. The negative side of proposed solutions is the fact that the introduction of an anaerobic treatment section in an existing activated sludge facility requires substantial overhaul including extra area and large investment.

Figu

re 4

. Pro

cess

and

Inst

rum

ent D

iagr

am fo

r the

Pro

pose

d C

AD

Pla

nt

152

The conclusion from this study underlines that the suitability of combined anaerobic and heat integration concept would be bigger for new designs as compared to retrofits. Acknowledgement Authors would like to acknowledge the support of EPA and DCENR, Ireland for this study

References [1] White Paper/Green Paper, Department of Communication, Energy and Natural Resources, Ireland (2007). [2] Metcalf & Eddy, “Wastewater Engineering, Treatment Disposal Reuse”, McGrow-Hill International Educations (1991). [3] Verma, S., “Anaerobic digestion of Biodegradable Organics in Municipal Waste”, Dept. of Earth and Environmental Engineering, Columbia University, May 2002. [4] European Commission Directorate - General for Energy (DG XVII) European Seminar on “Energy Efficient Options for the Treatment of sewage Sludge”, Dublin, Ireland, 6 May 1993. [5] Polprasert, C., Environmental Engineering Program, Asian Institute of Technology, “Organic Waste Recycling, Technology and Management”, Report, Bangkok (1996). [6] Mahony, T., Feasibility study for centralised anaerobic digestion for treatment of various wastes and wastewaters in sensitive catchment areas - Johnstown Castle, Co. Wexford: Environmental Protection Agency (Project: 1840950870), 2002.


153

Minimisation of Energy Use in Multipurpose Batch Plants Using Heat Storage: A Mathematical Programming Approach

Thokozani Majozi

Department of Chemical Engineering, University of Pretoria, Lynnwood Road, Pretoria, 0002, South Africa

email: [email protected]

Abstract The concept of heat integration in batch chemical plants has been in literature for more than a decade. However, most publications tend to advocate direct rather than indirect heat integration in batch chemical processes. Direct heat integration is encountered when both the source and the sink processes have to be active over a common time interval, assuming that the thermal driving forces allow. On the other hand, indirect heat integration allows heat integration of processes regardless of the time interval, as long as the source process takes place before the sink process so as to store energy or heat for later use. The thermal driving forces, nonetheless, must still be obeyed even in this type of heat integration. It is, therefore, evident from the foregoing statements that direct heat integration is more constrained than indirect heat integration. Presented in this paper is a mathematically rigorous technique for optimization of energy use through the exploitation of heat storage in heat integrated multipurpose batch plants. Storage of heat is effected through the use of a heat transfer fluid. The resultant mathematical formulation exhibits a mixed integer linear programming (MILP) stucture, which yields a globally optimal solution for a predefined storage size.

Introduction Until recently, heat integration has always been the privilege of continuous rather than batch chemical processes. This is mainly due to the fact that, in general, heat integration techniques assume steady-state behaviour, which is a feature of continuous processes. Moreover, batch operations tend to be less energy-intensive than their continuous counterparts. However, the increasing popularity of batch plants and the continuing global emphasis on emissions reduction is starting to warrant either the adaptation of the well-established heat integration techniques to or the development of novel techniques for batch processes. The increase in popularity is due to the flexibility and adaptability of batch plants, which is crucial in the current volatile market trends. It is also worthy of note that, although external utility requirement is a secondary economic issue inmost batch facilities, e.g. agrochemicals and pharmaceuticals, it can be significant in others, e.g. dairy and brewing [11]. Early work on heat integration of batch plants was proposed by Vaselanak et al. [18]. These authors explored heat integration of batch vessels containing hot fluid that required cooling and cold fluid that required heating. Four cases were investigated. In the first case, the fluid from one vessel was allowed to return to the same vessel after exchanging heat with the fluid of another vessel via a common heat exchanger. In

154

the second case, a heating or cooling medium was used to transfer heat between the hot fluid vessel and the cold fluid vessel, thereby maintaining the heat integrated fluids within the vessels throughout the heat exchange process. The third case entailed the transfer of fluids from their original vessels to receiving vessels while being heated or cooled. The fourth case was the combination of the above cases. Implicit in their analysis was the given schedule of the operations. A heuristic procedure was proposed for the cases where the final temperatures were not limiting and an MILP formulation for the cases where the final temperatures were limiting. Subsequent to this, other mathematical formulations for heat integration of batch processes have been proposed by Peneva et al. [14], Ivanov et al. [3], Corominas et al. [2], Papageorgiou et al. [13], Vaklieva-Bancheva et al. [17], and Adonyi et al. [1]. Peneva et al. [14] and Ivanov et al. [3] addressed the problem of designing a minimum total cost heat exchanger network for given pair wise matches of batch vessels. An implicit predefined schedule was also assumed. Corominas et al. [2] considered the problem of designing a minimum cost heat exchanger network and a heat exchange strategy for multiproduct batch plants operating in a campaign mode. The objective was to maximize heat exchange in a pre-specified campaign of product batches with hot streams requiring cooling and cold streams requiring heating. The emphasis on campaign mode implies that the proposed methodology cannot be applied in situations where equipment scheduling is of essence. Papageourgiou et al. [13] extended the discrete-time formulation of Kondili et al. [7] for scheduling of multipurpose batch plants by including heat integration aspects. Direct and indirect heat integration configurations were addressed. The main drawback of all discrete-time formulations is their explosive binary dimension, which requires enormous computational effort. Vaklieva-Bancheva et al. [17] improved the work of Ivanov et al. [3] by embedding the heat integration framework within an overall scheduling framework. However, the authors only addressed a special case in which the plant is assumed to operate in a zero-wait overlapping mode, where each product must pass through a subset of the equipment stages, and production is organized in a series of long campaigns. Recently, Pinto et al. [15] presented a discrete-time mixed integer mathematical formulation for the design of heat integrated multipurpose plants based on superstructure approach. A graph theory based technique that incorporates heat integration within scheduling of multipurpose plants has also been proposed by Adonyi et al. [1] with emphasis on make span minimization. Other established attempts on heat integration of batch plants are based on the concept of pinch analysis [8], which was initially developed for continuous processes at steady-state. As such, these methods assume a pseudo-continuous behaviour in batch operations either by averaging time over a fixed time horizon of interest [9] or assuming fixed production schedule within which opportunities for heat integration are explored [5,6,12,4]. It is, therefore, evident that these methods cannot be applied in situations where a production schedule that maximizes heat integration whilst optimizing production demands is sought. The methodology presented in this paper is the extension of the methodology developed by Majozi [19] which was only aimed at direct heat integration ogf batch plants. The main advantages of this methodology are that the start and end times of processes need not be specified a priori and requires very few binary variables due to uneven discretization of the time horizon of interest. The extension pertains to the inclusion of heat storage as a possibility for saving more energy and allowing overall

155

flexibility of the process. The mathematical model is linear which implies that the solution corresponding to a predefined size of storage is globally optimal.

Problem Statement The problem addressed in this paper can be stated as follows. Given: (i) production scheduling data, i.e. equipment capacities, task durations, time

horizon of interest, recipe for each product as well as cost of raw materials and selling price of final products,

(ii) hot and cold duties for tasks that require heating and cooling, respectively, (iii) cost of cooling water and steam, (iv) operating temperatures for the heat source and the heat sink operations, (v) allowed minimum temperature difference and (vi) available heat storage capacity determine the production schedule that results in minimum energy use or maximum profit. In the context of this paper, profit is defined as the difference between revenue and operating costs. The latter constitute raw material costs and external utility (cooling water and steam) costs. It is assumed that sufficient temperature driving forces exist between matched tasks for process–process heat transfer. Also, each task is allowed to operate either in an integrated or standalone mode. The integrated mode, in the context of this paper, is twofold, since a unit is allowed to be integrated with either heat storage or another operating unit. If heat integration cannot supply sufficient duty, external utility is supplied to complement the deficit. Whilst direct heat integration requires involved tasks to be active within a common time interval to effect direct heat transfer, they need not necessarily commence nor end at the same time. Moreover, the heat integrated tasks can either belong to the same process or distinct processes within reasonable proximity.

Mathematical Model As aforementioned, the mathematical model proposed in this paper is an extension of the earlier work by the same author. It entails the following sets, variables and parameters. Sets

unit storage heat a isuuU |= unit processing a isjjJ |= JjjJ ccc ⊂= cooling requires that unit processing a is| JjjJ hhh ⊂= heating requires that unit processing a is|

point time a isppP |= unit processing a to stream input an isjinjinjin ssS ,,, |= unit processing a from stream output an isjoutjoutjout ssS ,,, |=

156

Variables

( )pst joutp ,, = time at which the stream is produced from unit j ( )pst jinu ,, = time at which the stream enters the processing unit j ( )puT ,0 = initial temperature in the storage vessel at time point p ( )puTf , = final temperature in the storage vessel at time point p

( )pjCW , = amount of external cooling required by operation j at time point p ( )pjST , = amount of external heating required by operation j at time point p

( )pujQ ,, = amount of heat exchanged with storage at time point p

( )⎩⎨⎧

←←

=otherwise

unit storage withheat exchanging is unit if 01

,,uj

pujy

( )⎩⎨⎧

←

′←=′

otherwise unit another withheat exchanging is unit if

01

,,jj

pjjx

( )⎩⎨⎧

←←

=otherwise

point time at active is unit if 01

,,

pjpsy jin

Parameters

( )jT = operating temperature for processing unit j ( )jτ = duration of operation j in standalone mode ( )jj ′′ ,τ = duration of operation j when directly heat integrated ( )uj,τ ′′ = duration of operation j when integrated with storage

minTΔ = minimum temperature difference ( )jQ = amount of heat required by or removed from the operating unit j ( )uM = capacity of heat storage u

Constraints The mathematical model is based on the superstructure shown below. The heat transfer fluid in heat storage remains in the storage vessel during heat transfer with only the process fluid pumped around. The superstructure also shows that each unit is capable of receiving external heating or cooling in addition to direct and indirect heat integration

157

Figure 1 Superstructure for the mathematical model

In addition to scheduling constraints that have been presented in detail in another publication [10], the following constraints are necessary to cater for heat storage. Constraints (1) and (2) ensure that direct heat integration involves exactly one pair of units so as to simplify process operability. In essence, these constraints state that if 2 units are heat integrated at any given point in time, then these units must also be active at that point in time. However, if a unit is active at a given time point it is not necessary that it be heat integrated with another unit. ( ) ( ) jinjinhjin

Jj

SsJjPppsypjjxc

,,, ,,,,,, ∈∈∈∀≤′∑∈′

(1)

( ) ( ) jinjincjin

Jj

SsJjPppsypjjxh

,,, ,,,,,, ∈∈∈∀≤′ ′′

∈∑ (2)

Constraints (3) and (4) quantify the amount of heat transferred and received from storage unit, respectively. They ensure that if there is no heat integration between a processing unit and storage, then the amount of heat related to storage is not disturbed.

( ) ( ) ( ) ( )( ) ( )

UuPpJJj

pujypuTpuTcuMpujQ

c

fp

∈∈⊂∈∀

−−−=

,,

,1,,1,,,, 0

(3)

( ) ( ) ( ) ( )( ) ( )

UuPpJJj

pujypuTpuTcuMpujQ

h

fp

∈∈⊂∈∀

−′−−=′

,,

,1,,,1,,, 0

(4)

Constraint (5) ensures that only one unit is heat integrated with storage at any given point in time. Constraints (6) and (7) ensure that the temperature of the storage unit is not changed if there is no heat integration with any unit. These constraints carry the same meaning as constraints (3) and (4). Nonetheless, they are necessary since

158

they pertain to temperature whilst the latter pertain to the amount of heat. Overall, constraints (3), (4), (6) and (7) govern the relationship between heat and temperature of storage.

( ) ( ) UuPppujypujyhc JjJj

∈∈∀≤′+∑∑∈′∈

,,1,,,, (5)

( ) ( ) ( ) ( ) ( )

UuPp

pujypujyjTpuTpuThc JjJj

jf

∈∈∀

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛′−+≤− ∑∑

∈′∈

,

,,,,,max,1,0

(6)

( ) ( ) ( ) ( ) ( )

UuPp

pujypujyjTpuTpuThc JjJj

jf

∈∈∀

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛′−+≥− ∑∑

∈′∈

,

,,,,,max,1,0

(7)

Constraint (8) ensures that the initial temperature in heat storage at any given point in time is the same as the final temperature at the last time point. This condition is always true, regardless of the heat integration status in the previous time point. ( ) ( ) UuPppuTpuT f ∈∈∀−= ,,1,,0 (8) Constraints (9) and (10) ensure that if there is heat integration between any unit and heat storage, then the stipulated minimum driving force should be obeyed. Constraint (9) applies if heat storage is integrated with the heat source, whilst constraint (10) applies if heat storage is integrated with the heat sink.

( ) ( ) ( ) ( )( )

UuJJjPp

pujyjTTjTpuT

c

jf

∈⊂∈∈∀

−+Δ+≤

,,

,,,1max, min

(9)

( ) ( ) ( ) ( )( )

UuJJjPp

pujyjTTjTpuT

h

jf

∈⊂∈∈∀

−+Δ+≥

,,

,,,1max, min

(10)

Constraint (11) states that cooling in any heat source will be accomplished either by direct heat integration, external cooling or heat integration with storage. Constraint (12) is similar to constraint (11) but applies to a heat sink.

( ) ( ) ( ) ( ) ( ) ( ) ( )

UuPpJJj

pjjyjQjQpjCWpujQpsyjQ

c

Jjjjjin

h

∈∈⊂∈∀

′′++= ∑∈′

′

,,

,,,min,,,,,,

(11)

159

( ) ( ) ( ) ( ) ( ) ( ) ( )

UuPpJJj

pjjyjQjQpjSTpujQpsyjQ

h

Jjjjjin

c

∈∈⊂∈∀

′′++= ∑∈

′

,,

,,,min,,,,,,

(12)

Constraints (13) and (14) state that if a unit is directly heat integrated with another unit, then it cannot be simultaneously integrated with heat storage. This is also a condition imposed solely to simplify operability of the overall process. ( ) ( ) UuPpJJjpujypjjy c

Jj h

∈∈⊂∈∀≤+′∑∈′

,,,1,,,, (13)

( ) ( ) UuPpJJjpujypjjy h

Jj c

∈∈⊂∈′∀≤′+′∑∈

,,,1,,,, (14)

Constraint (15) shows how the variation in duration due to the heat integration mode is accounted for in the mathematical model. It is very likely that the duration times will be affected by the mode of operation and this should not be ignored in the formulation.

( ) ( ) ( ) ( ) ( )( )

( ) ( ) ( ) ( )SssUuPpJj

pjjyjjpujyuj

pujypjjyjpstpst

joutjin

jinujoutp

∈∈∈∈∀

′′′+′′+

−′−+−=

,,

,,

,,,,,,,,,,,

,,,,11,,

ττ

τ

(15)

The foregoing constraints constitute the full heat storage model. With the exception of constraints (3) and (4), all the constraints are linear. Constraints (3) and (4) entail non-convex bilinear terms which render the overall model a non-convex MINLP. However, the type of bi-linearity exhibited by these constraints can be readily removed without compromising the accuracy of the model using the so called Glover trans

( ) ( ) ( ) ( )( )

UuPpJJj

pujpujcuMpujQ

c

fp

∈∈⊂∈∀

Γ−Γ=

,,

,,,,,,, 0

(3’)

( ) ( ) ( )( ) ( )( ) ( ) ( )( )

UuPpJJj

pujyuTpuT

pujpujyuTpuT

c

Lff

fUff

∈∈⊂∈∀

−−+≤

Γ≤−−−

,,

,1,,1,

,,1,,1,

(16)

( ) ( ) ( ) ( )

UuPpJJjpujyTpujpujyuT

c

Uff

Lf

∈∈⊂∈∀

−≤Γ≤−

,,,1,,,,1,,

(17)

160

( ) ( ) ( )( ) ( )

( ) ( ) ( )( )UuPpJJj

pujyuTpuT

pujpujyuTpuT

c

L

U

∈∈⊂∈∀−−+−≤

Γ≤−−−−

,,,1,,11,

,,1,,11,

00

000

(18)

( ) ( ) ( ) ( )

UuPpJJjpujyTpujpujyuT

c

UL

∈∈⊂∈∀−≤Γ≤−

,,,1,,,,1,, 000

(19)

Constraints (16) – (19) constitute the linearised version of constraint (3). The advantage of this linearization technique is that it is exact, which implies that global optimality is assured. The disadvantage, however, is that it requires the introduction of new variables and additional constraints. Consequently, the size of the model is increased. A similar type of linearization is also necessary for constraint (4) in order to have an overall MILP model which can be solved exactly to yield a globally optimal solution.

Literature example Figure 2 is the representation of the case study that was used to demonstrate the performance of the proposed model it is taken from literature [13]. Whilst direct heat integration resulted in 18.5% improvement in terms of profit maximization, use of heat storage showed more than 25% improvement.

Figure 2 Process flowsheet for the case study

161

Conclusions A mathematical approach for optimization of energy use in heat integrated multipurpose batch plants has been presented and tested in a literature example. The results have shown that heat storage certainly results in more energy savings than direct heat integration.

References [1] R. Adonyi, J. Romero, L. Puigjaner, F. Friedler, Incorporating heat integration in

batch process scheduling, Appl. Therm. Eng. 23 (14) (2003) 1743–1762. [2] J. Corominas, A. Espuna, L. Puigjaner, A new look at energy integration in

multiproduct batch processes, Comput. Chem. Eng. 17S (1993) S15–S20. [3] B. Ivanov, K. Peneva, N. Bancheva, Heat integration of batch vessels at fixed

time interval. I. Schemes with recycling main fluids, Hung. J. Ind. Chem. 20 (1992) 225–231.

[4] I.C. Kemp, A.W. Deakin, The cascade analysis for energy and process integration of batch processes. Part 1. Calculation of energy targets, Chem. Eng. Res. Des. 67 (1989) 495–509.

[5] I.C. Kemp, E.K. MacDonald, Energy and process integration in continuous and batch processes, Inst. Chem. Eng. Symp. Ser. 105 (1987) 185–200.

[6] I.C. Kemp, E.K. MacDonald, Application of pinch technology to separation, reaction and batch processes, Inst. Chem. Eng. Symp. Ser. 109 (1988) 239–257.

[7] E. Kondili, C.C. Pantelides, R.W.H. Sargent, A general algorithm for short-term scheduling of batch operations. Part I-MILP formulation, Comput. Chem. Eng. 17 (1993) 211–227.

[8] B. Linnhoff, D.R. Mason, I. Wardle, Understanding heat exchanger networks, Comput. Chem. Eng. 3 (1979) 295–302.

[9] B. Linnhoff, G.J. Ashton, E.D.A. Obeng, Process integration of batch processes, Inst. Chem. Eng. Symp. Ser. 109 (1988) 221–237.

[10] T. Majozi, X.X. Zhu, A novel continuous time MILP formulation for multipurpose batch plants. 1. Short-term scheduling, Ind. Eng. Chem. Res. 40 (25) (2001) 5935–5949.

[11] D. Mignon, J. Hermia, Using BATCHES for modeling and optimizing the brewhouses of an industrial brewery, Comput. Chem. Eng. 17S (1993) S51–S56.

[12] E.D.A. Obeng, G.J. Ashton, On pinch technology based procedures for the design of batch processes, Chem. Eng. Res. Des. 6 (1988) 255–259.

[13] L.G. Papageorgiou, N. Shah, C.C. Pantelides, Optimal scheduling of heat-integrated multipurpose plants, Ind. Eng. Chem. Res. 33 (1994) 3168–3186.

[14] K. Peneva, B. Ivanov, N. Bancheva, Heat integration of batch vessels at fixed time interval. II. Schemes with intermediate heating and cooling agents, Hung. J. Ind. Chem. 20 (1992) 233–239.

[15] T. Pinto, A.Q. Novais, A.P.F.D. Barbosa-Po´voa, Optimal design of heat-integrated multipurpose batch facilities with economic savings in utilities: a mixed integer mathematical formulation, Ann. Oper. Res. 120 (1-4) (2003) 201–230.

[16] G. Schilling, C.C. Pantelides, A simple continuous-time process scheduling formulation and a novel solution algorithm, Comput. Chem. Eng. 20 (1996) S1221–S1226.

162

[17] N. Vaklieva-Bancheva, B.B. Ivanov, N. Shah, C.C. Pantelides, Heat exchanger network design for multipurpose batch plants, Comput. Chem. Eng. 20 (8) (1996) 989–1001.

[18] J.A. Vaselenak, I.E. Grossmann, A.W. Westerberg, Heat integration in batch processing, Ind. Eng. Chem. Process Des. Dev. 25 (1986) 357–366.

[19] Majozi, T., Heat integration of multipurpose batch plants using a continuous-time framework, Appl. Therm. Eng., 26 (2006), 1369 – 1377.


163

Early Stage Innovative Technologies for the Treatment of Wet Biomass and Organic Wastes: the HTU® Process and Supercritical

Water Gasification Jan Zeevalkink, Jaap Koppejan, Wilfrid Hesseling

TNO, PO Box 342 , 7300 AH, the Netherlands, www.tno.nl, email: [email protected]

Summary Many biomass feedstocks, especially organic wastes, contain considerable amounts of water. Examples are roadside grass, the organic fraction of household waste and sludge from waste water treatment facilities, organic residues from agro and food industries. At TNO several processes are under development aiming at an efficient use of the energy content of the wet biomass fractions. Examples are hydrothermal liquefaction for the production of a diesel fuel; supercritical gasification to produce an energy rich gas for electricity production. In this study the performance these processes is compared with that of existing technologies such as drying followed by combustion and anaerobic digestion. The comparison is based on data obtained from the Eminent software tool. The Eminent software tool is developed in two European projects for the assessment and promotion of early stage technologies.

Introduction In Europe large amounts of wet biomass or organic wastes are available for the production of renewable energy or energy carriers. Wet biomass is defined here as having a water content of more than 50 %. Traditional methods for the processing of these biomass feedstock are using only a part of the organics contained in the biomass (i.e. anaerobic digestion) or could have a low energy efficiency due to water evaporation (combustion). In this paper a few novel technologies are described that are especially suitable to process these resources.

Availability of wet biomass Huge amounts of wet biomass are available in Europe at a low price level, i.e. below 2€/GJ [3]. The present amount is estimated at 200 million t (dm, dry matter). An example is olive waste of which 5 million t is annually produced. Worldwide, the production of wet biomass is estimated at 5 billion t (dm) of which bagasse is an example with 100 million t (dm) annually. In the Netherlands, the potentially available biomass is estimated at 5 to 12 million t/a (dm). With an energy efficiency of 65 %, the recovered energy could amount to 70 to 166 PJ being 2 to 5 % of the Dutch energy consumption. Important biomass flows contributing to this amount are (for the Netherlands):

- the organic fraction of household waste (further mentioned OHW), separately collected from the residual “grey” waste: 1.5 M t/a (50% dry matter)

- road side grass: 600 kt/a with 50 % dm - residual organic wastes from the agro- and food industry such as pent brewery

grain and sugar beet pulp are produced on a large scale in the Netherlands. The production is estimated at 4.8 Mt/a dm with an average dry matter content of 10 %.

164

For the time being most residues from the agro- and food industry are used as feed for livestock. The other fractions mentioned can be of particular interest if their disposal or recovery is a cost factor. For instance, the composting of OHW and road side grass costs respectively approximately 30 and 40 €/t while the resulting compost is sold at low prices. These feedstocks can be attractive as a feed for novel technologies as they can be acquired with this cost as a gate fee. However, often pretreatment is required before these organic wastes can be processed. For instance sand removal and milling can be necessary if the technology requires a pumpable slurry as feed as is the case for the ESTs described in this paper. Options for the future are aquatic biomass (e.g. algae cultures or residues from their harvesting and processing) and energy crops. Their growth results obviously in higher prices above 2€/GJ.

Feedstock As an example of wet biomass, organic household waste OHW, separately collected, is used in this study. The alternative processes are compared by exploring their efficiency in the production of electricity from this feed. For the feedstock the following composition is assumed [12]:

- water content 50 % - ash content 30 % of dm - organics 40 % dm basis;

o of which: volatile organics 50 %, digestible for biogas formation - Typical mass flow 35.000 t/a dry matter - Low Heating value 11,3 MJ/kg dm

Currently, in the Netherlands, most organic household waste is composted. A small fraction is anaerobically digested. Processing cost amount to approx. 75 €/t a.r. with little revenues. These costs contribute to the attractiveness of OHW as feedstock. The economic scale of for instance a full scale HTU® plant or a large scale biomass combustion installation can be larger than 35.000 t/a organic material. In this case it is supposed that other organic feedstocks are fed into the installation supplementary to the OHW to ensure that the installations considered will operate economically.

Description of key early stage technologies The central technologies studied in this assessment of technologies are:

- the HTU® process - supercritical water gasification (SWG).

These processes are described in more detail below. HTU® and SWG have the advantage that the wet biomass can be processed without drying, as slurry containing 10-20 % dry solids. Their performance is described with regard to the processing of wet biomass flows and the recovery of the energy contained in the biomass. The options mentioned above are not the only innovative technologies for processing wet biomass. For instance, at TNO the production of ethanol from organic household waste is explored and appeared to be technically feasible. The HTU and SWG performance is compared with the results achieved when wet biomass is processed with mature technology. Reference processes selected are the following technologies:

165

- combustion - anaerobic digestion.

The processing on an industrial scale of 35.000 t/a dm organic household waste separately collected at the source is used to illustrate the perspective of these technologies. The emphasis in the comparison of the different processing routes will be on energy efficiency and costs. The HTU® technology The HTU® process is a process in which several types of biomass can be directly converted to a heavy crude-type of oil, similar in composition and characteristics to natural crude oil but with somewhat higher oxygen contents. In principle it could therefore play an interesting role in the substitution of conventional crude oil. After hydrogenation, the oil produced can be transformed into diesel or another commercially available transportation fuels, using conventional refinery technologies. A process layout scheme is provided below (Figure 1). In the process, wet biomass of typically 80% moisture content is pressurized to 120-200 bar and heated to about 300-350 °C where it remains for 5-15 minutes. Under these conditions, the biomass releases CO2 with the formation of a liquid product that resembles the atmospheric residue of crude oil, that separates from water and that has a relatively high heat of combustion. The thermal efficiency, defined as the Lower Heating Value of the product relative to that of the biomass feed (dry basis), is 75-85%, depending on the process configuration [2]. The HTU® process has no significant emissions. The off-gas is almost pure CO2 (for potential recycle). Within the process, the dissolved organics are eliminated from the water effluent by anaerobic treatment followed by further purification and the minerals are concentrated for potential recycle.

PT P R SFeed

80 260

Gas

330

WaterRecycle water

Biocrude

PT P R SFeed

80 260

Gas

330

WaterRecycle water

Biocrude

Figure 1 Simplified scheme of the HTU® process

(furnace and waste water treatment omitted) PT = pretreatment, recycle water (only if water content of feedstock is low).

P = pump; R =reactor; S =separator; figures: temperatures in oC

166

Figure 2 Overview of pilot plant

At TNO pilot plant research (Figure 2) has been ongoing for approximately 10 years [1, 2, 3]. A very large variety of feedstocks has been tested in HTU® autoclave and pilot plant experiments. They include wood, waste streams from sugar, potato and other food industries, grass, olive waste, organic domestic waste and fractions thereof, peat, digestate from anaerobic digestion, frying fat, and slaughterhouse waste. Figure 1 presents a simplified process scheme. The biocrude product is an organic substance that it is a solid at room temperature and becomes a liquid at about 80oC. It contains 12-20%w (DAF) of oxygen. The atomic H/C ratio is generally 1.0–1.3 and the average molecular weight is around 400. The nitrogen and sulfur contents depend on those of the feedstock. The LHV (lower heating value) of the biocrude is 30-35 MJ/kg DAF.

The raw biocrude can be separated into a light and a heavy biocrude by either flashing or solvent extraction. The light biocrude (LCR) is minerals-free. It can be used for high-efficiency electricity production. However, for large-scale applications it is preferred to upgrade the LCR by catalytic hydrodeoxygenation (HDO) to produce a premium gasoil and other high-quality fuel fractions. The heavy biocrude (HCR) is a coal-like solid. It can be co-combusted in a coal-fired power station to raise green electricity. Figure 3 gives the result of an economic analysis for a commercial-size HTU® process. HTU® can compete with a premium diesel made from petroleum

167

02468

10121416

-4 -2 0 2 4Biomass feed price, Euro/GJ

Cost of C5+product package,Euro/GJ

Future Plant

First Plant

C5+ package from crude oil @ 50 $/bbl (*)

C5+ package from crude oil @ 30 $/bbl

02468

10121416

-4 -2 0 2 4Biomass feed price, Euro/GJ

Cost of C5+product package,Euro/GJ

Future Plant

First Plant

C5+ package from crude oil @ 50 $/bbl (*)

C5+ package from crude oil @ 30 $/bbl

Figure 3: Cost of HTU® plus HDO HTU capacity 200 kt/a db, 20%dm

(*) or 30$/bbl with tax reduction of 15 Euro cents/litre

at crude prices between 30 and 50 $/barrel when the biomass cost ranges from –1 to +2 Euro per GJ. No subsidy, tax reductions or CO2 certificates were considered in this evaluation. From [2] the costs of OHW processing are estimated at approx. € 30/t a.r. It is assume that the feedstock needs a pre treatment process to produce a slurry that can be pumped to high pressure. This is described in the next section sub supercritical water gasification. Supercritical Water gasification SWG is the conversion of an organic compound in supercritical water (Figure 4). At approximately 600 °C and 300 bar, organic compounds are converted into a gas containing hydrogen, methane and carbon dioxide. The product gas is produced at a pressure of 300 bar. The exact gas composition depends on feed composition and process conditions. This opens opportunities for the efficient upgrading of the product gas into pure high-pressure hydrogen and separation of carbon dioxide. Hydrogen for transportation purposes in a fuel cell has to be stored at 300 to 700 bar. The pressurizing of hydrogen from atmospheric to 700 bar consumes approximately a third of its energy content> This is largely superfluous with SWG that delivers hydrogen at high pressure. Expansion turbines could be used for power production if the product gas is used at low pressures. The composition of the produced gas depends on reaction temperature and shifts towards more hydrogen when the reaction temperature is raised from 400°C to 700 °C. The calculated energy efficiency is approx. 75 %, i.e. 75 % of the high heating value of the feed and the external heat input is recovered in the product gas. For the long term, because of its high hydrogen concentration and the high pressure the produced gas can be attractive for the production of hydrogen for transportation purposes in combination with fuel cells,. On short term, direct use of the gas, after some cleaning, is for heat and/or power production is more likely. From [9] the costs of OHW processing are estimated at € 36/t a.r.

168

Figure 4. Process scheme of SWG pilot plant

TNO has conducted research [9] on pre treatment of wet biomass by washing for sand removal followed by milling for size reduction to produce pumpable slurry for the gasification process The energy consumption of the pre treatment process is estimated in the order of 10 kWh/t. The costs of the slurry production are dependent on the nature of the feedstock, and are estimated for OHW at 25 €/t [13]. The same pre treatment is assumed to be necessary for the HTU® process. Reference technology For comparison, the performance of the ESTs is compared to the results of two mature reference technologies:

- combustion in a dedicated biomass fired power station biomass and - anaerobic digestion with a gas motor to produce electricity.

Anaerobic digestion is an established technology for the treatment of wastes and wastewater. The final product is biogas: a mixture of methane (55-75 vol%) and carbon dioxide (25-45 vol%) [4, 7]. Anaerobic digestion converts appprox. 50 % of the organic solids in OHW into biogas. The assumed biogas production is 110 Nm3/t ar, to be combusted in a gas motor with an electric efficiency of 30 %. For the direct combustion of OHW, it is supposed that it is (co)combusted in a dedicated biomass plant with a thermal efficiency of 90 %, followed by a power generation with an electric efficiency of 25 %.

Process chains The appreciation of a process obviously strongly depends on the goal one wants to achieve. If diesel is required the HTU® process will be preferred; if fuel cells have to be fueled and hydrogen is required, the SWG process has an advantage. In this study the selected final product is electricity to get a comparison on the same basis. The biocrude product of HTU® is assumed to be co combusted in coal-fired power stations with an electric efficiency of 40 %. The energy rich gas produced by Supercritical Gasification is assumed to be combusted in a gas engine with an efficiency of 30 %; a value selected because of the low LHV of this gas at atmospheric pressure.

169

Most of the processes mentioned above cannot operate without some pre treatment of the feedstock. In Figure 3 complete processing chains are depicted. Washing means the removal of sand from the organic fraction in a water volume by gravity separation. Milling includes particle size reduction until 0.5 - 1 cm in order to obtain pumpable slurries. Based on the data that is mentioned in the paragraphs above the amounts of electricity that can be produced with the use of the described processing chains are calculated

Table 1 Novel and mature processing chains for wet biomass conversion

Short name Step 1 2 3 4 Supercritical gasification SWG

Washing Wet milling SWG conversion

Gas combustion for power production

Hydrothermal liquefaction HTU®

Washing Wet milling HTU® conversion

Co combustion for power production

Anaerobic digestion

Milling Anaerobic digestion

Gas engine for power production

Combustion Combustion

Power generation

The Eminent software tool EMINENT (Early Market Introduction of New ENergy Technologies) is a European Union sponsored Project in which a software tool has been developed with an extended database on energy resource, energy demand and energy technology data per country. [11] One of the objectives of the EMINENT project is to evaluate early stage energy technologies. The first EMINENT project was launched by the EC DG TREN in 2003. This three-year-long project has achieved substantial results in all its key areas. The second EMINENT project has been granted for continuation until December 2008. The main objective of both initiatives is to identify and accelerate introduction and implementation of leading edge European energy and environmental technology into the market place in Europe and worldwide. A software tool was developed, which is capable to analyze the potential impact of new, yet underdeveloped energy technologies (early stage technologies, ESTs) in different sectors of society over different countries. Its advanced internet based version enable users to continuously exchange and work with the most recent data. Along with the EMINENT tool an extensive database compatible with the tool has been developed which serves for storing the data both on new technologies and sectoral energy supplies and demands. The tool will not become available for other users before 2009. Evaluation of EST is one of the objectives of the EMINENT project. This paper presents an example of the use of EMINENT in the evaluation of an EST.

170

Results and discussion Table 2 Energy outputs and products

Intermediate Electricity Product Cost

indication MW MWe M€/a

SWG 10.5 gas 3.1 hydrogen, methane 4.3*

HTU® 10.5 crude 4.1 liquid fuel for transportation 3.9*

Digestion 6.6 gas 2.0 SNG 2.5* Combustion 10.1 heat 2.8*** 5.6** *excl. power generation cost **incl. power generation cost ***reduces to 1.4 MW with 75 % water in the feedstock

Table 2 presents the energy outputs of the defined processing chains. In the column intermediate the energy content of intermediate products is given: biocrude for HTU®, a H2/CH4 mixture for SWG, biogas for anaerobic digestion and heat for combustion. The energy efficiency of the HTU and the SWG process and that of anaerobic digestion do not depend on the water content of the feedstock. The presently selected feedstock has a water content of 50 %, which is according to the definition in this study is the upper limit for “wet” biomass. The electricity production of combustion is also calculated in case the water content of the OHW would be 75 %. Then the electricity production capacity is reduced to 1.4 MWe. The quality of the cost data is not very high as they are from different sources and their basis is not uniformly and clearly defined.

HTU® and SWG produce the same intermediate energy low due to the same conversion efficiency but differ in the produced power because of the difference in efficiency assumed in the power generation. Anaerobic digestion clearly produces less energy due to the fact that only approx. 50 % of the organics is converted. The power produced will probably be considerably les than indicated here due to the relative high internal energy consumption and to the fact that the product is wet stabilized material that has to be dried to get the final product: dry compost.

Conclusions The novel ESTs described, HTU® and SWG, perform better than the existing mature technologies anaerobic digestion and combustion. However the power production of the combustion technology is for a feedstock containing 50 % water not inferior. Combustion produces a considerable amount of power despite of the high water content. It should be taken into account that many wet biomass/organic wastes have a water content that is considerably higher. The difference, and the advantage of the novel technology, is more evident when the feed has higher water content as is shown if the water content would be 75 %. This underpins the definition giving in the beginning of this paper that wet biomass is defined to have more than 50 % water. The cost comparison for electricity production obscures somewhat the insight in the strengths of the technologies described here. SWG would have a higher conversion factor to electricity if it was assumed that the produced gas could be co combusted in power station or if fuel cells could be applied. Anaerobic digestion results in compost,

171

HTU® in an alternative liquid fuel, SWG in an energy rich gas or hydrogen or methane. It also depends on the anticipated applications of the product which conversion technology is to be preferred.

Acknowledgements The financial support from the EC project EMINENT2 – TREN/05/FP6EN/ S07.56209/019886 is gratefully acknowledged.

Abbrevations ar as received dm dry matter HTU® Hydrothermal Upgrading LHV low Heating Value OHW Organic fraction of household waste, separately collected SWG Supercritical Water Gasification

Publications and relevant literature [1] Naber JE, Goudriaan F, Zeevalkink JA, Berends RH. The HTU® process is ready for commercial demonstration, TNO, 2004. Paper presented at the European Biomass Conference, Rome, 2003. [2] Goudriaan F, van de Beld B, Boerefijn FR, Bos GM, Naber JE, van der Wal S, and Zeevalkink JA. Thermal efficiency of the HTU® Process for Biomass Liquefaction. Paper presented at conference “Progress in Thermochemical Biomass Conversion”, Tyrol, Austria, 18-21 September 2000 [3] Naber JE, Goudriaan F, Zeevalkink JA. Conversion of biomass residues to transportation fuels with the HTU® Process Paper presented at the 14th European Biomasss Conference in Paris, 2005 [4] Reith JH, Wijffels RH, Barten H. Bio-methane & bio-hydrogen; status and perspectives of biological methane and hydrogen production. ISBN: 90-9017165-7, 2003 [5] Penninger JML. Hydrogen from Supercritical Water by in-situ oxygen transfer to organic substrates presented at ECOS 2000, University Twente, The Netherlands, July 5-7, 2000, Proceedings Part 2:1181-1193 [6] Penninger JML, Rep M. Hydrogen-rich fuel gas from wet biomass by gasification in supercritical water. Proceedings International Hydrogen Energy Conference, July 2005, Istanbul, Turkey [7] Vandevivere P, De Baere L, de and Verstraete L. Types of anaerobic digesters for solid wastes in: "Biomethanization of the Organic Fraction of Municipal Solid Wastes", Editor(s): Mata-Alvarez J. September 2002 . ISBN: 1900222140 IWA Publishing. [8] Penninger JML, Wagenaar BM, Beld L, van de and Assink D. SWS process for production of hydrogen integrated with generation of clean energy. Proceedings First European Hydrogen Energy Conference, Sept. 2003 Grenoble [9] Zeevalkink JA, Meddeler B. Pre treatment of organics to produce a pumpable for the Super critical water gasification process, (TNO study 2006) in SuperHydrogen final report of EC project ENK6-CT-2001-00555 (BTG project coordinator) [10] Zundert E van, Doorn J van, Verhoeff F, Tienen T van. Biomass drying with (residual) heat (in Dutch) 2003, www.senternovem.nl (accessed 22/04/2008) [11] Zeevalkink J, Koppejan J, Hesseling W. Promotion and assessment of early stage technologies by the Eminent Initiative. Proceedings of the 10th conference on

172

process integration, modeling and optimisation for energy saving and pollution reduction 2007:12:703-708 [12] Environmental Impact Report Dutch Waste Policy Plan: Background document 14 on Organic Household Waste (in Dutch) 2002, www.senternovem.nl (accessed 22/04/2008) [13] Penninger JLM, (Sparqle International), private communication 2/03/2008


173

Comparison between EMINENT and other Energy Technology Assessment Tools

Raquel Segurado, Sandrina Pereira, Ana Pipio and Luís Alves

Department of Mechanical Engineering - Instituto Superior Técnico Pavilhão Mecânica I – 2º. Andar

Avenida Rovisco Pais 1049-001 Lisboa


Abstract This paper presents a comparative study between EMINENT and other tools for energy technology assessment. The EMINENT tool is an assessment model with two databases regarding energy data (resource and demand; and technologies – commercial and early stage). It is a software tool to evaluate the performance and market potential of early stage technologies (EST) in a pre-defined energy chain, under national conditions, in terms of financial, energy and environmental criteria. In addition, the software can suggest an improved energy chain with given availability of resources and energy demand patterns, by rearranging automatically the chain or with the introduction of other novel technologies available in a technology database. Among the tools analysed, CO2DB is thought to be the most similar to EMINENT, but focused on carbon mitigation technologies. The main difference is the non existence of resources and demand databases in CO2DB, that can be a barrier for the end-user since it needs to provide all the energy data to make the technology assessment. MARKAL does not aim to assess the technology per se, but to optimize all energy chain. This tool describes all possible flows of energy from well-to-wheel, and selects the set of technology options that minimize the total system costs. While EMINENT makes an environmental analyses, in MARKAL the end-user needs to force the environmental boundaries. IKARUS, optimizes the energy supply chain for minimum overall system costs, but this analysis is restricted to Germany. IKARUS also forecasts future energy demand in this country. E3Database is a tool to provide policy-makers information on energy chains and pathways. This tool evaluates energy supply chains, from primary energy to end-use, and compares to reference case, while EMINENT is focused on technology. Although there are many different energy technology assessment tools, EMINENT seems to be the only one targeting EST.

Keywords: early stage technologies, energy technologies, assessment tools, energy models, EMINENT

Introduction One of the questions that it is necessary to answer in order to show the relevance of this paper is: “Why do we need models and tools to make an assessment of an energy technology or an energy solution?” The answer is easy; the world lives a

174

moment where the issues related with climate change, security of energy supply, energy efficiency and sustainability have an important role. The energy technology assessment tools can help in the design of the best energy solution, in the selection of the more efficient and less pollutant technologies, for each reality (country, region, etc). The European Union (EU) is not an exception to this fact and is committed in supporting the development of low carbon energy technologies, in improving the use of renewable energy and increasing the energy efficiency, to reach these global objectives of sustainability, competitiveness and security in the energy scope. Ambitious goals were launched in 2007, namely the intention to cut in 20% the European Greenhouse Gas (GHG) emissions, to reach 20% of energy use through renewables and the reduction of 20% of energy consumption by 2020. Following these notions the European Commission (EC) presented the European Strategic Energy Technology Plan (SET-Plan), where it states: “Low carbon technologies will play a vital role in reaching our energy and climate change targets. The main goal of the SET-Plan is to accelerate the development and implementation of these technologies.” In this plan, the key EU technology challenges for the next 10 years are illustrated, in order to meet the 2020 targets and the 2050 vision (to reduce the GHG emissions by 60-80% by 2050). These new European policies show the relevance of the climate change and the energy issues for Europe, the weight of the development and exploitation of new technologies and, finally the importance of Research and Development (R&D) in this scope. This paper begins with the introduction of the theme – Energy Assessment Models, defining it briefly, and describing the several classification options of energy models based in its features. The EMINENT tool is described and framed in the previous classification. After, an analysis of potential competitive tool is made, and four of them are chosen – CO2DB, MARKAL, IKARUS and E3Database, and compared with EMINENT. Finally the EMINENT’s relevance is established among the new European policies and goals.

Energy Assessment Models A model is a mathematical representation of a certain system, therefore is always a simplification of reality. Additionally, any model that deals with future situations makes use of estimates and assumptions which may or may not be valid under certain circumstances, but will be inevitably uncertain. Energy models (models with focus on energy issues) can be classified according to many characteristics, the difficulty is that usually none of the models fits into only one category. There are many different ways to characterize energy models: its purpose, structure (internal and external assumptions), analytical approach (top-bottom vs. bottom-up), simulation vs. optimization techniques, mathematical approach, geographical coverage (global, regional, national, local, project), sectorial coverage, time horizon (short, medium or long term), data requirements. The classification most suited for the comparison analysis done in this paper is the analytical approach and the simulation vs. optimization techniques. Optimization models are prescriptive because they determine the best way to achieve a certain goal. On the other hand, the simulation models are descriptive, since they describe the functioning of the system.

175

Bottom-up models usually contain the description of specific energy related tasks, instead of energy demands, which are to be accomplished at minimum costs by a certain number of technologies. This type of models ask how can a given emission reduction task be accomplished at minimum cost. On the other hand, top-down models do not consider energy related tasks, but energy demand in the form of functions that depend, among other, on total or sectorial economic product and on energy prices. This type of models determine by how much does a given price movement changes energy demand or energy related carbon emissions.

EMINENT Tool EMINENT, an energy technology assessment tool, was developed to evaluate the performance and potential impact of the early stage technologies (EST), in a pre-defined energy supply chain, under national conditions, in terms of financial, energetic and environmental criteria, allowing a faster introduction of new energy technologies and new energy solutions in the market. The tool is composed by two databases – a database of national energy infrastructures, which contains information of the number of consumers per sector, type of demand, typical quality of the energy required and the consumption and installed capacity per end-user; and a database with the EST and commercial technologies. This database contains key information on new energy technologies that are currently under development. The energy technologies currently available and in use are also included, to enable the design of the most favourable energy chains. An EST can be selected from the database or a new EST can be introduced. The tool is able to assess a technology at financial, environmental and energetic level, comparing it with other technology that already exists in the market. The model developed in the tool is illustrated in Figure 1.

Figure 1 – EMINENT tool

The EMINENT tool is user-friendly and can be used by several types of entities, namely technology financers, technology developers and promoters, universities, research centres; etc.

176

With the enhancement and acceleration of the market introduction of EST, EMINENT provides relevant inputs to the achievement of some strategic and policy objectives of the EU, namely: the reduction of greenhouse gases and pollutant emissions (Kyoto), the strengthening of the security of energy supply, the improvement of the energy efficiency and the intensification of the renewable energy use, the enrichment of the competitiveness of European industry; etc. On the other hand, the EMINENT tool can facilitate developers, venture capitalists and governments to take decisions more effectively on further development of technologies that are yet immature, leading to more effective spending of R&D budgets. The EST assessment tool together with the EST database that has been developed for the evaluation of the technologies characteristics, in national context, provide a powerful tool for the target groups to perform fast analyses on new technologies. Regarding the classification of the EMINENT tool, it can be considered a simulation tool with a bottom-up approach.

Analysis of Potential Competitive Tools In the energy sector, computer models have been developed and have become standard tools for energy planning and optimization of energy systems which try to increase the share of renewable energy. These models are developed for different purposes, different technologies and focus on different sizes of energy systems. For example, the model HOMER is made particularly for small isolated power systems, although it allows for grid connection. The model includes most of the relevant technologies, but not all of them. Other models include precise physical details of specific technologies, like Hydrogems. Such models are often too detailed for energy planning purposes and often lack other relevant technologies. RETScreen software can be used to evaluate the energy production, life-cycle costs and greenhouse gas emission reductions for various types of proposed energy efficient and renewable energy technologies but it does not provides tools for joint energy balancing with different renewable energy sources. The model EnergyPLAN has been designed for national and regional analyses and it is a deterministic input/output simulation model. The H2RES model simulates the integration of renewable sources and hydrogen in the energy systems of islands or other isolated locations; it is based on hourly time series analysis of demand, storage and resources, but it does not perform market analysis. Most energy models are focused on energy planning and, when the technologies are analysed, they are integrated in the energy chain. The criteria used to choose the competitive models of EMINENT was based on the technology capacity assessment of the tools. The main tools considered to have some level of competition with EMINENT were CO2DB, MARKAL, IKARUS and E3database. These tools have a similar objective to EMINENT, since they can perform comparisons between technologies in the energy conversion chain, and that assessment is related with economic and environmental effects, namely CO2 emissions. Table 1 summarizes the comparison between all the tools considered.

177

Table 1 – Summary of the tools analysed

Resources Database

Demand Database

Technologies Database EST Cost

analysisEnvironmental

analysis Forecasting / Scenarios Optimization

EMINENT Yes Yes Yes Yes Yes Yes No No CO2DB No No Yes No Yes Yes No No

MARKAL No No Yes No Yes No No Yes IKARUS Yes No Yes No Yes Yes Yes Yes

E3Database No No Yes No Yes Yes No No

Description of the Select Tools

CO2DB The International Institute for Applied Systems Analysis (IIASA) developed CO2DB (Carbon Dioxide Technology Database) in order to assess options and measures for mitigating global CO2 emissions. The CO2DB tool is able to select technologies and calculate the efficiency of the energy conversion chains, and also estimate the environmental and economic costs. The database includes technical, economic and emissions data, labour and materials data, regional data, among others. The tool allows adding, organizing and comparing data with the technologies listed in the database. The user can also make energy chain calculations as well as comparisons. This tool assembles detailed data on carbon mitigation technologies in a standardized format. The technical, economic and environmental characteristics of these technologies, that now reach approximately 3,000, are meticulously described in a qualitative and quantitative way. In order to perform a comparative assessment of different energy technologies, the energy output and input have to be supplied for the chain calculations. In this way the tool is able to help with decision-making processes. The CO2DB tool carries out a macroeconomic analysis, as it performs an economic analysis of diverse technology chains from primary energy production to end-use. The IIASA disseminates CO2DB free of charge, in return, they request that users share the data they enter into the database. Among the tools investigated, CO2DB is thought to be the most similar to EMINENT tool, but focused on carbon mitigation technologies. The main difference is the non existence of resources and demand databases in CO2DB that, in the case of EMINENT, allows the end-user to assess the technology without the introduction of national energy profile.

MARKAL MARKAL (MARket ALlocation) uses the input data to calculate the evolution of a specific energy-environment system at the national, regional, or community level, over a period of usually 20 to 50 years. The software describes all possible flows of energy from resource extraction, through energy transformation and end-use. Each link in the chain is characterized by a set of technical, environmental and economic coefficients. This model selects the set of options that minimize the total system costs, finding the best renewable energy source, for each period of time. MARKAL enables the user to represent a complex energy system (local, regional, national or sectorial) as a linear program. The database indicates the energy demand, the available sources of supply of energy and the technologies. The time horizon is medium and long term, usually with 5 year periods.

178

The data requirements of this model include a base year energy balance and demand projections, this is needed in order to know the energy demand that will need to be satisfied in the next years or even decades. Usually, this analysis is done by sectors: industrial, commercial, household and transport. The available sources of energy supply can be either domestic resources or imports: oil, coal, natural gas, nuclear fuel and renewables. The database of technologies include extracting, transporting, converting and using energy technologies, both the existing ones and the ones that are expected to be available within the time horizon of the model. It is required to specify the characteristics of these technologies, for instance, the investment costs, the operating and maintenance costs, service life, fuel use, efficiency, availability, output, maximum expected market penetration and environmental coefficients. The model is a linear program that selects the best combination of these technologies to satisfy the projected energy demands. The solution to the linear program describes a set of energy technologies and energy flows that constitute an energy system that satisfies the inputted constraints (for example, the maximum allowed CO2 emissions) and minimizes the objective function (for instance, the total cost). MARKAL finds the least cost of technologies to satisfy end-use energy demands and user specified constraints, and calculates the resulting environmental emissions. MARKAL does not aim to assess the technology per se, but to optimize all energy chain. This tool describes all possible flows of energy from well-to-wheel, and selects the set of technology options that minimize the total system costs. While EMINENT makes an environmental analyses, in MARKAL the end-user needs to force the environmental boundaries.

IKARUS In 1991, the IKARUS project - Instruments for Greenhouse Reduction Strategies – was initiated. The objective was to supply tools to develop strategies to reduce energy related emission of greenhouse gases in Germany. In order to accomplish this aim, there was the need to establish a standardize database and models on which consistent energy scenarios and greenhouse reduction strategies could be formulated and calculated. The IKARUS software was developed in this project and consists of energy models and a database that is available as an information system, where is included detailed technology data, environmental data. IKARUS is composed by three energy models - a classical bottom-up energy optimization model (based on linear programming) that describes the energy system on a national level; a simulation model for different individual buildings as well as for a partial or an entire building stock; and a transport simulation model, for passenger and freight transport, where energy consumption, costs and emissions are broken down to travel purposes and detailed specific vehicles. The optimization energy model represents statically a linearized energy system of Germany in the reference years (1989, 2005 and 2020). It covers the energy flow from the primary energy sector to the energy end-use sectors, namely industry, transport, households. Links are given as energy sources, conversion and transport

179

of energy, and energy sinks. Conversion and transport losses arise throughout, starting from the provision of primary energy until the demand for useful energy has been satisfied, and the relationships between the energy flows are achieved by the techniques used, which are described by specific inputs per unit outputs. Cost and emission flows are modelled simultaneously. Based on a given energy demand, the model calculates a cost-minimizing solution for the energy system satisfying this demand. This includes the optimal energy technology structure as well as the optimal energy carrier combination. Restrictions can be imposed, for example, the maximum amount of CO2 emissions. There is also an economic model that calculates not only the direct investments induced by the reduction of CO2, but also the indirect effects through the investment goods producing sectors. IKARUS, as MARKAL, optimizes the energy supply chain for minimum overall system costs, but this analysis is restricted to Germany. IKARUS also forecasts future energy demand in this country.

E3Database The company LBST (Ludwig-Bolkow-Systemtechnik) developed E2database (Energy Emission Database), an energy and emissions balance tool that was able to calculate all possible energy chains from the energy source to end-user. The French Atomic Energy Commission (CEA), the Institut Français du Pétrole (IFP) and LBST jointly developed E3database (Energy Emission Economy Database). This tool is based on E2database, and intents the promotion of an economic evaluation of hydrogen energy chains in key energy sectors. This tool has been used mainly to model and calculate hydrogen energy chains, but it is not restricted to hydrogen and other supply chains also can be modelled. E3database calculates energy and material flows (input, output, emissions) as well as transport processes, and evaluates energy efficiencies and emissions of energy supply chains. It features a geographic approach, starting from a local, or regional, hydrogen demand, and it is able to evaluate all potential technological options that meet this demand. A fuel chain begins with the primary energy and ends with the final energy at the interface to consumption, or ends with the service provided for the energy consumer. Several processes can be combined to form energy chains by connecting corresponding inputs and outputs of the processes in the chain. The user can evaluate the regional hydrogen demand, select the required unit sizes of installations for hydrogen production and select the appropriate options for defining the hydrogen chains related to the regional study case. The calculation of costs requires the estimation of the levelized (annualized) costs for one unit of hydrogen produced, stored, delivered, dispensed or used. The economic calculation uses current state-of-the-art data for estimating the costs related to predefined installations, systems and devices, taking into account the costs reduction by technological progress, industrial innovation and learning-by-doing. This calculation also allows the comparison of the costs of hydrogen chains with the costs of a reference gasoline chain. E3Database is a tool that evaluates energy supply chains, from primary energy to end-use, and compares to reference case, while EMINENT is focused on technology.

180

Synopsis As the tools analysed reproduce the performance of all or a part of the energy chain, they can be classified as simulation tools. MARKAL and IKARUS can also accomplish the optimization of the energy system, providing the best solution based in the objective function. All the tools use a bottom-up approach, since they determine the energy supply mix in order to satisfy the energy demand. Although there are many different energy technology assessment tools, EMINENT seems to be the only one targeting EST. This is one of EMINENT’s strength, since it can provide a database with new and innovative energy technologies that are not available elsewhere and, on the other hand, it is able to assess in a fast way the market potential of these technologies. The EST assessed and identified has having market potential can be afterwards introduced in the technology database of the other tools. In this way it can be considered that EMINENT can be used to complement the other analysed tools in the evaluation and promotion of EST.

EMINENT and the new European challenges The climate change and the energy issues are a very important focus in the European Commission President agenda. In the speech “20 20 by 2020: Europe's Climate Change Opportunity”, the EC President enhanced the struggle against climate change and the quest for secure, sustainable and competitive energy for Europe. In this scope, he launched very ambitious goals, namely: • To reduce GHG emissions by 20% by 2020 (target that can be increased to 30%

in case of a global agreement); • In the longer term, to reduce GHG emission by 60-80% by 2050 (to meet a global

reduction of GEE of 50%). • To ensure 20% of renewable energy sources in the EU energy mix by 2020; • To reduce EU global primary energy use by 20% by 2020.

To reach the objectives presented it is necessary an enthusiastic policy to accelerate the development and implementation of cost-effectiveness low carbon technologies. Keeping this in mind, the EC has developed the SET-Plan, where are stated the key technology challenges for the next 10 year in order to meet the goals of 2020 and the vision of 2050. With this SET-Plan EC aims to deliver some important results: a new joint strategic planning, a more effective implementation, an increase in resources, and a new and reinforced approach to international cooperation. These new European policies show the relevance of the climate change and the energy issues for Europe, the weight of the development and exploitation of new technologies and, finally the importance of R&D in this scope. The investment in knowledge, R&D and innovation, by the EU is clear, as is stated in the communication from the EC “A European Strategic Energy Technology Plan (SET-Plan) - Towards a low carbon future” - “…In the longer term, new generations of technologies have to be developed through breakthroughs in research if we are to meet the greater ambition of reducing our greenhouse gas emissions by 60-80% by 2050.” In this sense the EMINENT tool is very useful due to its characteristics in the EST collection and assessment. The EMINENT tool can help in the identification and choice of the most efficient and the lower pollutants energy solutions, making an assessment for each solution and comparing the different solution, at energetic, environmental and financial level. On the other hand, the tool can provide some

181

technical insight on the Early Stage Energy Technologies and some figures related to its market potential in different countries. It is important to refer that the EMINENT tool is the only tool analysed that is focused in the EST assessment, in the technology innovation, is the only one based in technologies that are not in the market. The tool is able to analyse the market potential of a new technology, so it can be very useful in the analyses of the technology needs. In this sense, the tool will provide the missing link between recent R&D developments, thematic networks and European industries by the identification of recent technological innovations (EST) on energy and environment and the dissemination of key information in the market. The tool can contribute to an enhanced competitiveness of the European energy industry by bringing together technology developers, manufacturers and end-users. This presents key advantages in exporting European energy technology for developing countries.

Conclusions There are many Energy Assessment Models, and most of them are focused on energy planning and, when the technologies are analysed, they are integrated in the energy chain. The objective of the EMINENT tool is not to perform energy planning, but to analyse EST in order to evaluate its market potential and to accelerate its market introduction. Tools that carry out technology assessment were chosen as potential competitive to EMINENT – CO2DB, MARKAL, IKARUS and E3Database. All the tools analysed use a bottom-up approach and are simulation tools. MARKAL and IKARUS can also accomplish the optimization of the energy system. The main conclusion of this comparison study is that EMINENT is the only energy technology assessment tool that targets EST. This provides an advantage to EMINENT, as it is a tool where it is possible to find information regarding several new energy technologies that are not in the market. On the other hand, the EU is betting strongly in GHG reduction, security of energy supply, energy efficiency and renewables, giving special importance to the development of low carbon and more efficient energy technologies. In this scope, it was launched the SET-Plan, that concentrates a group of technological challenges to carry out until 2010 in order to accomplish the objectives of the EU to 2020 and 2050. As the EU is expected to invest in these research areas, there will be more research results that need to be promoted. EMINENT appears at the right time, because it can assess these results and accelerate its introduction in the market, filling the gap between the research and the market. In this sense, EMINENT can help the EU to have a leading role in the energy technology area. The EMINENT tool, besides having a database of EST that are low carbon technologies, can also support the EU in the assessment and selection of the technologies that can help accomplish the objectives outlined.

Acknowledgements The authors would like to thank the European Commission and its DG TREN for supporting the project EMINENT, that resulted in this work.

References [1] SET-Plan site, http://ec.europa.eu/energy/res/setplan/communication_2007_en.htm

182

[2] Hirematha RB, Shikhab S, Ravindranath NH. Decentralized energy planning: modeling and application – a review. Renewable and Sustainable Energy Reviews, 2007: 11: 29–752. [3] Schrattenholzer L. Energy Planning Methodologies and tools. IIASA, June 2005. Available on: www.iiasa.ac.at/Publications/Documents/RP-05-002.pdf [4] Assefa G, Frostell B. A systematic approach for energy technology assessment. www.iiasa.ac.at/Research/ECS/IEW2003/Abstracts/2003A%20_assefa.pdf [5] Nicole van Beeck. Classification of Energy Models, May 1999. Available on: http://arno.uvt.nl/show.cgi?fid=3901 [6] Lund H, Duić N, Krajačić G, da Graça Carvalho M. Two Sustainable Energy System Analysis Models: A comparison of methodologies and results. Energy, 2007: 32:948–954. [7] World Bank site on Tools for Assessment: Models and Databases, www.worldbank.org/html/fpd/em/power/EA/methods/tools.stm [8] COMMEND (community for energy environment and development) site about Modeling Software, www.energycommunity.org/default.asp?action=71 [9] IIASA (International Institute for Applied Systems Analysis) site on Databases and Software, www.iiasa.ac.at/Research/ECS/docs/data_index.html [10] Manfred Strubegger. CO2DB Software, Carbon Dioxide (Technology) Database, Users Manual, Version 3.0; April 2003; IIASA, Austria. [11] ETSAP (Energy Technology Systems Analysis Programme) site on MARKAL, http://www.etsap.org/Tools/MARKAL.htm [12] APPENDIX A. THE MARKAL MODEL. www.etsap.org/reports/markal-a4-a1.pdf [13] Shafiei E, Saboohi Y, Ghofrani MB. Model for R&D Planning in Energy Sector. International Engineering Management Conference, Canada, September 11-14, 2005. [14] Markewitz P, Kuckshinrichs W, Martinsen D, Hake J-F. IKARUS – A fundamental concept for national GHg-mitigation strategies, Energy Conversion and Management, 1996:37(6):777-782. [15] Martinsen D, Krey V, Markewitz P. Implications of high energy prices for energy system and emissions – the response from an energy model for Germany, Energy Policy, 2007:35:4504-4515. [16] Martinsen D, Krey V, Markewitz P, Vögele S. A New Dynamical Bottom-Up Energy Model for Germany Model Structure and Model Results. www.saee.ch/saee2004/Martinsen_dag.pdf [17] Krey V, Martinsen D, Markewitz P, Horn M, Matthes F, Graichen V, Harthan RO, Repenning J. Impacts of high energy prices on long-term energy-economic scenarios for Germany. www.risoe.dk/rispubl/reports/ris-r-1608_43-54.pdf [18] Agator JM. Decision Aiding Tool E3 Database for Energetic, Emissions-Related and Economic Regional Evaluation of Hydrogen Fuel Chains. 1st European Hydrogen Energy Conference, Grenoble, France, 3 September 2003.


183

Quick Progress in High-Efficiency Motors and Drives Opens New Markets and a Large Saving Potential

Norbert N. Vasen1, Alessandra Boffa2

1 ETA Renewable Energy, Piazza Savonarola 10, 50132 Firenze, Italy, email: [email protected] and [email protected]; fax: +39 055 573425

2 ABB S.p.A. – ABB Sace Division, V. L. Lama 33, 20099 Sesto S.G. (MI), Italy, email [email protected]; fax: +39 02 2414 3979

Abstract The work shows the quick progress in high efficiency electric motors and still more in drive technology. It points in particular to the many new potential applications, which are not followed by the awareness of the market. It will show the very significant impact on energy efficiency in industry and many other sectors if the new markets for drive technology were better addressed by energy saving initiatives. In fact, motors consume the major part (an average of 65%) of electrical energy in all sectors in industrial countries. Finally it is shown that already installed drives are not evenly distributed in the EU, so there is much to do in technology transfer and awareness initiatives.

Introduction The problem that is addressed in this paper is the large number of electrical motors used in the society and the inefficient use that is often made of it. The solution that is offered are high efficiency motors and frequency converters (from now referred to as “motors” and “drives”). These technologies have undergone such improvements during the last years, that they can help save much energy in an increasing range of applications. They offer not only energy and money savings, but also other benefits, that may attract even more interest of the users. The attention to electric motors originates from the high amount of energy that they consume. It is estimated that about 40% of the global electric power ends into industry. Of this part, over 65% is used to run electrical motors. However, also outside industry, many electrical motors are present.

High efficiency motors Today electric motors can be constructed with much higher efficiency levels than those of the past. This is thanks to a careful choice of materials and a design in which all the active parts are optimized. To encourage the use of high efficiency motors, different countries have adopted different kinds of legislative action. Probably the most efficient is that of the United

184

States where a federal law (Energy Policy and Conservation Act - EPAct) defines minimum efficiency standards and imposes fines on anyone failing to meet them. Europe, on the other hand, has decided to adopt a different approach. CEMEP (the European Committee of Rotating and Electronic Power Machine Constructors) has established, along with the European Community, a voluntary scheme that classifies motors into three different categories (Eff1, Eff2, Eff3) depending on power, the number of poles, and efficiency. This scheme is promoted to stimulate the use of motors with higher yield levels. Unfortunately, this initiative has obtained only a partial success: still in 2006, only 9% of the new installations involved the highest efficiency category Eff1. The main reason for this lies in the difference in the price of higher efficiency motors as well as in over-traditional purchasing habits that tend to favour a small discount today compared to far greater savings tomorrow. In fact, the cost of a high efficiency motor counts approximately for less than 2% of the costs of the energy that it will consume in its whole lifetime. In other words, in the average, the slightly bigger investment in the purchase of a higher efficiency motor is paid back in not more than one and two years. Improvement on the efficiency of the existing motors is even more conspicuous when the installed base is old and the motors have been rewound, resulting in a further efficiency loss. This often justifies also the substitution of an old motor even when it is still running.

Frequency converters The energy savings achieved by using frequency converters, also known as “electronic drives” or just “drives”, are particularly significant in fluid dynamic machines (centrifugal pumps and fans) especially when used with a variable flow rate. Fluid dynamic applications have a cubic energy-speed characteristic. This means that at 80% speed, only 50% of the energy is needed. So it is evident how much energy is wasted with the practice of dissipative regulation. It is much better to control these applications upstream, rather than downstream. An asynchronous electric motor without an drive is, by its own nature, destined to rotate at its nominal speed. If the electric motor is connected directly to the power supply and if a machine has to operate with a variable flow rate the plant must have a mechanical system (downstream of the motor) to adjust the flow rate, such as throttling valves, dampers or bypass systems, all of which are extremely wasteful. If, on the hand, an drive is used between the power supply and the motor (upstream), the flow rate is adjusted by directly reducing the speed of the motor. Applications such as pumps and fans are quite common in industrial applications, as well as in buildings and the potential savings ranging from 20 to 50 % and over. They show us how important it is to apply this technology to our installed base or new installations. Figure 1 will show the difference between upstream and downstream speed control, and the impact on energy efficiency. Here it is also essential to consider the great improvements that technology has provided to drives compared to the past.

185

Figure 1: The difference between upstream and downstream speed control

One of the most important developments in the frequency converters is that this technology becomes affordable also for smaller motor powers, which are probably the large part of all motors installed both inside and outside industry. Price decrease for these devices has been substantial in the past few years. This means that recently a very large number of motors would have good technical and economic reasons to be equipped with the new generation drives. As with all new developments, the market slowly follows the opportunity. Therefore, much should be done to promote the application of the electronic devices in energy wasting motors. There are a number of other benefits related with drives, which can also increase their market share. These include: • More comfort • Less noise • Longer lifetime of motors and the equipment that is driven by it • Improved production and quality through better control of the motor • Elimination of drive equipment (belts, valves, dampers, fluid couplings, gears, ...),

which will save capital, maintenance and energy. Some of these advantages are related to the drives themselves, others are related to the new developments in the electronic drive technology. To take a few examples of the market in Italy, it is estimated that only 8% of the 2 million ventilators or fans are equipped with the devices and that half of them would need them (and give a reasonable return time). In Northern Europe, these drives are installed on fans and pums often during their design and construction. Another development that can convince potential users of motor drives is the financial support in many countries, which is a response to the need to save energy

186

and the acknowledgement of the significant role of the electronic drive technology. The motor drive can be applied on all kinds of conventional motors but the use of modern and high efficiency motors will increase the savings and other benefits of the complete configuration. The task that lies before us is to bring the opportunities to the attention of the market and create mechanisms to apply the technology quickly to a large number of users. This is particularly urgent in countries with less experience in this field. In these countries, two main actions are necessary: retrofit on existing motors and promotion of the drives already during the design and construction of motor applications.

Technical considerations The electronic motor drives are presented under many different names: VFD (Variable Frequency Drive), VSD (Variale Speed Drive), VVVF (variable voltage variable frequency). The following picture shows the VSD in a typical configuration from user to motor.

Figure 2: Typical configuration of a drive in a user circuit.

Figure 3: Puls Width Modulation gives a smooth sinus.

187

The VSD in the central upper part of the figure receives grid power and instructions from the user and converts them into a power profile to the motor. The input power is converted to Direct Current (DC) that goes into small energy storage, and is converted from there to a power output with varying frequency and voltage or current. One method for adjusting the motor voltage is called pulse width modulation PWM. An example of PWM is given in Figure 3 [1] (from Wikipedia, “Pulse-width modulation”). The supply voltage (pulsed line) modulated as a series of pulses results in a sine-like flux density waveform in the magnetic circuit of the motor. The resultant waveform can be controlled by the varying width of the subsequent modulated impulses (per given cycle). The smoothness can be influenced by the frequency of these pulses, which is obviously (much) higher than the sine wave that has to be fed into the motor. Nowadays, usually Insulated Gate Bipolar Transistors (IGBTs) are used as the switches for this conversion. The latest development in drive control, which is specially followed by ABB, is the Direct Torque Control, which till recently was reserved for Direct Current (DC) motors. It allows to produce at all times a give torque, without need to add tachometric encoders to the motor, but just by analysing the current and voltage behaviour with microprocessors. The response time is much quicker than with AC or DC drives. Thanks to these developments, the motor can run on variable speed and torque. This variation can give benefits during start up, steady operation but also during deceleration. The benefits of control during transitional use (starting and stopping) are mainly the reduction of stress on the grid, the motor and everything that is connected to it. The control during steady state leads mainly to the benefits of energy saving and easiness in the system control. Other benefits, common to all circumstances are better production, phase correction, sensible noise reduction etc. A list of technical advantages for VSD in general is given below. It is followed by a list with advantages that are specific for the new generations of VSD, let’s say, from 2004. Advantages of drives in all cases (in the past and also recently). • Phase correction: the parameter cos phi of the user is optimal, which means that

current and voltage are in the same phase. This will avoid the technical problems and penalties related to reactive power.

• Energy supply with single phase or even Direct Current can be used to power normal 3 phase motors. This must be done with some care, derating the power of the VSD, so it can drive a smaller motor than specified;

• No high starting current. A normal motor at start takes 300% of nominal current, giving only 50% torque. With a VSD, 50% of current is absorbed and 150% of torque is available. This is because power (current x voltage or torque x speed) is better managed. These values can be programmed with the drive, in order to

188

comply with the requests of the application. It gives enormous flexibility in the use of motors and it will allow eliminating many other mechanical or hydraulic solutions that are used today.

• No mechanical stress to equipment connected to the motor (which the owner often cares much more about than about the motor).

• Increase the quality of the product by better control of the mechanical behaviour, reduce production times and reduce the cost of maintenance.

• Braking is possible, especially when a circuit is added to absorb the energy (this can be dissipative but or regenerative, usually the grid).

• More than one motor on one drive (but care must be taken for the power flow distribution among the various motors).

Innovations and its advantages (leading to new applications and new markets): • Better approach to EMC (more suitable to operation rooms, near computers, other

sensitive environments, less harm to health); • Less harmonics on the power lines of the factory or building, especially with

today’s low harmonic drives which can reduce the THD (Total Harmonic Distorsion) from above 30% to about 3 ÷ 4%. Moreover, when this is possible with a 6 pulse input stage, the system does not need to add dedicated trasformers;

• More flexibility of control (often free control software) and easier user interface (intuitive, similar to mobile phone);

• Availability of regenerative drives for wide power ranges, this allows to retrieve the energy from braking systems, hence obtaining further energy savings;

• Smaller size, more robust and easier installation • Higher MTBF values, so increased reliability; • Cost: modern VSDs are more affordable, also for smaller size (pay back

acceptable for power down to 0.75 kW, opening a large new market) • More energy saving, especially thanks to the new and faster hardware and

software allow even better motor flux optimization; • High torque, low rpm applications have been the domain of DC drives. This has

changed recently with the introduction of a new breed drive, the Direct Torque Control (DTC). This is also an advantage for flexibility of control, and adds a lot of new applications to the market of drives.

Some of the precautions to take with drives in general: • Normal motors that are designed for fixed-speed mains voltage operation are

often used, but certain enhancements to the standard motor designs offer higher reliability and better drive performance.

• A VSD or DTC is not the same as a soft starter, choose the latter if you need operation at only one speed. Otherwise an “overkill” is applied, leading to higher cost and sometimes to more complications.

• Take care of reduced ventilation of motors at lower speed; avoid continuous high torque under these cercumstances.

• Size the drive on max torque (monitor the current of the motor under worst conditions) rather than power. Ignoring this rule led to failure of many VSDs. Oversizing of the VSD may be required.

189

The list of innovations gives an important message to the market: drives are suitable for many more applications than in the past: • They cost less, and are therefore worthwile also for smaller motors. Small motors

exist in much larger numbers than large motors. This is true for industry but maybe even more in buildings, where they have functions in pumping and ventilation. These two applications (pumps and ventilators) have also the highest savings potential, due to their cubic power-flow characteristic.

• They are easier to use, therefore they will not hinder operation but give added value. This is indeed one of the non technical barriers and one of the strongest. The fear is that complicated operation will lead to problems and therefore to decreased functionality. With the easier installation and intuitive user interface, the impression will be rather the contrary: the users will discover the energetic and other benefits and how easily these are accessable. The main non energy related benefits are: better production through better control, more comfort (for example less noise by decreasing the speed when possible) and decreased stress on whatever is attached to the axis.

• They have higher reliability and will therefore contribute to the quality of the activity done by the motor, also in difficult conditions like voltage peaks, high currents or high temperature/humidity/dust. Moreover, they will have less impact on other equipment (decreased disturbance).

This message implies that a much wider market can benefit from VSDs. Moreover, it will lead to lower consumption of energy by the whole society. Therefore, stronger market mechanisms should be created, which promote the use of VSDs, where ever it can give benefits in terms of energy efficiency, but also better control of the application and thus quality/comfort.

Conclusions Electronic motor drives are since many years an important instrument in energy efficiency on electric motors, which consume a very large part of the total produced electric energy. The innovations in drives since a few years give an important message to the market: they are suitable for many more applications than in the past. • They cost less, and are therefore worthwile also for smaller motors. • They are easier to use, and will not hinder operation but give added value. • They have higher reliability and will contribute to the quality of the motors activity. This means that the opportunity exists to reduce the energy by most of the motor applications and to decrease the environmental impact and energy scarcity of nearly all sectors of society. Therefore, stronger market mechanisms should be created, which promote the use of drives, where ever it can give benefits in terms of energy efficiency, but also better control of the application and thus quality/comfort. The promotion should address retrofit of existing motor applications that are not regulated, or regulated in a dissipative way. Moreover, it should address the construction of machines that include motors, so that high efficiency motors and drives are already built into machines and systems at the

190

moment of their purchase. This latter promotion is much related to the energy certification of industrial machines and of buildings.

References [1] “Pulse-width modulation”. Wikipedia. en.wikipedia.org/wiki/Pulse_Width_Modulation


191

Environmentally Friendly Energy Supply in Wineries

Maximilian Lauer, Reinhard Padinger Joanneum Research, Institute of Energy Research, Elisabethstrasse 5, A-8010 Graz,

Austria, email: [email protected]; fax: +43 316 876 1320

Abstract The paper describes technical details and operational experiences of a combined heat, cold and power (CHCP) demonstration plant at the “Peitler” winery in Leutschach, Austria. The cooling demand of the winery is provided by a newly developed solar driven 10 kWc ammonia / water absorption chiller using heat from a thermosolar system and a wood chip boiler. The electrical power demand of the cooling system is produced by a 3 kWel Stirling engine, which is driven by the flue gas of a 50 kWth wood chip furnace. The ammonia / water absorption chiller was successfully installed 2003 and has operated maintenance-free for more than four years. The engine was installed together with a wood chip furnace three years later at the end of 2007 and has also been successfully tested on site. A period of two years operation and measurement has started in January 2008. The ammonia / water absorption chiller has proven market maturity and is already available on the market. The operating experiences with the 3 kWel biomass-driven Stirling engine demonstrate the technical maturity of the system design whereas further investigation should consider mass production issues in order to finally achieve market maturity. The annual costs of the system are approximately in the same range, as the costs of a comparable conventional compressor cooling system (without Stirling engine). The equivalent CO2 emissions of the absorption refrigeration system are only one fifth of a comparable conventional compressor cooling system.

Introduction Climate change is caused by CO2 and other greenhouse gases which mainly evolve from the combustion of fossil fuels for electrical power generation. A main factor for the world wide increase of electrical power demand is cooling. This leads to increasing electrical peak loads especially on hot summer days. In addition the electricity grid can only handle a certain amount of load at a time which can easily lead to power failure. To meet the problems of climate change and power distribution, new technologies for the decentralized combined generation of heat, cold and power (CHCP) with renewable resources like biomass and solar energy need to be developed. The demonstration project “Environmentally friendly energy supply in wineries” is carried out at a winery (“Peitler”) in Styria, that has demand of heating, cooling and electricity. In 2001 - 2003, in a first step a solar driven cooling system for the fermentation control of wine production as well as for the cooling and

192

dehumidification of the wine bottle storage unit was built. A solar driven 10 kWc ammonia / water absorption refrigeration machine was designed and a prototype was installed on site. This installation has been successfully tested for four years [1]. In parallel, a prototype 3 kWel Stirling engine, which was designed for the use with solid biomass, was constructed and built [2]. In 2006, the “Peitler” winery has become one of twelve CHCP demonstration sites of the EU project PolySMART (www.polysmart.org). In the frame of this project, the Stirling engine was implemented in the existing cooling system of the winery together with a 50 kWth wood chip boiler as a thermal source. Since then, CHCP is realized with heat from solar panels and from the wood chip boiler, cold from the absorption refrigeration machine, and electricity from the Stirling engine. Waste heat from the Stirling engine is used of course in the heating system for the building. The combination of these components within such a trigeneration system maximizes the degree of efficiency of the fuel input and reduces the CO2 emissions to a minimum by the usage of thermally driven cooling and solid biomass for electrical power production.

Cooling demand of wineries In the following chapters, the most important steps of the wine production process, in which cold is needed, will be described in order to give an overview on the implementation and use of the system.

Must cooling The need for must cooling depends on the environmental temperature during the harvest time. Conditions in Austria sometimes require must cooling, sometimes not. However in more southerly regions, must cooling might be very important for the production of high quality wine. By cooling the must, a separation of solid matter in the wine can be expected. Furthermore, a lower temperature of the must hampers the enzymatic activity, the oxidation reactions and the production of acetic acids, and therefore the special flavour of a specific wine will be preserved. Must cooling however needs high cooling power for a short time.

Temperature control of the fermentation process Temperature control of the fermentation process is very important for the quality of the wine. To get a high quality wine, the following requirements have to be respected: • Keeping the necessary starting temperature of the fermentation process:

For a well controlled start of the fermentation process, the temperature of the must should not be too low, in any case somewhere in the range between 17 and 20 °C. Below 15 °C, undesired yeast strains such as Kloeckera- and Candida-types can become active, which can hamper the desired yeast strains Sacharomyces cerevisiae.

• Temperature control of the fermentation process in a later stage: The fermentation temperature during the main stage of the fermentation process should not exceed a temperature range between 17 and 18 °C. Increased cooling demand is usually caused by exothermal chemical reactions of the fermentation

193

process. In special cases, specific short time temperature changes are desired. In the final stage of the fermentation process, an increasing of the fermentation temperature is beneficial.

• Deliberate inhibition of intermediate fermentation: By inhibiting the intermediate fermentation the winemaker can influence the sugar content of the wine. If the desired sugar content has been reached, the wine will be cooled in order to intermediately stop the fermentation process, and the yeast will be filtered. For carrying out this procedure effectively, a high cooling power for a short time is needed.

Stabilization of potassium bitartrate (“Winestone”) Stablization of potassium bitartrate is important in regions, where people usually drink very young wine as usually in Styria. These young wines are not stable in terms of stabilization of potassium bitartrate. It can be avoided by chemical additives, a better way however is a short shock cooling of the wine, before racking and bottling. Normally, the wine will be cooled down to a temperature between 0 and +2 °C for some 3 to 5 days. Also for this process high cooling power for a short time is needed.

Wine storage The optimal temperature for wine storage is some 10 to 14 °C. To keep this temperature in the summertime normally cooling is needed. However, due to the fact, that wine bottles, ready for selling, are normally already packaged in cardboard boxes with 6 bottles each, climate conditions in the storage should not fall below the dew point. If the temperature is too low, cardboard boxes become damp, loose their stability and become useless. Excessively high temperature however might be bad for the wine. At the surfaces of the cooling fins of the heat exchanger however, condensation is desired, because this will reduce the relative humidity in the room. This reduces possible destabilization of the cardboard boxes and other damages as for example mildew on the bottle labels. To avoid condensation in the room at a temperature level of 10 C, humidity should not exceed a level of some 7 g/kg dry air.

Combined Cold, Heat, and Power production (CHCP) at The “Peitler” Winery

General information about the “Peitler” winery The “Peitler” winery at Leutschach is one of the producers of high quality wines of southern Styria. The area of wine production is 5,5 ha. The yield of wine ranges, depending on the weather conditions, between 22.000 and 30.000 litres a year. Table 1 gives an overview on the fermentation capacity of the “Peitler” winery. The cooling demand of the “Peitler” winery is some 8.000 kWh per year.

194

Table 1: Overview about fermentation tanks and their capacity at the “Peitler” winery

Capacity of tank litres

Number of tanks Total capacity litres

4000 2 8000

3600 1 3600

2500 1 2500

2000 2 4000

1500 1 1500

1000 7 7000

750 2 1500

700 1 700

600 1 600

Total 18 29400

Overall design of the CHCP system A CHCP (combined Heat Cooling Power) generation system is relatively sustainable and environmentally sound, if possible harmful impacts on the environment are avoided or considerably reduced in comparison with other systems. Criteria for the evaluation of the sustainability are the “Global Warming Potential” (GWP), the “Ozone Depletion Potential” (ODP) and the “Indirect Global Warming Potential” (ID-GWP). Following a study [4], continuous absorption cooling with ammonia and water show the best results related to the above mentioned terms. By additionally producing the needed electricity, (or at least a part of it) using a biomass Stirling engine, the sustainability of the whole system can be once more increased. Figure 1 shows a scheme of the CHCP system at the “Peitler” winery

195

Figure 1: Scheme of the CHCP system at the “Peitler” winery In Figure 1 on the left side the wood chip furnace is shown, which produces the hot flue gas for the Stirling engine. After having passed the heat exchanger of the Stirling engine, the hot flue gas heats the boiler. Hot water from the boiler loads the hot water buffer tanks, together with the hot water from the solar collectors. Heat from the buffer tanks is transformed by the ammonia/water absorption refrigeration machine into cold. This is used in the wine storage and in the wine cellar. A cooling tower is needed for the absorption refrigeration process. Technical data of the components are: • Heat capacity of the wood chip furnace: 50 kWth • Power capacity of the Stirling engine: 3 kWel • Capacity of the hot water buffer storages: 4.800 l in total • Surface of the solar collectors: 100,8 m2 • Cooling power of the NH3/H2O absorption refrigeration machine, special

development of company Pink GmbH, Langenwang (A): 10 kWc

Ammonia / water absorption refrigeration chiller A market analysis concerning small absorption refrigeration machines has shown, that in this size range only products for natural gas as fuel are available. Therefore a new absorption refrigeration machine, suitable for the given boundary conditions, has been developed in cooperation with company Pink GmbH, Langenwang (A). A detailed description of this machine is given in [1].

Stirling engine The Stirling engine, a so called “Alpha Type Stirling”, is constructed on the basis of an industrial motorcycle crank mechanism. The pistons of the original engine are used as crossheads for the Stirling pistons, which are built on the top. Figure 2 shows the Stirling engine with the crank mechanism (1), the hot cylinder (2), the cold

196

cylinder (3), the heat exchange (4), the cooler (5), and the regenerator (6). The heat exchanger is open, so one can see the heat exchanger tubes.

Figure 2: 3 kWel Stirling engine, with crank mechanisem (1), hot cylinder (2), cold

cylinder (3), open heat exchange (4), cooler (5), and regenerator (6) The smooth pipes are aligned in flue gas direction in order to achieve a self cleaning effect. Ribbed pipes or pipes with fins would be clogged by combustion residues within a short time. The size of the engine is determined by the desired shaft power and thus by the heat exchanger surface area as well as by the maximal power transmission that can be achieved by the motorcycle crank mechanism. The heat exchanger has to be designed considering the smooth surface of the pipes. The surface area determines the heat input into the process and the size of the dead space. This leads to a disadvantageous ratio between expansion piston stroke volume and dead space. Certainly the piston stroke volume could be increased for more active volume, but this is limited by the maximum transferable power into the crank mechanism and by an increase of pressure drop within the heat exchanger. The final design takes this complex interrelationship into account and represents the optimal solution of biomass driven Stirling technology. Table 2 gives an overview on the main technical data of the Stirling engine.

Table 2: Principal technical data of the Stirling engine

Power capacity 3 kWel

Revolutions per minute 512 rpm

Working fluid air or nitrogen

Working pressure up to 32 bar

Temperature of heat input / output 1.000 °C / 750 °C

Electrical efficiency 23,5 %

1

2

4

3 5

6

197

The Stirling engine was installed at the already existing solar driven cooling system at the “Peitler” winery together with the wood chip furnace at the end of 2007.

Operational experiences and economical and ecological evaluation of the CHCP plant

The operation experiences with the CHCP plant at the “Peitler” winery are basically shown in Figure 3. It shows a comparison of the energy flow of a conventional compression cooling system (left) and the absorption refrigeration cooling system (right), both for 10 kWc cooling power.

Figure 3: Comparison of the energy flow of a conventional compression cooling

system (left) and the NH3 / H2O absorption refrigeration cooling system (right), both for 10 kWc cooling power

The comparison shows, that the conventional compression cooling system needs 4 kWel electrical power for 10 kWc cooling power, whereas the absorption refrigeration system only needs 0.25 kWel electrical power for the same cooling power. The heat demand and the heat flow however increase of course. This is not a problem, if the heat is produced from biomass and the waste heat can be used for room heating. An economical comparison of the two systems (without Stirling engine) is shown in Table 3. The investment cost of the absorption refrigeration system is 47.150,-- €. This is higher than the investment cost of 33.314,-- € of the conventional compression cooling system. However, mainly due to the lower electricity consumption, the absorption refrigeration cooling system has total annual cost of 4.738,-- €/a, slightly below the total annual cost of 4.758,-- of a conventional compression cooling system. The capital costs in this example have been calculated for a useful lifetime of 20 years for the absorption refrigeration system and 15 years for the conventional compression cooling system assuming an annual interest rate of 6,5 % for both systems. A very satisfactory result is the comparison of equivalent CO2 emissions of the two systems, which have been calculated based on a life cycle analysis. As shown in Figure 4, the absorption refrigeration system has equivalent

Cooling 10 kWc

Electricity 4 kW

Heat sink

Cooling 10 kWc

Heat sink

Heat 15 kWth

Electricity 0,25 kW

NH3 / H2O absorption refrigeration cooling system

Conventional compression cooling system

198

CO2 emission of some 1.200 kg/a, which is only one fifth of the equivalent CO2 emissions of the comparable conventional compression cooling system, which emit more than 6.000 kg/a. An economic evaluation of the Stirling engine cannot be made at the moment, because the investment cost and the maintenance cost of a future serial Stirling engine are no known and would be speculation at the moment. Table 3: Economical comparison of a conventional compressor cooling system with

the absorption refrigeration cooling system (without Stirling engine)

Absorbtion Compression Remark

A. Investment € €

1. Chiller 18.500,-- 12.914,-- inkl. Re-Inv.

2. Solar collectors, 50 m2 4.000,-- 0,-- 50 % f. cooling

3. Solar installation (50 %) 6.000,-- 0,--

4. Cooling installation 7.650,-- 8.900,--

5. Electonics and Control 11.000,-- 11.500,-- Visualisation

Subtotal Investment 47.150,-- 33.314,--

B. Operating costs € / a € / a

1. Capital costs 20 a / 15 a, 6,5 %

4.279,-- 3.023,--

2. Electricity 238,-- 1.349,--

3. Service 221,-- 386,--

Total, annual costs 4.738,-- 4.758,--

Conclusions A solar driven 10 kWc ammonia/water absorption chiller was successfully installed at the “Peitler” winery in Leutschach (A) and has operated maintenance-free for more than four years now. A 3 kWel Stirling engine was intergated at the end of 2007 together with a wood chip furnace and also successfully tested on site. A period of two years operation and measurement of the whole CHCP system was started in January 2008.

199

0

1000

2000

3000

4000

5000

6000

7000

Absorption Compression

[kg CO2- equivalent/a]

Solar collector

Biomass combustion

Electricity input

Conventional compression cooling system

NH3 / H2O absorption cooling system

Figure 4: Comparison of the equivalent CO2 emissions of the absorption refrigeration system without the Stirling engine(left) and a comparable conventional compression

cooling system (right) The ammonia/water absorption chiller has proved technical maturity and is already available on the market. The Stirling engine is a prototype and is not ready to be produced before further engineering. Preliminary operating experiences with the 3 kWel biomass driven Stirling prototype engine demonstrate the technical feasibility of the system design. An economic evaluation of the Stirling engine cannot be made at the moment, because the investment cost for future product is no known yet.

Acknowledgment The following partners mainly contributed to the success of this project.

Production of High Quality Wine

Steel Processing Company

Electronic Process Control Software & Hardware

Solar Installation & Design

Thürschweller Energy Systems & Consulting

Polygeneration with advanced Small and Medium scale thermally driven Air-conditioning and Refrigeration Technology

200

A similar contribution on the technology was presented by Dr. Reinhard Padinger at the 10th Energy Managers Days 2008 at the Grand Hotel Bernardin, Portoroz, Slovenia on 8th and 9th April 2008.

References [1] Podesser, E., Peitler J, Meißner E, Türschweller S. Enzinger P.: Einsatz von Sonnenenergie und Bioenergie zu Kühlung von vergärendem Weintraubensaft und zur Weinlagerraumkühlung, Report No: IEF-B-09/03, JOANNEUM RESEARCH, Institute of Energy Research, (2003) [2] Podesser, E., Dermouz, H., Padinger, R., Wenzel, A., Entwicklung eines mit Holz betriebenen Stirling-Kleinkraftwerkes zur dezentralen Strom- und Wärmeerzeugung – Phase II, Report No: IEF-B-12/95, JOANNEUM RESEARCH, Institute of Energy Research, (1996). [3] Weinbaukataster der weinbaubetreibenden Bundesländer, 1999 [4] Podesser, E: Umweltverträgliche Kälteerzeugung, Report No: IEF-B-6/94, JOANNEUM RESEARCH, Institute of Energy Research, (1994).


201

Raw materials for fermentative hydrogen production

Krzysztof Urbaniec, Robert Grabarczyk CERED Centre of Excellence, Warsaw University of Technology, Jachowicza 2/4, 09-

402 Plock, Poland, email: [email protected]; fax: +48 24 262 65 42

Abstract The basics of thermophilic hydrogen fermentation and photofermentation are outlined. Various types of biomass which can be used as raw materials for hydrogen fermentation are named and the methods of biomass pretreatment are briefly presented. Hydrogen yield from processing of four specific raw materials is estimated.

Introduction The European Commission's 1997 White Paper on renewable energy sources sets out the objective of increasing the share of renewable energies to 12% of gross inland energy consumption by 2010 [1]. Amongst renewable energy sources, the biggest contribution comes from biomass. It can be converted to energy carriers by thermal (combustion, pyrolysis and gasification) or biological (fermentation, digestion) processes [2, 3]. An interesting and promising method of biomass utilisation is hydrogen production by the fermentation process. Hydrogen fermentation is the topic of Integrated Project HYVOLUTION, financed under the Energy Priority of the 6th Framework Programme of the EU [4]. The work is coordinated by Wageningen UR Agrotechnology and Food Innovations, The Netherlands, and the project consortium is composed of 21 partners representing 11 European countries plus Republic of South Africa. The project is aimed at developing a blueprint for a complete hydrogen production plant employing two-stage hydrogen fermentation.

Hydrogen fermentation The fermentation based biomass conversion to hydrogen builds on anaerobic digestion of carbohydrate-rich substances under influence of bacteria [5]. As a matter of fact, the well known fermentation process converting organic substrates to methane also involves production of hydrogen as an intermediate product but under natural conditions this hydrogen is promptly consumed by methanogenic bacteria. Microorganisms known to produce hydrogen include:

• strict and facultative anaerobic bacteria which convert fermentable sugars to organic acids, H2 and CO2; the highest rates of hydrogen yield were obtained with certain species of thermophilic bacteria [6, 7],

• photofermentative bacteria which use light energy for complete oxidation of substrates to H2 and CO2.

202

Consequently, the hydrogen production by thermophilic fermentation and photofermentation can be considered. Thermophilic fermentation consists in converting sugars or polysaccharides to hydrogen, carbon dioxide and organic acids. The fermentation can be carried out continuously or in batch mode. The typical chemical reactions of thermophilic fermentation of glucose, xylose and sucrose are presented below.

C6H12O6 + 2H2O → 4H2 + 2CH3COOH + 2CO2

3C5H10O5 + 5H2O → 10H2 + 5CH3COOH + 5CO2

C12H22O11 + 5H2O → 8H2 + 4CH3COOH + 4CO2

The maximum theoretical yields of hydrogen from glucose, xylose and sucrose, expressed in moles of hydrogen per 1 mole of substrate are 4, 3,3 and 8, respectively. Photofermentation consists in converting organic acids to hydrogen and carbon dioxide. Assuming that acetic acid is the substrate, the photofermentation can be represented by the reaction:

CH3COOH + 2H2O → 4H2 + 2CO2

The maximum theoretical yield of hydrogen is 4 moles per 1 mole of acetic acid. During photofermentation the photoheterotrophic microorganisms absorb and use the electromagnetic radiation as an energy source [7]. The advantage of photofermentation is the production of hydrogen-rich gas containing very little carbon dioxide (10-20% vol) due to the alkalinity of the fermentation broth. After simple upgrading, the gas can be supplied to a fuel cell. As the organic acids produced in thermophilic fermentation can be used as substrates for photofermentation, both processes can be coupled as schematically shown in Figure 1. Compared to the attainable hydrogen yield of thermophilic fermentation, the yield of the two-stage process can be higher by a factor of up to almost 3 making it possible to obtain from biomass about 70% of theoretically available hydrogen.

203

Figure 1. Scheme of two-stage fermentative hydrogen production

The main stages of the fermentative hydrogen production process are following [8]:

• Biomass pretreatment to give fermentable feedstock and nonfermentables, • Thermophilic fermentation in which fermentable feedstock is converted to

hydrogen gas and organic acids, • Photofermentation in which organic acids are converted to hydrogen gas, • Upgrading of hydrogen gas to meet product specification, • Separation and treatment of non-fermentables.

Thermophilic fermentation is rather well understood but the knowledge of photofermentation is still far from complete. From the engineering point of view it is difficult to design a photofermenter ensuring effective supply of light energy needed for achieving a satisfactory hydrogen yield. Preliminary economic estimates indicate that industrial-scale applications of the two-stage hydrogen production process can be economically viable.

Potential raw materials for hydrogen fermentation With regard to the chemical composition there are three different kinds of biomass, which can be used as raw materials for thermophilic fermentation [7]:

• Sugar containing biomass (i.e. sugar beet, sugar cane, sweet sorghum), • Starchy biomass (i.e. potato, cereals), • Lignocellulosic biomass (i.e. grass, wood, straw).

Apart from various kinds of crops, waste or byproducts from biomass processing industries (beet pulp, molasses, potato peels, apple pulp, wheat bran, brewers grain etc.) can be used for hydrogen production. Characteristic data including chemical composition of four selected kinds of biomass are given in Tab. 1 [9, 10, 11].

Biomass

pretreatment

Thermophilic fermentation

Photofermentation

Biomass

nonfermentable fraction

light

nonfermentables

H2 + CO2 H2 + CO2

Gas upgrading

H2

CO2

204

Table 1. Data of selected kinds of biomass

Raw material Component, % of dry

mass Sugar beet Potato Pressed beet pulp Wheat straw

Sucrose 68 - 4 -

Starch - 76 - -

Cellulose ~4 ~4 20 35

Hemicellulose ~5 - 32 25

Lignin ~1 - 4 18

Average yield, t fresh biomass/ha 45 47 - 4,1

Water content in fresh biomass, % 75 75 75 17

The substrates to thermophilic fermentation are simple sugars (glucose, xylose, arabinose etc.) and therefore each biomass type requires a proper method of pretreatment in order to obtain a water solution of the simple sugars. The method of biomass pretreatment depends mainly on the type and initial form of raw material. In case of sugar containing biomass, the pretreatment step consists in an extraction process of raw juice which can be directly supplied to the thermophilic fermenter. The pretreatment of starchy and lignocellulosic biomass is more complicated due to the hydrolysis process, in which polysaccharides are converted to simple sugars. An important problem is the removal of lignin before the fermentation unit. In contrast to cellulose and hemicellulose, lignin is not converted to simple sugars and moreover it may hamper the growth of microorganisms used to hydrogen production [7]. Nonfermentable fraction which contains lignin can be utilized by combustion or gasification to generate energy. It is also known that pure lignin is a good raw material for the production of chemicals and other products, e.g.:

• Biopolymer additives (antioxidant, UV stabiliser, colouring agent), • Plasticizers in cement, • Precursors for phenol-based chemicals, • Binder for briquette, fertilizers and feed, • Surfactants, • Emulsifiers, • Carbon fibres.

These wide directions of use of lignin have resulted in an increasing interest in the research on new or modified methods of lignocellulosic biomass pretreatment, in which lignin is separated before hydrolysis. For example in an organosolv process, ethanol is added to lignocellulosic material formed of chips [12]. It causes chemical

205

breakdown of lignin and solubilization of lignin components. Then dissolved lignin is recovered via distillation with almost 99% efficiency. The obtained results are shown in Table 2.

Table 2. Theoretical and real yields of simple sugars and lignin

Raw material Theoretical/real yield,

kg/t fresh biomass Sugar beet Potato Pressed

beet pulp Wheat straw

Sucrose 170 163.2 - 10

9.6 -

Glucose - 211.1 198.4

55.6 50

322.8 290.5

Xylose - - 90.9 81.8

235.8 212.2

Lignin - - 10 9.9

149.4 147.9

On the basis of data given in Table 1 and information from literature, the theoretical and real yields of fermentable simple sugars and lignin were calculated. Calculations were carried out under the following assumptions:

• 96% efficiency of raw juice extraction process in beet sugar factory, • 94% efficiency of potato starch hydrolysis, • 90% efficiency of lignocellulose saccharification with 99% efficiency of lignin

recovery.

Hydrogen yield of the fermentation process At the present stage of development of hydrogen fermentation it is known that the real efficiency of two-stage process described in Section 2 above is up to 70%. The data shown in Tab. 2 were used to estimate the theoretical and real hydrogen yields of two-stage process. The results are shown in Fig 2. For crops the real yields were also converted to kilograms of hydrogen per hectare of growing area and the results are shown in Figure 3.

206

22,926,5

18,9

67

1618,5

13,2

46,9

sugar beet potato pressed beet pulp wheat straw0

10

20

30

40

50

60

70

kg H

2/t fr

esh

biom

ass

theoretical yield

70% conversion

Figure 2. Theoretical and real hydrogen yield of two-stage fermentation process

721,5

870,5

192,4

sugar beet potato wheat straw0

200

400

600

800

1000

kg H

2/ha

Figure 3. Real hydrogen yields from crops

Conclusions Due to the low moisture content - about 17% - the highest yield of hydrogen per kilogram of biomass is obtained from wheat straw. The yields from sugar beet, potato and beet pulp are considerably lower. On the one hand, high moisture of raw material increases the costs of biomass transport, but on the other hand it facilitates hydrogen production process by reducing fresh water requirement.

207

Another advantage of wheat straw is a high content of lignin, which after separation process becomes a valuable co-product. When the yield is expressed in kilograms of hydrogen per hectare of growing area, the highest yield is obtained from potato. However, the pretreatment of potato is more expensive than the pretreatment of sugar beet.

Acknowledgement Support of the European Commission under contract 019825 HYVOLUTION is gratefully acknowledged.

References [1] Energy for the future: Renewable sources of energy. White paper for a community

strategy and action plan. COM(97)599. [2] Claassen PAM, van Lier JB, Lopez Contreras AM, van Niel EWJ, Sijtsma L,

Stams AJM, de Vries SS, Weusthuis SA. Utilisation of biomass for the supply of energy carriers. Applied Microbiology and Biotechnology, 1999; 52: 741-755.

[3] Hofbauer H. Conversion technologies: gasification overview. Paper presented at 15th European Biomass Conference and Exhibition, Berlin, 2007.

[4] Claassen PAM, de Vrije T. Non-thermal production of pure hydrogen from biomass: HYVOLUTION. International Journal of Hydrogen Energy, 2007; 41: 1416-1423.

[5] Claassen PAM, de Vrije T, Grabarczyk R, Urbaniec K. Development of fermentation based process for biomass conversion to hydrogen gas. Paper presented at PRES Conference, Prague, 2006.

[6] van Niel EWJ, Budde MAW, de Haas GG, van der Wal FJ, Claassen PAM, Stams AJM. Distinctive properties of high hydrogen producing extreme thermophiles, Caldicellulosiruptor saccharolyticus and Thermotoga elfii. International Journal of Hydrogen Energy, 2002; 27: 1391-1398.

[7] Reith JH, Wijffels RH, Barten H. Bio-methane & Bio-hydrogen. The Hague: Smiet offset, 2003.

[8] Wukovits W, Friedl A, Markowski M, Urbaniec K, Ljunggren M, Schumacher M,Zacchi G, Modigell M. Identification of a suitable process scheme for the non-thermal production of biohydrogen. Paper presented at PRES Conference, Ischia, 2007.

[9] van der Poel PW, Schiweck H, Schwartz T. Sugar technology. Berlin: Bartens, 1998.

[10] Urbaniec K, Grabarczyk R. Selected kinds of biomass as raw materials for hydrogen fermentation. Inzynieria Systemow Bioagrotechnicznych, 2007; 15:11-17 (in Polish).

[11] Jorgensen H, Vibe-Pedersen J, Larsen J, Felby C. Liquefaction of lignocellulose at high-solids concentrations. Biotechnology and Bioengineering, 2007; 96: 862-870.

208

[12] Mabee WE, Greeg DJ, Arato C, Berlin A, Bura R, Gilkes N, Mirochnik O, Pan X, Pye EK, Saddler JN. Updates on softwood-to-ethanol process development. Applied Biochemistry and Biotechnology, 2006; 129: 55-70.


209

Development of a Graphical Analysis Method for Renewable Energy Supply Chain

Hon Loong Lam, Petar Varbanov1, Jiri Klemeš

EC MC Chair (EXC) INEMAGLOW 1 EC MC ERG – ESCHAINS

Research Institute of Chemical Technology and Process Engineering, FIT University of Pannonia, Egytem u. 10, H-8200 Veszprém, Hungary.

Tel: +36 88421664, Email: [email protected]

Abstract A graphical representation method is proposed to reduce the energy waste and improve the efficiency of energy supply chain systems. The energy demand and supply loads in the studied region are represented with a cumulative curve. It provides a visual analysis where zones with their energy surplus and deficit can be matched and combined forming energy supply chain clusters. The energy balance within each cluster has to be adjusted to overcome the problem of energy waste. It explores the tradeoffs between closer proximity of energy sources and sinks reducing transportation losses. This will help decreasing the Carbon Foot Print (CFP) and could improve stability and reliability of the power supply.

Introduction Renewable energy is generated from natural resources as solar radiation, wind, water flow, geothermal heat and biomass [1]. These sources are naturally replenished on a seasonal basis. They are converted from the primary sources to energy carriers to provide the required services. Figure 1 illustrates the energy transformation in the energy supply chain. Renewable energy plays an important role of reducing greenhouse gas (GHG) emissions when the world has to cope with rising energy demand. The development of the renewable energy supply receives increasing attention from energy policy makers and operators. This paper presents an overview of the energy supply chain development. It typically involves replacement of fossil fuels by various types of renewable energy sources (RES). A framework which incorporates energy sources (ES) into the energy supply chains is proposed. This includes the generation of an inventory model for energy demand and supply. It follows by the energy supply chain region clustering. The Life Cycle Assessment and CFP concepts are taking into account, trade-offs between emissions, cost and energy efficiency of the different technologies [2].

Overview of Renewable Energy Supply Chain (RESC) The development of energy supply chain especially RESC is getting momentum. The major challenges in RESC development are: the availability of ES, the efficiency of energy conversion technologies, energy distribution and economic analysis of coping with the energy demand.

210

Figure 1: Renewable Energy Chain An important step to secure the sufficient and reliable energy supply responding to the development of energy demand is the formulation of an appropriate supply model. This requires a multi-disciplinary approach. Energy, economy and the environment should be included as model components [3]. These components are illustrated in Figure 2. They are represented by the energy demand and supply information flow, the basic information for each stage of supply chain pathway. It also includes the constrains that should be considered for the energy model. Figure 2 Components of A Renewable Energy Supply Chain.

Consumer

Energy supply

Conversion

Distribution

Demand information flow Supply information flow

Basic Information Supply Chain

Pathway Constrains

i. Environmental impact

ii. Conversion and distribution efficiency

iii. Operating capacity

Resource availability and limitation

Policy legislation Demand forecasting model

Distribution networks

Conversion Technologies

i. Geographical information ii. Energy sources

Biomass

Flowing water

Wind

Geothermal

heat

Solar radiation

…

Electricity

Biofuels

Hydrogen

Hot utilities

…

Transport Manufacturing Residential Business &

commerce

…

Transportation

Production

Heating

Cooling

Customer Services …

Primary Sources Energy Carriers Facilities Services

211

Iniyan and Sumathy [4] proposed an integrated renewable energy model called Optimal Renewable Energy Model (OREM). It aims to minimise the cost and the efficiency ratio. It determines the optimum allocation of different renewable energy sources for various end-uses. The pattern of renewable energy distribution depends on the potential of RES, energy demand, reliability of renewable energy systems and their acceptance level. Nakata [3] reviewed the various issues related to the energy-economic model and its application to national energy policies, renewables energy systems and global environment. The energy model is used as a tool for energy planning and decision support systems. It typically includes forecasting module for estimation of future demand and optimisation model for optimal allocation of the load and its supply chain [5]. Energy models and energy supply chain strategies are based on the availability of the sources of renewable energy. WEA, 2000 stated that the world potential of renewable energy is sufficiently large to meet future world energy requirements [1]. However, the main focus is how to harness this potential energy in an economically, environmentally and socially acceptable way. The renewable energy harnessing is limited by various barriers. Renewable resources are faced with considerable economic disadvantages. Under current conditions renewable resources tend to be more expensive than fossil resources [6]. The availability of the RES as the electricity production by wind is linked to fluctuations [7]. The fluctuations are caused by the availability of the renewable sources, back up possibility, trends and predictability. Strengres et al [8] demonstrated how renewable energy potential is influenced by its land restrictions. Studies presented by Carta and Mentado [9] and Hoffmann W [10] demonstrated the fluctuations caused by wind speed and solar radiation. Several methods have been proposed to overcome the problem of fluctuations in energy supply system through combination of various RES in an integrated system. Lund and Munster [11] introduced the combination of wind power and distributed energy generation by combine heat and power, (CHP) has been proposed to overcome the fluctuations in electricity production. Korpas and Greiner [12] and Greenblatt et al [13] introduced integrated renewable energy systems to produce hydrogen and compressed air. These can be used as an energy carrier and a storage medium to back-up the intermittency and fluctuating output of many renewable resources [14]. A schematic diagram which integrates RES and waste to energy into energy supply chain is proposed and illustrated in Figure 3.

Renewable Energy Supply Chain Inventory Database The energy supply chains make use of the energy sources and the energy distribution network to meet the energy demand of all the tasks in an optimal manner. This optimality depends on the selected objectives

i. Minimising the total annual costs of energy by reducing transportation distance.

ii. Maximising the local energy resources iii. Miinimising the emission of GHG and reducing CFP along the energy

suppy chain .

212

Figure 3 Schematic diagram for integrated renewable energy supply chain A novel inventory approach is proposed to develop an energy supply chain database as well as the energy demand and supply mapping. Geographical Information System (GIS) is used as a tool for capturing, storing, analysing, and displaying geographical information. It is based on some reported applications. Voivontas et al [15] applied GIS to assess the potential of RES in Greece. Dominguez and Amador [16] used GIS as a powerful tool to manage information and planning of the RES as most of the information in the RES has some special geographical qualities. A formulation of an appropriate inventory database by a multi-layered input is suggested by this work. The geographical referenced information is analysed and ready to be extracted as parameters in the energy supplied chain system. Figure 4 illustrates the way geographical objects and the attributes are used for the implementation of the inventory approach within the GIS environment. The first section of the multi-layer database is the information about RES. It contains the value of parameters in the energy conversion calculation. The second section is the database for a distribution network. It includes the location of power plants, the infrastructure network (road, railway, water transport, grid network and pipeline) to indicate the transportation distance and possible required transport. The last section of the database contains the information of energy demand from the consumers. This can be obtained and estimated from the economic activities, population census and the weather forecast.

Waste treatment

Electricity Fuel Heat

Distributed Energy

Generation

Consumer Energy Carriers

- Hydrogen - Compressed air - Energy Storage

Centralized Grid &

Distribution Network

Industry Resident

Commerce Transport

Renewable energy sources - Wind - Hydro - Geothermal - Solar - Biomass

Energy Sources

Emission (Included GHG)

Waste to energy

Fossil energy sources

Waste

213

Figure 4 Objects and attributes of GIS inventory database The energy supply and demand model can be obtained by extracting the information from GIS database. The energy demand model is formulated using different variables from the regional economic activities, population, and weather information. The energy supply model can be estimated from the energy sources information which are incorporating a wide variety of the latest energy conversion technologies. They are including electricity from photovoltaic cells, wind and hydropower, heat from thermal solar panels and geothermal sources and energy storage from diverse renewable sources. The cogeneration of heat and power is also covered.

Regional Energy Balance The studied region contains a numbers of zones. The energy demand and supply index for each zone depend on their economic activities and the availability of renewable energy. Certain region may supply the energy and possibly on renewable energy that exceeds its demand during certain periods of time. This creates a surplus. Another region may be confronted with a deficit. These situations have been graphically represented in Figure 5. The regional energy balance is given in Eq. (1)

bi = si - ndi Eq.(1)

where,

bi - energy balance in Zonei. The positive and negative signs indicate energy surplus or deficit.

si - Estimated renewable energy from Zonei

RES Supply

Distribution network

Consumer Demand

Multi-layer database

& intermediate-info mapping

Inventory Database

214

n - Renewable energy contribution share di - Forecasted energy demand in Zonei

Figure 5 Regional energy surplus and deficit.

Regional Energy Clustering Regional energy clustering method is proposed to manage the energy balancing among the zones. It aims to overcome the problem of energy waste. The energy surplus and deficit from different zones can be matched and combined to form energy supply chain clusters. Each cluster is formed to secure sufficient energy provision within the combined zone. This can be done by integrating the conventional energy sources and the targeted contribution share of RES. The cluster with closer proximity of energy sources and sinks could reduce the distance of energy transmission and transportation. The shorter distance of energy sources distribution within the cluster is beneficial from the LCA point of view. It brings reduction of:

i. The energy waste ii. The energy cost iii. The CFP

Peç as Lopes et al [17] reported that a cluster system could promise significant advantages in power supply operations. The advantages are:

1. Enhanced local reliability of energy supply 2. Reduction of feeder losses 3. Support of local voltage 4. Increased transmission efficiency 5. Uninterruptible power supply functions

Priority Analysis Priority analysis has been introduced to identify the sequence of clustering. A cluster is formed by combining the zones until the summation of energy balance is close to zero see Eq. (2) and Figure 6.

Zone 2 Zone 3 Zone 4 Zone 1

Energy Balance

0 Surplus

Deficit

Surplus Surplus

Zone

215

Σbi ≅ 0 Eq. (2)

Figure 6 Priority analysis for m x n zones matrix The zones in the first cluster will be labeled by the sequence of Z1, Z2, Z3 ... The next cluster will be formed by combining the other zones. Combining zones in the best sequence is an optimisation problem. The optimisation outcome depends on the selected priorities:

i) Cost of energy ii) Environmental impact iii) Other possible factors (e.g. transmission limitation and geographical

barriers) The cost of energy is obtained from several estimations. This includes the cost of energy production, the cost of distribution and the cost of penalty for the intermittency and fluctuating energy supply. The fluctuating RES availability carries penalties for sufficient energy storage facilities and also the back-up power generation plant whenever power generation is not matching grid demands. The environmental impact depends on the assessment of CFP for transmission, distance, volume and density of transported energy.

Energy Supply Chain Cluster The concept of the energy supply chain cluster can be further illustrated as shown in Figure 7. The energy demand and supply in the studied zones are displayed by a monotonic cumulative curve. It provides a visual analysis of pointing to regions with energy surplus and deficit can be combined and matched as a new energy supply chain cluster. The sequence of zone is based on the priority analysis. The solid line in Figure 7 represents the cumulated profile for energy demand requested in a specific zone and the dotted line is the energy supply profile. The end of the cumulative curve at the right edge shows the total energy demand and supply in the region. The vertical distance between the energy demand curve and supply curve indicates the energy balance, bi. Every combined cluster should achieve energy balance close to zero see Eq. (2). This allows certain level of flexibility of energy surplus or deficit gap. This gap could be covered by energy import or export among the clusters or from the outside of the region. The figure also indicates the size of each cluster and the total energy involved in the supply chain within the cluster.

Z1

Z

Z

Z

Z Z

Z

Z

Z

216

Energy demand

cumulative curve

Energy supply cumulative curve

Region Size

Cluster 1

Cluster 2

Cluster 3

Energy

∑=

n

iib

1

Z1 Z3 Z2 Z5 Z6 Z9 Z7

Energy balance to be imported or exported

(Area) Z4 Z8

Figure 8 Diagram illustrate the formation of the energy supply chain clusters This new graphical method is a step towards development of an integrated supply chain for renewable energy. RESC should take into account each and every stakeholder along the supply-demand pathway. It includes RES suppliers, transportation and transmission of the energy carriers, conversion operators, electricity suppliers, storages, and consumers. This has to consider social concerns, environmental and economical impacts related to establishing the renewable energy systems and the specific conditions of each studied region. The core part of integrated supply chain is the information base. It includes both the basic and detailed decision information. Basic information includes the multi-layered GIS database. It contains RES materials database, RES logistic database and the RES conversion database. The detailed decision information considers the consumers requirement and energy policies which form the parameters values database (default values and the variables database). The GIS based interface allows visualization of the interacted relationship between the intra-cluster and inter-cluster. This supports the users in the process of making decisions and evaluation according to their own interests (environmental, economical, social or combination).

Conclusion A graphical concept is proposed incorporating RES into energy supply chains. This includes the generation of inventory database for energy demand and supply, and the energy supply chain region clustering. Energy surpluses and deficits in the

217

studied region can be matched and combined forming energy supply chain clusters. The supply chain cluster can be represented with a cumulative curve to provide a clear insight. This extension provides as a tool for energy planners and operators, who work with energy supply chain under the environmental and economical constrains. The clustering system could help reducing the energy waste, energy cost and CFP. It offers a possibility to improve power supply quality and reliability.

Acknowledgments The financial support from the EC MC Chair (EXC): INEMAGLOW, MEXC-CT-2006-042618 and EC MC (ERG): ESCHAINS, MERG-CT-2007/46579 , FP6 project TEMPUS-TACIS JEP_26045_2005, ECORSE "Ecological and Resource Saving Engineering" are gratefully acknowledged.

References [1] WEA, 2000. World Energy Assessment: energy and the challenge of

sustainability. UNDP,UN-DESA and the World Energy Council United Nations Development Programme, New York.

[2] Klemeš J. and Huisingh D., Economic use of renewable resources, LCA, cleaner batch processes and minimising emissions and wastewater. Journal of Cleaner Production 16 (2008) 159 - 163

[3] Nakata, Energy-economic models and the environment. Progress in Energy and Combustion Science 30 (2004) 417–475.

[4] Iniyan and Sumathy, An optimal renewable energy model for various end-uses. Energy 25 (2000) 563 – 575

[5] Marik, Schindler, Stluka, Decision support tools for advanced energy management. Energy (2007). doi:10.1016/j.energy.2007.12.004

[6] Narodoslawsky et al, 2008, Utilising renewable resources economically: new challenges and chances for process development, Journal of Cleaner Production 16 (2008) 164-170

[7] Salgi and Lund, 2008, System behaviour of compressed-air energy-storage in Denmark with a high penetration of renewable energy sources. Applied Energy 85 (2008) 182–189.

[8] Strengers, B.J., Leemans, R., Eickhout, B., de Vries, B., Bouwman, A.F., 2004. The land use projections in the IPCC SRES scenarios as simulated by the IMAGE 2.2 model. GeoJournal 61, 381–393

[9] Carta J. A., Mentado D., 2007, A continuous bivariate model for wind power density and wind turbine energy output estimations. Energy Conversion and Management 48 (2007) 420–432.

[10] Hoffmann W., 2006, PV solar electricity industry: Market growth and perspective. Solar Energy Materials & Solar Cells 90 (2006) 3285–3311.

[11] Lund and Munster, 2003, Modelling of energy systems with a high percentage of CHP and wind power. Renewable Energy 28 (2003) 2179 – 2193.

[12] Korpas and Greiner, 2008, Opportunities for hydrogen production in connection with wind power in weak grids. Renewable Energy 33 (2008) 1199–1208.

[13] Greenblatt et al, 2007, Baseload wind energy: modeling the competition between gas turbines and compressed air energy storage for supplemental generation. Energy Policy 35 (2007) 1474–1492.

218

[14] Ibrahim H., Ilinca A., and Perron J. , 2008, Energy storage systems—Characteristics and comparisons. Renewable and Sustainable Energy Reviews 12 (2008) 1221–1250.

[15] Voivontas et al, 2001, Assessment of biomass potential for power production: a GIS based method. Biomass and Bioenergy 20 (2001) 101 – 112)

[16] Dominguez and Amador, 2007, Geographical information systems applied in the filed of renewable energy sources. Computer and Industrial Engineering 52 (2007) 322 – 326.

[17] Peç as Lopes et al, 2007, Integrating distributed generation into electric power systems: A review of drivers, challenges and opportunities. Electric Power Systems Research 77 (2007) 1189–120


219

Life Cycle Assessment as an Environmental Assessment Tool in Municipal Solid Waste Management

Luca De Benedetto, Jiří Klemeš EC Marie Curie Chair (EXC) “INEMAGLOW”, Research Institute of Chemical

Technology and Process Engineering, FIT, University of Pannonia Egyetem ut 10, Veszprém, H -8200, Hungary

Tel: +3688421664, Email: [email protected]

Abstract Life Cycle Assessment (LCA) is a tool for analyzing environmental impacts on a wide perspective and with a reference to a product system and economic activity. The history and the main concepts of the LCA methodology have been briefly reviewed. The LCA framework has been introduced, with regards to goal and scope definition, inventory analysis, impact assessment and interpretation. The main developments in Municipal Solid Waste (MSW) and application of LCA in waste management have also been addressed. A model of waste treatment system that can be applied to the Hungarian reality has also been introduced.

Introduction Life Cycle Assessment involves the evaluation of specific elements of a product system to determine its environmental impact. The implementation varies depending on the adoption pattern and on the precision that needs to be achieved. Due to the constraints on resource or data availability, industrial companies perform most of the time analyses based on a more simplified approach (Life Cycle Approach). In some cases they just apply the general principles to certain aspects of the production system (Life Cycle Thinking). Nevertheless all are usually referred as LCA activities. The LCA is also called Life Cycle Analysis or cradle to grave approach. It comprises a conceptual framework and a set of tools that have been studied and developed in the last 30 years [1]. The core of the concept is the assessment of the impacts at each stage of the product life cycle. The term 'product life cycle' is used with a reference to the extraction of raw materials, production, manufacturing, distribution, use and disposal including all necessary transportation steps. The proposed view is therefore holistic and includes the entire lifespan of a product. The first studies on LCA date from the late sixties, early seventies. In 1969, the Coca Cola Company funded a study to compare resource consumption and environmental releases associated with beverage containers [1]. Similar studies were then started in UK, Switzerland and Sweden [2]. In these early studies LCA was closely linked with energy analysis. Due to the energy crisis of the early seventies, waste and effluents were initially not considered as priority and attention was concentrated on calculating the total energy used in production of various household goods. Bousted [2] studied at that time various types of beverage containers, including glass, plastic, steel and aluminium in the UK. After the oil crisis subsided both the energy issues and the use of LCA applications lost some prominence. It was only in the late eighties and early nineties when a new interest in the tool was found [9] and coupled with efforts to bring standardization to

220

its use. In 1989 the Society of Environmental Toxicology and Chemistry (SETAC) started working on defining a common terminology and a methodology framework. One of the first results of this work was the definition of functional unit. This is a quantified description of the product systems to which impacts are attributed. This unit sets the scale for comparison of two or more products and one if its main purposes is to provide a reference to which the input and output data are normalised. Three aspects have to be taken into account when defining the functional unit [9]:

• Efficiency • Durability • Performance

When performing an assessment of more complicated systems (e.g. multifunctional systems like waste treatment) special attention has to be paid to by-products. This standardization work was then picked up by the International Standard Organization (ISO) in 1994 with the first of its 14040 series. The rigid context of the ISO offered coherence to the different methodologies and approaches in LCA without imposing one. The ISO work has resulted in the definition of specific steps that allow the separation of the subjective and objective phases within the proposed method. The principles and framework for LCA in these documents include: goal and scope definition, life cycle inventory analysis (LCI), life cycle impact assessment (LCIA), life cycle interpretation phase and reporting. These phases are the codification of the same steps individuated by SETAC in the previous years. The exception is that Life Cycle Improvement has been considered an activity that should permeate all other phases and not one of its own. The Interpretation phase was added instead. The interest on this topic is witnessed also by newer versions of the above mentioned series, the latest of which published in 2006 (ISO 14040:2006 effectively replaces 14040:1997, 14041:1998, 14042:2000 and 14043:2000).

Figure 1 Phases and application of LCA (adapted from ISO 14040:1997)

Goal and scope definition

Inventory Analysis

Impact Assessment

Direct Applications: • Product

Development • Strategic planning • Marketing • Public policy making

Other aspects: • Technical • Economic • Market • Social

Inte

rpre

tatio

n

221

During the last years the methodology of the LCA has been consolidated. Generally LCA is accepted as a tool, which allows progress towards full environmental responsibility for all corporate and public stakeholders. Some difficulties in the methodology have been recognized by Frankl [26]:

• Complexity • High cost and long time scales • Uncertainty about valuation • Continuing invisibility of much of LCA work.

Another challenge lies with the communication of the results. Long reports might put off many users. If the results are too simplistic then there is difficulty in validating them. A survey by the European Environment Agency [9] pointed at the following social impacts of LCA:

• LCA is now seen as necessary by all stakeholders as integral part of environmental management tool kit

• Use of this tool is also seen important in the process of corporate strategy formulation

• Level of knowledge of LCA remains worryingly low in the general public • Level of progress in LCA adoption varies between countries • Quality control mechanism remains relatively weak • Involvement of external stakeholders in defining study boundaries is seeing

increasingly important.

LCA: the general framework The methodology for Life Cycle Analysis includes the phases described in Figure 1. It should be noted that ISO 14040 does not describe the technique in detail, nor does it specify which methodology should be used for each phase. It provides mainly a framework in which these elements can be developed.

Goal and Scope Definition This is the first subjective phase of the application of LCA. At this stage it is necessary to identify the aim of the analysis and the system boundaries. This is to ensure that no relevant part of the system to be investigated is actually left out. The definition of the goal and the scope are critical elements, since the results will depend greatly on them. The goal needs to state clearly and without ambiguity which is the application, what are the reasons, why the study is carried out and who the recipients of the results of the study are. A goal identified in such a way will allow the practitioner to perform the correct choices throughout the study. The goal could also be adjusted depending on specific and relevant findings of some later steps of the analysis. The scope sets the borders of the assessment. Different elements should be considered in order to specify the scope correctly: product group, functional unit, the system and its boundaries, impact assessment boundaries, the data quality requirements, limitations. The definition of the system boundaries (inputs and outputs) are critical in order to determine the amount of work to be done. This activity is quite subjective and requires decision to be taken on the following areas: geographical boundaries, life cycle boundaries and boundaries between the

222

technology and the biosphere. Given the subjective characteristics of this activity, it is necessary to be very transparent with regards to all decisions and assumptions taken at this stage. The impact categories need to be chosen from a list of standard ones: assessing the boundaries will also limit the categories to be considered during the study. As for the goal, even the scope can be adjusted during the iterative process of the analysis.

Inventory Analysis Aim of this second phase is to perform mass and energy balances to quantify all the material and energy inputs, waste and emissions from the system causing the environmental burdens. The following main issues (as defined by ISO 14040:1997) should be considered during this phase: data collection, refining system boundaries, calculation, validation of data, relating data to the specific system and allocation. Data can be specific (to the process, the company, and the geographical area) or more generic (extracted from trade organizations or governmental institutions). To perform this phase it is possible to rely on quantitative or qualitative data, according to their availability. Data collection can prove to be the most work intensive activity of a Life Cycle Assessment. In some cases it is possible to rely on average values from trade organizations or literature. Generally the results obtainable with LCA are very sensible to the set of data used. It is important to understand the intrinsic criticality of this phase. More generic and qualitative data could be used for a first simplified analysis, to be reiteratively repeated on more specific and system related data. The data initially collected can be used to review the system boundaries, as defined in the previous phase. If the system is very complex it might be necessary either to review the system boundary or to include more data. Another possibility is the allocation of the relevant environmental burdens to the system, ensuring that the approximation of the input-output relationship and the main characteristics is possible. Allocation might be necessary in case of multi-input or multi-output systems. The inventory should then be interpreted considering all the specified uncertainties and lack of data. In particular validation should also be considered and carried out during the whole process of data collection to reduce eventual discrepancies and data quality issues later on.

Impact Assessment The third phase is based on the aggregation of the environmental impacts quantified in the Inventory Analysis into a limited set of recognizable impact categories (e.g. global warming, ozone depletion, acidification etc.). This phase comprises the following steps: classification, characterisation, normalisation and weighting. According to ISO 14042:2000 there are three groups of categories to be considered: Resource use, Human Health consequences and Ecological consequences. These broad groups should include all categories like climate change, stratospheric ozone depletion, photochemical oxidant formation (smog), eutrophication, acidification, water use, noise, etc. These categories should be selected from a list of examples and be relevant to the system under investigation.

223

The second step in this phase is mainly a quantitative step: characterisation. In this step it is necessary to assign the relative contribution of each input and output to the selected impact categories. Pennington [8] proposed a generic equation to calculate indicators from the inventory data. For each impact category he used generic characterisation factors.

∑ ×=S

sInventoryEmissionsFactorsationCharacteriIndicatorCategory )(.)(.. (1)

Where s indicates the inventory data (input). The characterisation factors can be found in literature as databases or are available in various LCA support tools. The following equation takes into account some of the potential variables of non-generic characterisation factors in the context of human health and natural environment:

∑=j jsEmission

tjsEffecttisFactorsationCharacteri),(

),,(),,(. (2)

Subscript i is the location of the emission; j is the related location of exposure of the receptor and t is the time period during which the potential contribution to the impact is taken into account. The next step in this phase is the normalization. This activity is described by Stranddorf [10] as a necessity to calculate the magnitude of the category indicator results relative to reference values where the different impact potentials and consumption of resources are expressed on a common scale. The goal of normalization is to set a common reference enabling comparison of different environmental impacts. Quantitative results of the above mentioned characterisation of impact categories are not always comparable and an additional step is necessary: weighting. This activity aims at comparing the impact categories against each other. This would allow ranking and possibly defining the relative importance of different results. Weighting can be a quantitative or qualitative activity based on social or political considerations. Different weighting methods have been developed [11].

Interpretation This is the last phase as indicated by ISO 14040:1997. Interpretation is a systematic procedure to evaluate information from the conclusions of the inventory analysis and impact assessment of a product system. The following tasks should be accomplished in this phase [9]: 1. Identify the significant environmental issues. 2. Evaluate the methodology and results for completeness, sensitivity and

consistency. 3. Check that conclusions are consistent with the requirements of the goal and

scope of the study, including, in particular, data quality requirements, predefined assumptions and values, and application oriented requirements.

4. If yes report as final conclusions. 5. If not, return to Task 1 or 2.

224

LCA Applications in MSW Management Municipal solid waste (MSW) is a type of waste that includes predominantly household waste (domestic waste) sometimes with the addition of commercial waste collected within a collection area. The waste is either solid or semisolid and generally excludes industrial hazardous waste. MSW consists of everyday items such as product packaging, furniture, clothing, bottles, food scraps, newspapers and other paper. Depending on their nature and composition the treatment of solid waste requires various technologies and a coordinated mix of practices. That includes incineration, gasification, pyrolysis, recycling, composting, and disposal to landfills. A tool generally used to classify different approaches to waste treatment is called Waste Hierarchy [23]. The waste hierarchy implies that some methods to handle waste are more opportune than others from an environmental point of view. Reducing the amount of waste is generally indicated as first priority. The order of the other priorities (reuse, recycle, incinerate and landfill) is based on the type of waste itself and the location where the waste arises. Different studies dealt with the environmental impacts of MSW systems [15], or compared different scenarios of solid waste management ([16], [17]). Both Finnveden [19] and Lee [18] provide good examples on how to use LCA as a tool to analyze different MSW. The goal of the first study was to identify advantages and disadvantages of different methods to treat solid waste (landfilling, incineration, recycling, digestion and composting) and to identify critical factors, including background systems which may affect the results. The study confirms that waste hierarchy is usually a good starting point for waste policy making, however LCA can further support it in identifying situations in which this approach is not valid. Analogous results were also presented by Moberg [20] and Olsson [21]. The study also exposes the fragility of LCA to data gaps and base assumptions. In particular all assumptions on time aspects, system boundaries and characterisation models affect the final results significantly. In few cases this resulted in changing the order of the preference of the proposed waste hierarchy. Clift [14] provided an exhaustive review of the use of LCA in this field. He is evidencing the criticality of the definition of the system boundaries as well as the importance of screening between direct, indirect and avoided environmental burdens in the LCI phase. Waste treatment and Waste Management systems and techniques have been developed in recent years to reduce the environmental burden of land-filling, and also to address the energy aspect. Disposal of organic waste in landfills produces methane (CH4), the reduction of which is significant in fighting greenhouse gases emissions. An interesting experience in this field has been presented by Liamsanguan [22] on the Thai reality. He proposes the use of LCA in determining the best option, from an environmental perspective, to improve the current waste management techniques in Thailand. The study evidences a reduction of more than 50% when land-filling is coupled with source separation (for recycling) and anaerobic digestion.

225

The role of national policies is paramount in the development of waste management and Waste to Energy (W2E) issues. Lee [18] provides an LCA based comparison of landfill, incineration, composting, and feed manufacturing to treat food waste (that accounts in Korea to 30% of MSW). Most interestingly this study compares the situation of food waste treatment in 1997 and 2005. While the categories of global warming and human toxicity decreased significantly (50% and 70% respectively for the reduction of landfilling), acidification, eutrophication, and ecotoxicity increased impressively. This is due to emission in the recycling process, and highlights the need for better control and inclusion of work environment. The role of LCA as decision support tool waste management and W2E issues is further confirmed by studies on bottom ash originating from municipal solid waste incineration (MSWI) as a potential road construction material [21], and on waste hierarchy for MSW and paper treatments [22], [23].

The Hungarian case According to [24] in Hungary almost 105 Mt of waste is generated each year. Approximately 4.5 Mt/y (approx. 450 kg/head) is communal solid waste and approximately. 20 Mt/y is treated liquid communal waste. The remaining approx. 80 Mt/y of waste is generated by industrial, agricultural or other business activities. Within commercial waste approximately 4.2 Mt/y is hazardous waste. The volume of industrial waste is declining and communal waste is slightly increasing. Figure gives an overview of material content of collected MSW in Hungary.

MSW distribution in Hungary

32.4%

16.9%11.4%

4.1%

3.1%

2.8%

29.3%

organicpaperplastictextileglassmetalother

Figure 3 Material content of collected MSW (2004). Adapted from [25] It is important to note that Hungary is member of EU since 2004 and all legislation is harmonised with EU laws, including all laws and directives on environmental protection and waste management. The level of application of these directives is in the progress. Istvàn [25] mapped the Hungarian reality from a MSW perspective. He evidenced that a great effort in the last decade has been put in reducing the number of landfills from 2700 to 150 to cope with the EU legislation. Further work is required in improving the use of MSW in recycling, composting and mostly in energy production.

226

In Figure 3 is the developed model representing the Hungarian waste management systems.

Figure 3 Model of Waste Management and System boundaries According to [25], 85% of the waste is still dumped in landfills. This creates problems of illegal disposals and recultivation of old landfills. Great improvements could be achieved in bio-stabilizing the content of MSW in landfills and increasing the energy recovery percentage, particularly from incineration and recycling. LCA is a tool that can be very beneficial to support the decision process in adopting a waste management solution that is optimal from an environmental and socio-economic perspective. The following scenarios could be analysed with an LCA approach:

1. Landfill with recovery of gas 2. Incineration 3. Incineration and materials recycling 4. Incineration and anaerobic digestion

A process simulation approach could be applied to material and energy flows in the waste management system based on LCA to quantify emissions and financial costs and to determine the most viable solution (Figure 3). Uncertainty estimation can be

Landfill 85%

Collection in Waste centres

Transportation

Materials Recovery Facility

Kerbside collection

Materials

Energy

Emissions

Costs

Energy

Composting Incineration

3.5% 3.5%

8%

227

accounted via repeated simulations. To define the system boundaries, material and energy balances in detail, it is suggested to limit the analysis at municipality level.

Conclusions This contribution is a first step towards analysis and development of a model to represent and further study the local and regional MSW systems in Hungary. As a base, the history and the main concepts at the basis of the LCA methodology have been briefly reviewed with particular attention to Waste to Energy applications. The main developments in MSW and application of LCA in waste management have also been addressed. As a potential solution a process simulation approach has been proposed. This initial study needs to be further developed to include validation of the proposed model and possible applications.

Acknowledgement: Financial support from MC Chair (EXC) INEMAGLOW: FP6-042618 Integrated Waste to Energy Management to Prevent Global Warming and ESCHAINS: MERG-CT-2007/46579 is gratefully acknowledged.

References [1] Udo de Haes HA, Heijungs R. Life-cycle assessment for energy analysis and

management. Applied Energy 2007; 84: 817-827 [2] Boustead I. LCA – how it came about: the beginning in the UK, International

Journal of Life-Cycle Assessment, 1996; 1: 147–150 [3] ISO. Environmental management – life cycle assessment—principles and

framework. Geneva, Switzerland (ISO 14040)1997 [4] ISO. Environmental management – life-cycle assessment; goal, scope definition

and inventory analysis. Geneva, Switzerland (ISO 14041)1998 [5] ISO. Environmental management – life-cycle assessment; life-cycle impact

assessment. Geneva, Switzerland (ISO 14042) 2000 [6] ISO. Environmental management – life-cycle assessment; life-cycle

interpretation. Geneva, Switzerland (ISO 14043) 2000 [7] Rebitzer G, Ekvall T, Frischknecht R, Hunkeler D, Norris G, Rydberg T, et al.

Life cycle assessment Part 1: Framework, goal and scope definition, inventory analysis, and applications. Environment International 2004; 30: 701-720

[8] Pennington DW, Potting J, Finnveden G, Lindeijer E, Jolliet O, Rydberg T, et al. Life cycle assessment—part 2: current impact assessment practice. Environment International 2004, 30: 721-734

[9] Jensen AA, Hoffman L, Møller B, et al. Life Cycle Assessment - A guide to approaches, experiences and information sources. European Environment Agency, 1997. www.lca-center.dk/cms/site.asp?p=2867 (visited 20/04/2008)

[10] Stranddorf HK, Hoffmann L, Schmidt A. LCA technical report: Impact categories, normalisation and weighting in LCA. Serititel FORCE Technology/Denmark 2005. www.lca-center.dk/cms/site.asp?p=2867 (visited 03/02/2008)

[11] Lindeijer E. Part VI: Normalisation and valualtion. In: Udo de Haes (ed.). Towards a methodology for life cycle impact assessment. Society of Environmental Toxicology and Chemistry (SETAC) - Europe. 1996, Brussels

[12] Udo de Haes HA, Jolliet O, Finnveden G, Hauschild M, Krewitt W, Mu¨ller-Wenk R. Best available practice regarding impact categories and category indicators in life cycle impact assessment, SETAC-Europe. Part 1. International Journal of Life Cycle Assessment 1999; 4: 66–74

228

[13] Udo de Haes HA, Jolliet O, Finnveden G, Hauschild M, Krewitt W, Mu¨ller-Wenk R. Best available practice regarding impact categories and category indicators in life cycle impact assessment. Part 2. International Journal of Life Cycle Assessment 1999; 4: 167–174.

[14] Clift R, Doig A, Finnveden G. The application of life cycle assessment to integrated solid waste management, part I—methodology. Transactions of the Institution of Chemical Engineers, Part B: Process Safety and Environmental Protection 2000; 78: 279–287

[15] Harrison KW, Dumas R., Barlaz MA. Life-cycle inventory model of municipal solid waste combustion. Journal of the Air and Waste Management Association 2000; 50: 993–1003

[16] Denison R. Environmental life-cycle comparisons of recycling, landfilling, and incineration: a review of recent studies. Annual Review of Energy and Environment 1996; 21: 191–237

[17] Mendes RM, Aramaki T, Hanaki K. Comparison of the environmental impact of incineration and landfilling in Sa˜o Paulo City as determined by LCA. Resource, Conservation and Recycling 2004; 41: 47–63

[18] Lee SH, Choi K, Osako M, Dong J. Evaluation of environmental burdens caused by changes of food waste management systems in Seoul, Korea Science of The Total Environment 2007; 387: 42-53

[19] [Finnveden G, Johansson J, Lind P, Moberg G. Life cycle assessment of energy from solid waste—part 1: general methodology and results. Journal of Cleaner Production 2005; 13: 213–229

[20] Moberg G, Finnveden G, Johansson J, Lind P. Life cycle assessment of energy from solid waste—part 2: landfilling compared to other treatment methods. Journal of Cleaner Production 2005; 13: 231–240

[21] Olsson S, Kärrman E, Gustafsson JP. Environmental systems analysis of the use of bottom ash from incineration of municipal waste for road construction Resources, Conservation and Recycling 2006; 48: 26-40

[22] Liamsanguan C, Gheewala SH. LCA: A decision support tool for environmental assessment of MSW management systems. Journal of Environmental Management 2008, doi:10.1016/j.jenvman.2007.01.003

[23] Schmidt JH, Holm P, Merrild A, Christensen P. Life cycle assessment of the waste hierarchy – A Danish case study on waste paper. Waste Management 2007; 27: 1519–1530

[24] Competition in local services: solid waste management in Hungary. OECD Submission report 2000, www.gvh.hu/domain2/files/modules/module25/pdf/028 Solid_ Waste_1999_10_ok.pdf (visited 20/04/2008)

[25] István Zs, Present situation and future trends of solid waste management in Hungary, International Conference Decision Support for Waste Management in the EU, June 2007, /holiwast.brgm.fr/ConferenceJune2007/ Presentation/S23%20Z.%20Istvan%20Current%20situation%20and%20future%20trend%20of%20solid%20w.pdf (visited 20/04/2008)

[26] Frankl P, LCA in Industry and Business, Springer 2000, ISBN3-540-66469-6


229

Training Methods of Multiple-Skill Specialists in the Field of Energy-Saving in the Fuel and Energy Balance Complex

Sergey Mikhailov1, Valery Meshalkin2

1Ministry of Industry and Power of RF, Kitaygorodskiy proezd 7, Moscow, 109074, Russian Federation, E-mail: [email protected]

2Mendeleyev University of Chemical Technology of Russia, Miusskaya ploshchad 9, Moscow, 125047, Russian Federation, E-mail: [email protected]

Abstract Modern state and main trends in resource- and energy-saving processes in industry define the necessity of training of specialists, bachelors and masters who should have proper competence in the field of process control based on complex approach to technical and economic energy-saving aspects. It is shown, that, on one hand, problem-solving in the field of energy-saving is impossible without serious technical training which considers the knowledge of technological processes specifics in various industries. On the other hand, commercialization of energy-saving suggestions defines necessity of obtaining certain knowledge in the field of economy and management by a specialist.

Current trends in Russian energy intensity Modern state and main trends in resource- and energy-saving processes in industry define the necessity of training of specialists, bachelors and masters who should have proper competence in the field of process control, based on complex approach to technical and economic energy-saving aspects. It is shown that on one hand, problem-solving in the field of energy-saving is impossible without serious technical training which considers the knowledge of technological processes specifics in various industries. On the other hand, commercialization of energy-saving suggestions defines necessity of obtaining certain knowledge in the field of economy and management by a specialist. At present, the specific energy intensity of Russian economy is high enough. This factor (in calculation on parity of purchasing power) as twice as higher than the specific energy intensity in the USA, 2,3 times - as a whole all over the globe and 3 times - in developed countries of Europe and in Japan. Thus, in Russian Federation, 0,89 tons of conditional fuel are counsumed for production of goods and services, which cost 1 thousand dollars. Whereas, Norway and Sweden, being situated in similar climatic conditions consume 0,36 and 0,26 tons of conditional fuel accordingly [1]. In fact, the recent increase rate of energy intensity reduction slightly exceeds factors, marked in "Energy strategy of Russia for a period before 2020". Particularly, specific energy intensity of GDP declined by 23.5% during 2000-2006 [2]. At the same time, absence of significant reduction energy intensity in different branches of industry with new increasing internal consumption of electric power can considerably reduce Russian competitive edge in the world industry.

230

Industry power efficiency increase at all steps of manufacturing and consumption of electric power is directly linked with the minimization of losses of the electric power. Various groups of losses require different response from each other measures of technical and organisational nature [3]. It is shown, that in conditions of Russian Federation, the reform of electric power sector and creation of the competitive electric power market, commercial losses are becoming of special importance. Those include besides conventional non-payments and losses as a result of theft, also financial losses caused by non-economical work of electric power manufacturers and consumers in the market. Those losses can be caused by, for example, deviation of real industrial electrical consumption schedules from planned [4].

Professionals in energy-saving required At present, professionals in the field of energy-saving are in demand in the labour market. Training methods of multiple-skill experts in the field of energy-saving, considering specific character of power consumption processes in the fuel and energy balance complex, are suggested. The specialists of this type should have regular technical background, including training which allows to reveal and estimate arising losses, but also they should have deep knowledge of the economy and management. This knowledge based on application of logistic approaches allows to develop guidelines of improving of power efficiency for all enterprises participating in the generation - transmission - electric power consumption chain. Demand for specialists of this type should be considered while developing state educational standards for higher education institutions. The program of training of specialists in the field of energy-saving should become a component of Russian Federation energy policy. At present, despite the certain achievements in energy-saving and understanding of the problem by federal and regional governors, the strategic approach to implementation of energy-saving policy is non-existent in the majority of regions. Obviously, introduction and use of resource- and energy-saving technology, organising economic mechanisms to increase power efficiency should be connected with the strategy of the regional development. But in practice, existing strategies of the regional development usually do not contain items related to power efficiency. To overcome this trend, inclusion of discipline "Organization bases of energy-saving management on the regional level" in regional curricula is an appropriate and desirable step forward aimed at education of specialists at technical and economic higher educational institutions. Topics, connected with the general procedure of strategic energy-saving management, should be studied within the framework of the discipline. They are shown in Figure 1.

231

Figure 1- The general procedure of strategic energy-saving management on the regional level

Methodological basis of the analysis models of regional energy balance should become another key item of education, as the most important component of regional energy strategy. While studying energy-saving regional aspects, basic strategies on territorial, functional and consumer levels should be considered as well. This is shown in Table 1. The complex approach to education of specialists will enable to form the necessary trained personnel in the field of energy-saving management at the regional level. As a result, territorial social-economic systems will get the boost for sustainable development.

Conclusions The modern labour market in Russia demands specialists, bachelors and masters who should have proper competence in the field of process control, based on complex approach to technical and economic energy-saving aspects. This requires development of specialised courses and training methods for multiple-skill experts in the field of energy-saving, considering specific character of power consumption processes in the fuel and energy balance complex.

Social-economic strategy of the region development

Energy strategy of the region

Dynamic model of the energy balance

Territorial strategy of energy-saving

Functional strategy of energy-saving

Consumer strategy of energy-saving

Model of the energy balance

Ref

orm

mod

el o

f P

LC R

JSC

“UE

S o

f Rus

sia”

Rus

sian

ene

rgy

stra

tegy

232

Table1 - The levels of strategic regional management of power efficiency

Levels Types of strategies

Strategies of the firm development Competitive strategies

Territorial level

Social strategies Strategies in the field of the electric powers generation Strategies in the field of transmission and distribution of electric power

Functional level

Strategies in the field of the electric powers consumption In the field of industry In the field of housing-public facility In agriculture

Consumer level

In social sphere

References [1] Kononov YD, Galperova EV, Mazurova OV, Posekalin VV, The dynamics of the energy intensity of the Russian economy against the background of global trends, International Journal of Global Energy Issues 2003: 20(4):364 – 374. [2] Ministry of Energy of the Russian Federation. The Summary of the Energy Strategy of Russia for the Period of up to 2020. Moscow 2003. [3] Meshalkin V.P, Dovi’ V., Marsanich A. Industrial Logistics Principles. Moscow, Genoa, 2002:722. [4] Pykhtin RV, Skripitsina TA. Increasing Investment Efficiency of Power Generating Companies by Means of Project Funding. Transportnoe Delo Rossii 2006:9(2):62-66 Пыхтин Р.В., Скрипицина Т.А. Повышение эффективности инвестиционной деятельности электроэнергетических предприятий на основе использования проектного финансирования.Транспортное дело России 2006:9(2): 62-66.


233

Technical Analysis of Desiccant Cooling System

François Boudéhenn 1, Françoise Burgun 2 1 INES-GRETh, 50, av. du lac Léman, 73375 Le Bourget du Lac, France,

mail: [email protected], phone: +33 479 444 589, fax: +33 479 72 20 16 2 INES-CEA, 50, av. du lac Léman, 73375 Le Bourget du Lac, France,

mail: [email protected], phone: +33 479 444 566, fax: +33 479 688 049

Abstract This paper is a technical analysis of desiccant cooling system, one of the technologies of solar cooling. Despite the fact that these systems present many advantages (ventilation system, low heating temperature (60-80 °C), thermal COP similar to single-lift absorption chillers (0.5-1), they stay at a research and development step, striking some technological and economical difficulties. Moreover, because of the cost of investment, only high power systems (over 75 kW) are developed. This article presents a description of these systems, their limits and the new development about these technologies.

Introduction Solar cooling systems represent one of the alternative solutions as regards the dual problem of summer comfort and low consumption cooling devices. Nowadays, within the different existing technologies (more or less advanced), two of them seem more adapted to represent the future of solar cooling:

Absorption chillers which have similar working principle than electrically-powered compression chillers, but with a thermal compression in place of the mechanical compression. These mature technologies are now marketed by several manufacturers (Yazaki, Rotartica, EAW, Phönix…), and they represent more than 60 % of the systems installed ;

Desiccant cooling systems, which operate in open cycle, using water as a refrigerant, in direct contact with air. Despite the fact that these systems present many advantages (ventilation system, low heating temperature (60-80 ° C), thermal COP similar to single-lift absorption chillers (0,5-1), they stay at a research and development step, striking some technological and economical difficulties. Moreover, because of the cost of investment, only high power systems (over 75 kW) are developed.

While economical data of experimental desiccant cooling systems are not representative of the real cost, a technical analysis of these systems would be very useful to identify key points to reconsider the future development.

Description of desiccant cooling systems

Technical description The Figure 1 presents a schematic drawing of a desiccant cooling air-handling unit.

234

Figure 1 - Schematic drawing of a desiccant cooling air-handling unit

The ambient air is dehumidified by adsorption into the desiccant wheel (1). After pre-cooling into the heat recovery wheel (2), the dried fresh air is cooled by direct evaporation into the supply air humidifier (3), to attain the desired supply air temperature.

During this time, the return air from the building is used to the pre-cooling of ambient air and also to the regeneration of the desiccant wheel. The return air is cooled by direct evaporation into the air humidifier (4) and used to cool the ambient air into the heat recovery wheel (2). This air pre-heated is heated into the heating coil (5) to reach the needed temperature of regeneration for the desiccant wheel (1). The slowly rotation of desiccant wheel (6-12 rph), assuring the transfer of adsorbed humidity from the supply air to the exit air, allows a continuous process.

Operating data Eicker [1] gives nominal points of functioning for a desiccant cooling system with typical temperature and humidity conditions (32 °C outside temperature and 40 % relative humidity). These nominal points are reported in the moist air diagram (Figure 2 ).

Table 1 - List of solar desiccant cooling systems in Europe, Delorme & al. [2] and Balaras at al. [3]

Country Place Cold power Solar collectors Operational year Deutschland Meeting room 60 Air, 100 m2 2001 Deutschland Conference room 30 Air, 115 m2 1998 Deutschland Meeting room 18 Water, 20 m2 1996 Deutschland Meeting room 18 Water, 23 m2 1997 Deutschland Exposition room 18 Air, 20 m2 1999 Deutschland Factory 108 Air, 100 m2 2000 Deutschland Installation test 24 Air and Water, 40 m2 2000 Deutschland Office, conference room 30 Water, 12 m2 2000

France Conference room 7 Water, 16 m2 2004 Portugal Office 36 Water, 48 m2 1999

Spain Library 55 Air, 105 m2 2002

235

Due to low temperature demands (60-80 °C), accessible with air-led and water-based flat plate collectors, desiccant cooling systems are particularly adapted for applications of solar thermal energy. Table 1 presents a non exhaustive list of solar desiccant cooling systems in Europe.

Figure 2 - Typical desiccant cooling process in the moist air diagram

These systems use solar energy as a hot source for the regeneration of the desiccant wheel, Figure 3.

Figure 3 - Schematic drawing of solar desiccant systems

236

The feedback on these installations has identified different key points:

Several facilities are limited by the design and hydraulic control (the removing of the hydraulic storage tank would also improve the overall performance);

Hot temperature regulation according to the demand would improve overall system performances;

An automatic and continuous monitoring is recommended in order to detect malfunctions and control system problems;

The desiccant wheels performances are not always suitable with operating conditions or climates.

In many cases, energy conservation is not as good as expected due to the consumption of auxiliary components (fans).

Technical analysis Despite the fact that all components of desiccant cooling systems are known for a long time (the heat recovery wheel and the heating coil are heat exchanger, air humidifiers are used into classical air handling unit and desiccant wheels are used for humidity control in the pharmaceutical or micro-electronics industry), they meet significant difficulties both in terms of reliability as well as performances. Two of the difficulties associated with these systems are related to the wheels dessiccantes and control units.

Desiccant wheels The desiccant wheel is a rotary sorption dehumidifier (Figure 4) based on the use of a solid desiccant. The solid desiccant (silica gel, LiCl, molecular sieves…) is coated and impregnated on a honeycomb matrix (Figure 5) made in glass, ceramic fibres, cellulose or in heat-resistant plastics.

Figure 4 – Example of a silica gel desiccant wheel integrated into a

cassette, NovelAire [4]

Figure 5 - Diagram of the airflow through a matrix honeycomb

The desiccant wheel works continuously and in two phases: the adsorption step during which humidity of the ambient air is fixed by the adsorbent and the regeneration step during which the heated return air can be used to drain water vapour stored.

The separation between regeneration air and process air is ensured by sealing strips. The sealing tape often leads to aerodynamic leaks. But the main problem is that desiccant wheels work in two opposed temperature areas (32 °C for adsorption step and 70 °C for regeneration step). These two opposed temperature areas leads to thermal losses in the structure degrading the sorption performances.

237

Table 1 gives a list of desiccant wheel manufacturers.

Table 1 - List of manufacturers of sorption dehumidifiers, Henning [5] Compagny Coutry of origin Desiccant Wheel size (m)

Munters USA US SiGel, AlTi, Silicates, New proprietary 0.25 – 4.5

Munters AB Sweden SiGel, AlTi, Silicates, New proprietary 0.25 – 4.5

Seibu Japan SiGel, AlTi, Silicates, New proprietary 0.1 – 6

Nichias Japan SiGel, Molecular sieves 0.1 – 4 DRI India SiGel, Molecular sieves 0.3 – 4

Klingenburg Germany Al oxides, LiCl 0.6 – 5 ProFlute Sweden SiGel, Molecular sieves 0.5 – 3

Rotor source US SiGel, Molecular sieves 0.5 – 3 NovelAire US SiGel, Molecular sieves 0.5 – 3

This technology comes from the market of the air-handling unit. Also, lot of desiccant wheels aren’t really adapted to solar desiccant cooling systems and their price depends strongly on their sizes and on their manufacturers. Eicker [1] specifies that the desiccant wheel represents 60% of total cost of a solar desiccant cooling installation.

Control units The control units for desiccant cooling systems have to meet several challenges. While the control unit of desiccant cooling systems is based on humidity variations, the accuracy of moisture sensors measurement is proportional to their prices. This price may vary by a factor five, depending on the desired accuracy and equipment selected. In addition, these control units have to adapt the functioning at different operating configuration (single ventilation, direct humidification, indirect humidification, humidification combined and desiccant cooling) as a function of temperatures and humidity. So, the investment cost of control units for desiccant cooling system is higher than those of conventional air conditioners with compression cooling systems, Luginsland [5].

New Developments For Desiccant Cooling In a view to optimise desiccant cooling systems, several technologies are developed both for heat and mass transfers (desiccant exchanger and liquid desiccant systems) and for the control system.

Desiccant heat exchanger In order to optimise the desiccant cooling systems, the CEA - INES, in collaboration with the CNAM - IFFI, developed a multifunctional exchanger (Figure 6) able to couple heat and mass transfers of the desiccation.

238

Figure 6- Schematic figures and photo of desiccant exchanger

The process air enters the heat exchanger by the centre and then flows in a radial direction between the desiccant bed and the periphery where the air is collected. To heat or cool the desiccant during regeneration or adsorption mode, a copper heat exchanger composed of 30 tubes of 10 mm diameter with fins is immersed in the desiccant bed. The exchanger is filled with 50 kg of Engelhard KC-Trockenperlen-N type silica gel. The particles are of spherical shape with a mean diameter of 3.4 mm. The total mass of the new prototype is 60 kg. A first commercial prototype of heat exchanger filled with adsorbent was tested, Clausse [7] and it appeared that the system is able to cool air from 33 °C (outside temperature) to 24 °C (blowing temperature) at 90% HR. Figure 7 presents a sketch of a solar desiccant cooling installation using desiccant heat exchangers.

Figure 7 - Solar dessicant cooling installation sketch

The two wheels (desiccant and heat recovery) are replaced by two desiccant exchangers (air and desiccant/water) functioning heated or cooled; the cooling corresponding to the drying step while heating is used for the regeneration. During the regeneration phase, hot water, resulting directly from solar collectors, is used. During the adsorption phase, cold water circulation, resulting for example from a cooling tower or geothermal sinks, maintains the adsorbent under optimal sorption conditions. A TRNSYS numerical model of desiccant heat exchanger was also developed to simulate the heat and mass transfer for the adsorption and regeneration processes in the prototype. The results of the numerical model are in agreement with the experimental results and systems performances could be evaluate, Demasles [8].

239

Liquid desiccant Liquid sorbent agents can also be used in desiccant cooling system, but just a small number of suitable materials can be used, due to the strict limitations that apply to hygroscopic solutions in direct contact with the environment. The main sorbents used are lithium chloride and calcium chloride. These systems are basically composed by an absorber, a regenerator and Air-solution heat and mass exchangers using to put in contact the air and the solution in counter-flow or cross-flow. The principal advantages of these systems are the ability to store cooling capacity by means of the regenerated desiccant. This form of storage is compact, requires no insulation and can be applied for indefinitely long period of time, Henning [5]. Today, this technology is at a research and development step.

Control units PSE GmbH and the Fraunhofer Institut of Freeburg have developed a standard control unit for solar desiccant cooling systems, Luginsland [6]. This control unit use different sensor types (temperature, humidity) and a modular design software, in order to adapt easily the control unit at the configuration of the installation (Figure 8).

Figure 8 - Standard control unit for solar desiccant cooling systems, Luginsland [6].

The functioning of the standard control unit developed was evaluated first on a tests hand and will be also used on a demonstration installation in Greece.

Conclusions and future prospects The desiccant cooling systems and the absorption chillers are technologies able to offer an alternative to traditional air-handling units. Even so, the desiccant cooling systems have some trouble in performances and reliability. Many developments are made, both in terms of heat and mass transfers (desiccant exchanger, liquid desiccant) and of control units, in order to optimise the operation and reliability of these systems and enable their market introduction. Not only technical features need to be reinforced regarding these technologies. Environmental and economical matters are also at stake. The EMINENT project provides an efficient framework to lead both the foot print and the economical viability analysis. Hence it allows running comparative analysis of different cooling technical solutions, and in differents countries. The results of these evaluations are of great interest and may give assets to promote these technologies.

240

In the actual context in wich emphasis will be put on technological solutions that may offer almost zero emission and a good economical ratio, public awareness, desiccant cooling systems and the absorption chillers represent a good challenge to traditional air conditioning. By evaluating these technologies on their potential merits and fields of application before offering them to the market, EMINENT will lead to increased awareness and in reduced lead-time in the transfer of results from pure research to industrial solutions.

Acknowledgements The financial support from the European Community EMINENT2 project TREN/05/FP6EN/S07.56209/019886 is gratefully acknowledged. This work is part of the “Développement d’un échangeur dessiccant pour les systèmes de rafraîchissement solaire de 10 kW” project sponsored the French Energy Agency (ADEME) in the Prebat program “Bâtiment horizon 2010”.

References [1] Eicker U. Solar technologies for buildings. Wiley editions, 2003, pp.123-200. [2] Delorme M, Reinhard S, Berthaud S, Mugnier D, Quinette J-Y, Richler N, Heu

Emann F, Wiemken E, Henning H-M, Tsoutos T, Korma E, Dall’o G, Fragnito P, Pitera L, Oliveira P, Barrose J, Ramon Lopez J, Torre Enciso S. La climatisation solaire. Rhônalpénergie-Environnement, 2004, pp.1-32.

[3] Balaras C A, Grossman G, Henning H-M, Infante Ferreira C A, Podesser E, Wang L, Wiemken E. Solar air conditioning in Europe, an overview. Renewable and Sustainable Energy Reviews, 2007, Volume 11, Issue 2, pp. 299-314.

[4] NovelAire. Desiccant dehumidification wheel. Technical information brochure, 2004, http://www.novelaire.com.

[5] Henning H-M. Solar-assisted air-conditioning in buildings. Spring Wien NewYork editions, 2004, 2nd edition, pp. 15-39 and pp. 59-115.

[6] Luginsland F, Häberle A, Hindenburg C, Wittwer C. Development of a standard control unit for solar desiccant cooling systems. Proceeding of the first international conference of solar air-conditioning, 2005, October 6th/7th, Bad Staffelstein, Germany.

[7] Clausse M, Perigaud Y, Meunier F, Boudéhenn F, Demasles H. Experimental characterisation of a novel adsorber heat exchanger for dessicant cooling applications. Proceeding of the 22nd IIR International Congress of Refrigeration, 2007, august 21th/26th , Beijing, China.

[8] Demasles H, Boudéhenn F, Clausse M. Numerical and experimental studies of a novel adsorber heat exchanger for desiccant solar air conditioning. Proceeding of the 2nd International Conference Solar Air-Conditioning, 2007, October 18th/19th, Tarragona, Spain.


241

Fixed Bed Gasification of Biomass Fuels: Experimental Results

Nikolaos Koukouzas1, Catalin Flueraru2, Anastasios Katsiadakis1, Evangelos Karlopoulos1

1CERTH/ISFTA, 4th km. Ptolemaida-Kozani, P.O. BOX 95, GR-50200, Ptolemaida, Greece. Email: [email protected]

2SC OVM-ICCPET SA, 266-268 Rahovei Street, 050912, District 5, Bucharest, Romania

Abstract A 100 kW fixed bed downdraft gasification reactor has been developed in a joint bilateral project between a Greek and a Romanian research team. The main purpose was to establish the main operational characteristics of the lab-scale unit and to examine the effect of oxygen increase in the gasification air mixture on the quality of the produced fuel gas, concerning combustion behavior. Two different types of biomass types have been used; oak wood saw dust pellets and sorghum pellets. The results demonstrated that when the oxygen content of the gasification air mixture was increased, the combustible content (CO, H2, and CH4) and heating value of the fuel gas was also increased, for both biomass types.

Introduction Gasification is the thermo-chemical process that converts a solid or liquid hydrocarbon feedstock of fossil or renewable origin to a gaseous fuel. The gasification agents commonly used include air, pure oxygen, steam or a combination of them. The fuel gas produced consists of carbon monoxide (CO), hydrogen (H2), carbon dioxide (CO2), methane (CH4), other light hydrocarbons and nitrogen (N2) (if air is used as the gasification agent) and contains impurities, such as small char particles, ash, tars and oils. The solid residue will consist of ash (composed principally of Ca, K, Na, Mg and Si oxides) and possibly some carbon or char. Although a lot of research and development work has been carried out on gasification during the past two decades, the commercial breakthrough for this technology has yet to come [1,2]. Research efforts in gasification technology have persisted, however, driven mainly by the need for cleaner and more efficient power generation technologies based on coal. In the past decade, biomass and municipal solid waste gasification have attracted increasing interested by the research community and significant advances have been made in the technology of biomass gasification and fuel gas utilization [2, 3]. Biomass gasification allows the conversion of different biomass sources to a more convenient gaseous fuel that can be then used in conventional equipment (e.g. boilers, engines and gas turbines) or more advanced and modern equipment such as fuel cells for the generation of heat and electricity. The conversion to a gaseous fuel provides a wider choice of technologies for heat and electricity generation for small to large-scale applications and significantly increases the potential of the biomass feedstock. In addition, the fuel gas could be used to produce transportation fuels, such as synthetic diesel or hydrogen.

242

The gasification reactions are mainly endothermic and thus, heat has to be supplied to the reactor. In directly heated gasifiers, the heat necessary for the endothermic reactions is provided by combustion or partial combustion of the biomass within the gasifier (autothermal gasification). In indirectly heated gasifiers, the heat is generated outside the gasifier and then exchanged with the gasifier by means of a heat exchanger or a heat carrier (allothermal gasification). Autothermal gasification reactors mostly use air as gasification medium because pure oxygen is economic feasible only in large-scale installations [1]. Gasifiers can operate at low (near atmospheric) or high (several atmospheres) pressure. Temperature and pressure operating conditions as well as residence time are key factors in determining the nature and quality of the produced fuel gas. Gasifiers have been designed in various configurations. Based on solid fuel combustion, gasification reactors can be divided into three main categories: fixed bed gasifiers (updraft and downdraft), fluidised bed gasifiers and the less established entrained bed gasifiers. Detailed reviews of gasifier options are available in the international literature. Various biomass gasification designs have been developed during the past two decades. Fixed bed gasifiers are mainly used in the small-scale range, whilst for larger scale fluidised bed gasifiers are proposed. Two different types of fixed bed gasifiers were originally developed: updraft and downdraft gasifiers. In updraft gasifiers, the gasification agent is introduced at the bottom and the fuel gas flow is upwards counter-current to the biomass, which is fed from the top. The fuel gas leaves the gasifier at the top and the ash is discharged at the bottom. Several zones are created in updraft fixed bed gasifiers, which are – starting from the top – the drying, the pyrolysis, the reduction and the oxidation zone. The temperature is increasing from the top to the bottom. The tars are produced mainly in the pyrolysis zone and leave the gasifier together with the fuel gas. Since there is no zone above the pyrolysis zone that has a higher temperature to thermally destroy the tars, a high amount of tars is expected in the fuel gas. Downdraft gasifiers try to avoid the disadvantage of high tar contents by injecting the gasification agent not at the bottom but to a certain height above the bottom. The main difference to updraft gasifiers is that the gas flows co-currently downwards with the biomass. This leads to a different order of the reaction zones from top to bottom, namely, drying, pyrolysis, oxidation, and reduction zone with the result of a low tar content in the fuel gas. The main disadvantage of this kind of gasifier configuration is the high carbon content in the bottom ash. This can be explained by the reduction zone, which is the first zone above the grate [4]. Another gasifier option is the twin-fire fixed bed gasifier, which is able to combine the advantages in one gasifier and avoid the disadvantage of both systems above. The main advantages of fixed bed gasifier configuration can be summarized as follows: Simple and reliable design; Capacity for wet biomass gasification and favourable economics on the small scale. However, the main technical challenges that have to be faced include: Long residence time; Non-uniform temperature distribution; Possible high char or/and tar contents in the fuel gas and low productivity [3].

243

Gasification fundamentals in fixed bed autothermal gasifiers Gases and solids move forward in the same direction in the downdraft gasifier and therefore different reactions occur in the different zones. In the drying zone, temperature is about 150-300 oC. Therefore, water vaporization mainly occurs in this zone. In the pyrolysis zone, temperature is about 600 oC, and pyrolysis of biomass happens and produces char, tar and gas as reaction (1) shows: Biomass → char + tar + gases (CO2, CO, H2O, H2, CH4, CnHm) (1) In the oxidation zone, because of the presence of oxygen, combustion reactions of biomass pyrolysis products proceed here to provide the required heat for the whole gasification as reactions (2) and (3) present. In all the reactions presented here, the value of enthalpy of reaction refers to temperature 298.15 K: C + O2 → CO2 ΔH = -408 kJ/mol (2) 2C + O2 → 2CO ΔH = -246 kJ/mol (3) In the reduction zone, secondary reactions of biomass pyrolysis and oxidation products proceed. Most of these reactions are endothermic and thus the temperature of reaction drops in this zone: C + CO2 → 2CO ∆Η = +172 kJ/mol (Badouard reaction) (4) C + H2O → CO + H2 ∆Η = +131 kJ/mol (Steam-carbon reaction) (5) CO + H2Ο → CO2 + H2 ∆Η = -41 kJ/mol (Water-gas shift reaction) (6) C + 2H2 → CH4 ∆Η = -75 kJ/mol (Hydrogasification) (7) CO + 3H2 → CH4 ∆Η = -206 kJ/mol (Methanization) (8) CH4 + H2O → CO + 3H2 ∆Η = +206 kJ/mol (9) According to the principles of chemical thermodynamics, the quantity of obtained products depends on the working temperature and pressure. As in any gas-solid heterogeneous reaction, the kinetics of the biomass gasification is determined by three basic mechanisms such as: diffusion of the gasification agent and of the products through the particle surrounding gas film, the diffusion of the gaseous components through its pores and the chemical reaction. At high temperatures the complex process is determined by the chemical reaction speed. The role of pores diffusion and of surrounding film diffusion increases in the same time with the temperature rise [1, 5].

Experimental section

Biomass Feedstock Two different types of biomass were used in the gasification experiments: Saw dust pellets from deciduous trees, oak wood in particular, and sorghum pellets. Oak wood

244

is widely found in mountainous regions of the Mediterranean countries. It is a fast-growing tree and is widely used in the timber and wood industry for the production of commercial lumber and furniture. Sorghum is an energy crop and represents one of the most promising solid biofuels in southern Mediterranean regions. It can be used in the production of 2nd generation liquid biofuels by the application of biomass gasification and further utilization of the produced fuel gas via the Fischer – Tropsch technique [6]. The combination of forestry and agricultural wastes for supllying power plants provides higher operational flexibility and minimizes the effect of seasonal supply variations. The smaller the biomass size, the better would be the heat transfer. The temperature would be uniform resulting in reaction taking place throughout the particle. Whenever reaction controls the gasification, the rate of reaction will be maximun and increases exponentially with temperature [7]. The wooden and agricultural biomass pellets selected for this study had a diameter of 40 to 50 mm. Biomass residues are generally powder in nature. In the present work, oak wood saw dust is pelletized before gasification in order to reduce the volume of the gasifier. The same approach was followed for the sorgum feedstock, as well. The proximate and ultimate analysis of the biomass feedstocks used, along with their heating values, are presented in Table 1. A Perkin-Elmer 2400 Series II CHNS/O analyzer allowed the estimation of carbon, hydrogen, nitrogen, and sulphur content of the samples according to standard procedures [8] while the oxygen content was calculated by difference. Prior to the ultimate analysis, the samples were completely dried in order to calculate exclusively the structural hydrogen of the biomass. The moisture and ash content was determined according to standard experimental procedures [9, 10], whilst the high heating value (HHV) was calculated by using a Parr 6300 bomb calorimeter. Experimental set up. A schematic of the experimental fixed bed downdraft gasifier used for the biomass gasification trials, along with the measurement points is shown in Figure 1. The reactor vessel is a stainless steel cylindrical tube of 20 cm diameter and 130 cm height. The output thermal power of the installation is 100 kW, based on a biomass flow rate of 25 kg/h having a calorific value of around 14,4 MJ/kg. Some other basic technical characteristics of the installation that have been established during the initial trial tests are presented in Table 2. The experimental gasification installation is composed of the following parts: Biomass bunker; Biomass dosing feeding system; Gasification reactor vessel consisting of: drying zone, oxidation and pyrolysis zone, air and steam distribution section, reduction zone and gas/solid separator composed of two cone parts; Cyclone; Ash discharge system; Steam feeding system; Air/oxygen mixture feeding system; Oil burning starting system; Measurement transducers and thermocouples; Three gas analysers and a data acquisition system. The total electricity demand of the experimental installation is 15.5 kW.

245

Table 1: HHV, proximate and ultimate analysis of the biomass feedstock.

Table 2: Technical data of the experimental fixed bed gasifier.

Biomass capacity 25 kg/h

Air/O2 flow 33 Nm3/h

Steam flow 3 kg/h

Gasification temperature 830 - 870 °C

Fuel gas flow 1,8 - 2 Nm3/kg biomass

The biomass bunker has a capacity of about 250 kg. The biomass dosing feeding system is driven by a 24 V engine, powered by an adjustable source in order to ensure the necessary speed. A 24 V engine also drives the dosing ash system. The steam feeding system is made of a pipe heated by an electric resistance. Water is pumped by an adjustable peristaltic pump, evaporated and heated at an adjustable temperature. The adjustment is assured by the gasification test rig-monitoring program and can be modified by the operator. The air/oxygen mixture feeding system assures the necessary air/oxygen preheated at an adjustable temperature controlled the same way as the steam. The oil burning starting system with 10 kW thermal output assures the installation ignition and is stopped after the biomass firing.

Component Units (as received

basis)

Pellets of saw dust from oak

wood

Sorghum pellets

Moisture % w/w 16,20 17,00

Ash % w/w 1,80 4,80

Carbon % w/w 45,60 41,30

Hydrogen % w/w 4,70 5,20

Nitrogen % w/w 1,20 1,60

Oxygen % w/w 30,50 30,10

Sulphur % w/w traces traces

kcal/kg

3945 3718 High Heating

Value kJ/kg 16517 15560

246

Ta = Air temperature, TO2 = Oxygen temperature, DP4 = Differential pressure on the diaphragm to calculate oxygen flow, Tpa = Preheated air temperature, P1 = Primary air pressure, DP1 = Differential pressure on the diaphragm to calculate primary air flow, Ts1 = Primary steam temperature, Qw1 = Primary steam flow, P2 = Secondary air pressure, DP2 = Differential pressure on the diaphragm to calculate secondary air flow, Ts2 = Secondary steam temperature, Qw2 = Secondary steam flow, P3 = Tertiary air pressure, DP3 = Differential pressure on the diaphragm to calculate tertiary air flow, P4 = Bunker pressure, n1 = Dosing feeding system speed, T1 = Oxidation zone temperature, T2 = Pyrolysis zone temperature, T3 = Gasification zone temperature, DP5 = Differential pressure in gasification chamber to calculate biomass height, n2 = Speed of the dosing ash system, P5 = Fuel gas pressure, DP6 = Differential pressure on the Venturi tube to calculate fuel gas flow, T4 = Fuel gas temperature, A1 = TESTO 350 XL portable emission analyser, A2 = H2Scan HY-ALERTA Model 500 hand held hydrogen detector, A3 = Geotechnical Instruments GA 2000 biogas analyser.

Figure 1: The lab-scale fixed bed gasification reactor configuration The measurement transducers are all produced by DWYER company. Airflows are calculated by measuring the pressure drop from diaphragms, air pressure and temperature. All the transducers are powered by a 24 V DC fixed source. The data acquisition system consists of analogical-digital converters with 4-20 mA input, 12 bits output. The system is connected to a computer by a RS232 communication

247

system. The acquired data are shown on the screen in numerical and graphical form and stored in a Microsoft Excel compatible file through dedicated software. The TESTO 350 XL gas analyser is used for determining the CO, CO2 and O2 content of the fuel gas and is coupled with a CO2 infrared cell and a dilution module in order to increase the CO measurement range. The H2Scan HY-ALERTA Model 500 is a hydrogen analyser and the GI GA 2000 is a biogas analyser used for the determination of CH4 content in the fuel gas. All the above-mentioned measurement equipment is portable, providing extra flexibility during the experimental trials. Experimental procedure The first experimental biomass gasification trials in the atmospheric pressure, fixed bed downdraft gasifier were performed in order to establish the operating conditions of the lab-scale installation. The overall aim was to produce a rich fuel gas with high combustible content. The effect of the main gasifier operating conditions, such as the reaction temperature, has been analyzed to establish the conditions allowing higher carbon conversion and fuel gas constituents (CO, H2 and CH4) concentration and production. The first tests were conducted with atmospheric air, without the addition of pure oxygen in the mixture, with the aim to monitor the whole process, regarding pressure and temperature profiles along the reactor vessel and pipelines. The preheated air and air enriched in O2 maximum temperature was set to 350 oC with an average flow of 33 Nm3/h. The overheated steam capacity was set to 3 kg/h. In the initial trials, the gasification temperature was in the range of 830 – 870 oC and the fuel gas exit temperature to around 500 oC. In all cases, the produced ash was below the melting point and therefore no slagging was observed in the ash discharge system. The experiments aimed to establish the functioning conditions of the gasification reactor and to determine the influence of the gasification agent composition on the fuel gas quality. Tests were performed with air as gasification agent and air enriched in oxygen. The modification of the gasification agent’s composition was realised by decreasing the nitrogen content by 5%, 10%, 20%, 30% and 50%, as rated to the initial nitrogen content of atmospheric air. After the initial ignition of the biomass feedstock, the gasification installation was operated in constant; steady state regime for a 4-hour interval per experimental trial, for each O2 concentration rate mentioned above. The fuel gas was drawn and analysed every 30 minutes during each experiment. The major fuel gas compounds were detected and quantified using the gas analysers mentioned previously. The N2 content of the fuel gas was determined by difference, by the application of mass balance, upon determining the participation of the other compounds of fuel gas. The fuel gas heating value was calculated by determining the fuel gas compounds, based on their volumetric composition.

Results The experimental results are presented in Table 3. The second row represents the fuel gas composition without N2. In both cases examined concerning the type of biomass, oak wood saw dust pellets and sorghum pellets, the increase of pure oxygen content in the gasification air mixture, led to an increase of the combustible

248

content of the generated fuel gas (CO, H2, CH4), thus leading to a fuel gas with higher heating value. By decreasing the nitrogen content of air by 50%, the increase in the heating value was 46% for sorghum pellets and 42.7% for oak-wood pellets, with the latter being slightly higher for all oxygen to nitrogen fractions applied. The CO content of the fuel gas increased by approximately 40% in both cases, whilst the H2 content increase was 42% for oak wood and 47% for sorghum pellets. Sorghum pellets exhibited the higher increase in CH4 content (57%) but without increasing dramatically its total contribution to the volumetric composition and heating value of the generated fuel gas. The higher increase was observed for CO2 (72%), as expected. The graphical variation of the fuel gas composition with the O2 introduced in the gasification process of oak wood pellets and sorghum pellets is presented in Figure 2 and Figure 3, respectively. The tar content of fuel gas was not examined at this stage, because the main concern was to increase the combustible content and heating value of the generated gas, rather than investigating the quality of the fuel gas concerning impurities that affect downstream utilization of the gas.

Table 3: Dry fuel gas composition and LHV as a function of the composition of air enriched in O2 and type of biomass used.

DRY FUEL GAS COMPOSITION AND LOW HEATING VALUE (LHV)

CO2 CO CH4 O2 H2 N2 LHV kcal/Nm3

Gasification agent O2 / N2 O* S** O S O S O S O S O S O S

10.9 10.6 21.0 20.5 1.8 1.4 0.9 0.2 26.3 25.1 39.1 42.2 21% O2 79 % N2 17.8 18.3 34.0 35.5 2.8 2.3 1.4 0.3 44.0 43.6 0 0

1331 1257

12.5 12.0 22.1 21.4 1.8 1.5 0.7 0.3 26.9 25.9 36.0 38.9 24,95 % O2 75,05 % N2 20.1 19.6 35.8 35.0 3.0 2.5 1.1 0.5 40.0 42.4 0 0

1376 1312

14.1 13.8 23.3 22.0 1.9 1.6 0.2 0.2 27.5 26.4 33.0 36.0 28,90 % O2 71,10 % N2 21.0 21.4 34.7 34.3 2.8 2.4 0.3 0.3 41.2 41.6 0 0

1433 1355

16.1 15.8 24.6 23.7 2.0 1.7 0.6 0.3 30.2 29.6 26.5 28.9 36,80 % O2 63,20 % N2 21.7 22.1 33.0 33.4 2.6 2.3 0.7 0.4 42.0 41.8 0 0

1538 1476

18.1 17.8 25.7 24.9 2.1 1.8 0.4 0.2 32.7 32.1 21.0 23.2 44,70 % O2 55,30 % N2 22.9 23.2 32.6 32.4 2.7 2.4 0.5 0.3 41.3 41.7 0 0

1638 1578

18.7 18.3 29.7 28.8 2.5 2.2 0.2 0.5 37.5 37.0 11.3 13.2 60,50 % O2 39,50 % N2 21.1 21.1 33.5 33.2 2.9 2.5 0.2 0.6 42.3 42.6 0 0

1899 1834

*O = Oak wood, **S = Sorghum

249

0

5

10

15

20

25

30

35

40

45

0 5 10 20 30 50

% Decrase of the N2 content of the air enriched in O2 mixture.

[% v

ol.]

of d

ry fu

el g

as.

CO2COCH4O2H2N2

Figure 2: Variation of produced fuel gas composition with the O2 content introduced

in the gasification process of saw dust oak wood pellets.

0

5

10

15

20

25

30

35

40

45

0 5 10 20 30 50

% Decrease of the N2 content of the air enriched in O2 mixture.

[% v

ol.]

of d

ry fu

el g

as.

CO2COCH4O2H2N2

Figure 3: Variation of produced fuel gas composition with the O2 content introduced

in the gasification process of sorghum pellets.

250

Conclusions – Discussion The following conclusions and remarks can be drawn from the above analysis:

• The two different kinds of biomass used in the gasification experiments demonstrated similar elemental composition and heating value. The produced fuel gas from oak wood and sorghum pellets demonstrated similar quality characteristics and both kinds of feedstock did not cause any problem in the feeding and ash discharge system of the installation, therefore; they can be used interchangeably in the same gasification system providing higher operational flexibility and minimizing the effect of seasonal supply variation.

• The initial trial tests established the functional characteristics of the gasification test unit, providing the basis for further experimental work.

• When the oxygen content of the gasification air mixture was increased, the combustible content and heating value of the fuel gas was also increased. When the nitrogen content was decreased by 50%, the CO content increased by 40% and the H2 content by 42 – 47% for both biomass types. The heating value was also raised by the same percentage.

Acknowledgements

The bilateral project between Greece and Romania under the title “Gasification of Biomass and Wastes in Existing Coal-fired Power Plants”, as part of the Joint Research and Technology Programmes 2005-2007, is gratefully acknowledged for providing financial support for this work. References [1] Pfeifer C., Hofbauer H. Development of catalytic tar decomposition downstream

from a dual fluidized bed biomass steam gasifier. Powder Technology 2008; 108: 9-16.

[2] Bauen A. Biomass gasification. Encyclopedia of energy. Vol. 1, Elsevier Inc, 2004. [3] Wang L., Weller C.L., Jones D.D., Hanna M.A., Contemporary issues in thermal

gasification of biomass and its application to electricity and fuel production. Biomass and Bioenergy 2008, doi:10.1016/j.biombioe.2007.12.007.

[4] Kramreiter R., Url M., Kotik J., Hofbauer H. Experimental investigation of a 125 kW twin-fire fixed bed gasification pilot plant and comparison to the results of a 2 MW combined heat and power plant (CHP). Fuel Processing Technology 2008; 89: 90-102.

[5] Lv P., Yuan Z., Ma L., Wu C., Chen Y., Zhu J. Hydrogen-rich gas production from biomass air and oxygen/steam gasification in a downdraft gasifier. Renewable Energy 2007; 32: 2173-2185.

[6] Fryda L.E., Panopoulos K.D., Kakaras E. Agglomeration in fluidised bed gasification of biomass. Powder Technology 2008; 181: 307-320.

[7] Kirubakaran V., Sivaramakrishnan V., Nalini R., Sekar T., Premalatha M., Subramanian P. A review on gasification of biomass. Renewable and Sustainable Energy Reviews 2007, doi:10.1016/j.rser.2007.07.001.

[8] ASTM E870-82 (2006). Standard Test Methods for Analysis of Wood Fuels. [9] ASTM E1756-01 (2007). Standard Test Method for Determination of Total Solids

in Biomass. [10] ASTM E1755-01 (2007). Standard Test Method for Ash in Biomass.


251

Mathematical simulation and optimization of forest management with consideration of carbon balance

Oleg Butusov1, Valery Meshalkin2

1 Forest Ecology Centre, Russia,117810, Moscow, Profsoyuznaya 84/32, email: [email protected]

2 RCTU, Russia, Moscow, Miusskaya pl. 9, email: [email protected]

Abstract Taking into account that carbon accumulation in the different kinds of forests might be used to evaluate carbon deposition it is clear that mathematical simulation of natural and anthropogenic dynamic of forest is important means for carbon balance forecasting. Carbon fluxes might be calculated using conversion coefficients. It is well known that carbon fluxes accounts are stipulated in Kyoto Protocol and Framework Convention on Climate Change. The forest areas are one of the important reserves of the carbon balance. Wide use of energy saving and resource-conserving technologies permit to retain the significant areas of forest stock and by this lead to increase of carbon deposition in forests. Main economic feature of forests consist in that they are renewable sources of raw material.

Introduction Mathematical simulation and optimization of forest management under condition of more wide usage of energy saving and resource-conserving technologies might be viewed as main base for natural resources conservation [1-6].To advance in this important direction we developed mathematical and computer model of forest stock dynamics under different kinds of forest management. Our model takes into account the following natural and anthropogenic factors: forest fires, burned places restoration, forest cutting, raw material consumption, species and ages forest structure, dead wood destruction, industrial pollution influence and succession. To simulate carbon fluxes we use conversion coefficients. The influence of energy saving and resource-conserving technologies is taking into account by changing of planned values of cutting areas. The model lets to take into account the carbon balance change under condition of carbon credit that is to forecast after-effects of artificial forestation of new areas with the purpose of additional carbon deposition. The basic feature of the nonlinear dynamic approach used in our model is statistical-regression identification of the model equations coefficients for predominant species in different kinds of forests. In our model we used afforestation data base for Vologda region. We used also the assumption that for the species interaction might be used only tree main predominant species which are typical for Vologda region forests namely: spruce, pine and birch. We used the assumption that nonlinear interactions between species might be simulated in the quadratic approach. The model

252

coefficients were calculated using statistical-regression identification. The main idea of our work lies in assumption that at the first step of modeling the differential equations coefficients might be found from statistical data and at the second step the resulting equations might be used for forecasting of predominant species dynamics in different kinds of forests and under different kinds of forest management. At the same time on the phase portrait might be singled out critical points and might be examined the ecosystem behavior in these points. As a result of statistical identification we received the differential equation coefficients for each kind of forests and calculated the graphs of the dynamics of forest area percents for each predominant species. Using mathematical modeling we undertake optimization study of different kinds of forest management. The criterion function of optimization is maximization of timber cutting under condition of preservation in the stationary limit the prescribed species and ages forest stock structure. Each forest model of cyclic cutting period dynamics should take into account three main processes: succession redistribution of forest areas, cuttings and burned areas regeneration time and regeneration redistribution of forest areas.

The succession redistribution model The model equations are

3210,i319,i218,i237,i

226,i

215,i34,i23,i12,i1,i

i XXAXXAXXAXAXAXAXAXAXAAdt

dX+++++++++=

where 1 2 3, ,X X X - forest area percents for which main predominant species are spruce, pine and birch; index i takes the following means: 1 – spruce, 2 – pine, 3 – birch, 4 – aspen, 5 – alder. The model equations coefficients were calculated using statistical-regression identification. These equations were used for closed cutting period and in that way we should take into account the forest renewal. Thereupon consider the problem of cutting or burned places renewal.

Cuttings and burned areas regeneration time model To solve the problem we used the following steps. First step – choice of the functional dependence for which we solve the problem of the least square (LSQ) approximation and second step - least square data processing. On this stage we used three kinds of modeling: step approximation, exponent and logistic approximation. Step approximation. In this case the single parameter is the renewal time - Tz. To calculate renewal times we used the following algorithm

253

1( )

p

a t at

Arg p R S−=

= >∑ 1

1

p

a a tt

a p

S Rdt

R

−

−=

−

−=

∑

(1) 1Tz p dt= − +

where Sa – the total cutting areas known for each 5 years from accounting documentation; Rt – cuttings for t-th – year; a – year for which burned and cutting areas are known. The second equation of (1) is linear interpolation intended for renewal time definition inside year period. Exponent approximation. In this model regenerating cutting areas are distributed by exponent probability law. In this case the total afforestation area for the p-th year from the beginning of cuttings are

( 1) ( 2)1 2 1

p pp pI R e R e R eα α α− − − − −

−= + + L (2) The coefficients were calculated using LSQ

( )2( ) minp p

pQ I Sα = − →∑ (3)

The problem might be reduced to

( ) 0,pp p

p

dII S

dα− =∑ (5)

( 1) ( 2)1 2 1( 1) ( 2)p p p

p

dIR p e R p e R e

dα α α

α− − − − −

−= − + − + +L

Logistic approximation. In this model regenerating cutting areas are distributed by exponent probability law. The total afforestation area for the p-th year from the beginning of cuttings might be calculated using following equation

1 2 1( 1) ( 2) (1)p pI R S p R S p R S−= − + − + +L (6)

where 0

0 0

( )( ) at

S bS tS b S e−=

+ −.

254

Regeneration redistribution of forest areas model For each forestry and for each forest area types we have the area distribution for mature and young forests. Let us take the following notations: Mp – areas of young forests for different species ( 1 5p = ÷ ); Sp – areas of mature forests for different

species ( 1 5p = ÷ ); pqA - transfer coefficient from p-th mature species to q-th young species. The problem is to calculate transfer regeneration matrix. First of all we normalized the areas values

5 5

1 1

,p pp p

r rr r

M Sm s

M S= =

= =

∑ ∑

The model equations are

1 1 2 2 3 3 4 4 5 5p p p p p pA m A m A m A m A m s+ + + + =

1 1 2 2 3 3 4 4 5 5q q q q q qA s A s A s A s A s m+ + + + = 5 5

1 11, 1pr rq

r rA A

= =

= =∑ ∑

The first equation describes the after cutting redistribution of p-th mature species over all regenerating areas. The second equation describes the after cutting formation of q-th young species from all mature areas. Other two equations describe the normalization condition. So the problem we need to solve contains n2 unknown quantities and only 4n equations. In this case we applied these equations to different forestry making the line problem overdetermined. The resulting problem was solved using LSQ-algorithm.

Simulation results Numerical experiments were carried out for different renewal times, stock targets and percents of forest fires areas. The example of resulting graphs is shown in Fig.1.

Conclusions Stationary states were reached after 170 years. After that period stable distribution of forest areas was settled. Similar results we but with slightly different distribution were received for another values of input parameters. This clearly shows that it is possible to reach necessary level of cutting turnover maintaining required carbon balance.

(7)

(8)

(9)

(10)

255

a b

c d

Figure 1. Simulation results (renewal time – 5 years, stock target – 92150000 m3, percent of forest fires areas – 5%): a – species areas dynamics, b - species wood

stocks dynamics, c – cut areas dynamics, d - cut wood stocks dynamics.

Acknowledgments The presentation was carried out under financial support of Trans European teaching and educational Project TEMPUS-TACIS JEP_26045_2005, ECORSE "Ecological and Resource Saving Engineering".

References 1. Butusov O. Air pollution computer aided city monitoring system // Proceedings of the ISPRS Commision VII Symposium "Resources and Environmental Monitoring. September 26-30, 1994. Rio de Janeiro, Brazil: INPE,1994:274-280. 2. Butusov O. Integral Indices as Ecosystem Health Measuring Means. Ecosystem Halth and Medicine. Ottawa, Canada: University of Guelph 1994:18-19. 3. Butusov O. Industrial sources air polluted zones in a city. Pollution in Large Cities. Proceedings. Padowa: Padova Fiere. 1995:241-250. 4. Butusov O, Meshalkin V. Computer-aided monitoring sys tem for city air pollution investigations. Pollution in Large Cities. Proceedings. Padowa: Padova Fiere. 1995:283-291.

256

5. Butusov O. Special integral indices for economic and ecological assessments in industrial regions. Internationales forum fuer informatisierung IFI-95. Des 1. - Anzahe, Baende: IAIA p.156-164. 6. Butusov O, Meshalkin V. New approach to the exposure-response study of the industrial impact on forest. Proceedings of the 15-th International congress of chemical and process engineering (CHISA-2002). Praha:Process Engineering Publishers, 2002:126-127.


257

Energy-Saving Issues in Reactive Distillation Schemes

Alexandra E. Pleşu1, Jordi Bonet2, Cristian I. Ciornei1, Valentin Pleşu1, Grigore Bozga1, Maria-Isabel Galan2

1University POLITEHNICA of Bucharest, Department of Chemical Engineering, Centre for Technology Transfer in the Process Industries,

1 Polizu St., Building S, Room S-001, RO-011061, sector 1, Bucharest, Romania; tel. +40 21 4023808, e-mail: [email protected], [email protected],

[email protected], [email protected] 2University of Barcelona, Department of Chemical Engineering,

Marti i Franques 1, floor 6, E-08028 Barcelona, Spain; e-mail: [email protected], [email protected]

Abstract Minimization of carbon dioxide emissions is one of the main priorities towards sustainability. Most of the energy consumed nowadays is obtained from fossil fuels and it is predicted that this trend will be maintained in short and medium term. The decrease of the energy demand for the energy intensive processes and the capture of carbon dioxide are required to reduce the environmental impact of fossil fuels. Distillation is a largely used separation operation, but in the same time is intensive energy consuming; more than half of the process heat distributed to plant operations ends up in the reboilers of distillation columns. The process intensification by combining a reactor with a distillation column (reactive distillation) opens new opportunities to save energy. The advantages of the use of reactive distillation are highlighted in illustrative examples.

Introduction In this study, main aspects regarding the reduction of energy consumption and CO2 emissions by reactive distillation are stressed. An efficient way that leads to CO2 avoidance is the application of advanced process integration methodologies, as presented by Klemeš et al [1] and Varbanov et al [2]. A great number of alternative schemes and operating conditions to perform chemical reactions and separations can come up in any chemical process. Most of the energy used in the chemical industry is consumed by the distillation columns and it represents most of the operating costs of the plant. Distillation columns need huge amount of energy because of the evaporation steps involved. Typically, more than half of the process heat distributed to plant operations ends up in the reboilers of distillation columns. By this, high-level energy is fed at the base of the column and about the same amount of energy is released at the top, unfortunately at a much lower temperature level. The difference between the two Gibbs energies can be seen as the necessary energy investment to reverse the mixing entropy and to separate the components of a given feed by a distillation process. Often, the energy feed cannot be used for heat integration, but is discharged to the environment. Therefore, a first selection of the separation scheme required for a system can be done according to the energy requirements of the distillation columns, in an early stage design. The selection of the right scheme determines the energy efficiency achievable by designing and optimizing the process. A wrong initial selection or if some promising process

258

schemes are not taken into account, leads to an increasing difficulty of improvement as the plant is designed in more detail. The first selection of promising alternatives can determine the success or failure of a plant in a world every time more competitive. The requirement of a higher energetic efficiency is not only economical, but it is also social, in the respect of a deep conscience about protecting the environment and using efficiently the resources available. The integration and intensification of processes such as the reactive distillation can lead to great advantages related to the increase of the reaction conversion and energy saving. Process intensification is of constant concern as includes a series of strategies aimed to develop more effective, cheaper, lower impact on environment equipment. Reactive distillation combines in the same equipment reaction and separation by distillation, representing state of art in process intensification. Main advantages of reactive distillation relate to the influence of reaction on separation and vice-versa: more compact equipment, less pieces of equipment in the plant, increased conversion overcoming reaction equilibrium limitations, improved selectivity, energy saving, energy integration, elimination of solvents, reduced by-product formation, etc. In reactive distillation, the reaction products are continuously evacuated by distillation shifting the equilibrium to the formation of products. A total conversion of the reactants can be achieved avoiding the recycling streams. The continuous evacuation of products from the reaction media can increase the reaction kinetics and avoids some side reactions that could produce fouling, contributing to the increase of the efficiency and diminishing the maintenance needs. The improvement of the reaction performance determines less severe operating conditions (pressure, temperature and quantity of catalyst needed), increasing the life of the construction materials. When the reaction is exothermic, the heat generated is directly used for the energetic requirements of the distillation column, saving energy. On the other hand, the boiling point of the mixture permits to keep the temperature constant along the column, avoiding the appearance of hot spots. The catalyst can be distributed strategically throughout the column, in this way decreasing the quantity of catalyst needed for the operation and prolonging its life, by avoiding the poisoning with side products and heavy components contained in the feed, when it is place above the feed location. The combination of reaction and separation operations within the same equipment avoids the usage of the reactor from the process scheme and decreases the number of distillation columns. The elimination of the recycling streams and units leads to energy savings and contributes to an obvious reduction of the plant volume and of storage necessities, as well as the number of pumps, pipes and measuring tools associated. Most of the costs associated to an equilibrium reaction performed in a traditional configuration are not related to the reactor, but to the sequence of distillation columns used for the purification of the products and for the recycling of the non-reacted species. The energy requirements of the distillation sequence increase according to the purity requirements. The decrease of the energetic consumption contributes to economize natural resources and attenuate the environmental impact of the industrial activity. Reactive distillation can also contribute to the ease of separation by avoiding the formation of an azeotropic mixture of the reactants, offering the appropriate media in which the azeotrope forming components react, yielding the desired product with less separation efforts, which involves less energy used by reboilers and condensers.

259

Nevertheless, reactive distillation is not suitable to all the existing processes, for instance it would not be worth to be applied in the case of a reaction with a high conversion and slow kinetics. Therefore, an early-stage analysis must be performed in this respect and accordingly, an easy to use and fast method must be applied in order to evaluate a high number of alternatives and operating conditions for energy-saving. Examples will illustrate the above detailed ideas.

Evaluation strategies Process intensification opens new improvement opportunities for nowadays processes and provides new profitable revalorization alternatives for residual streams. The infinite/infinite (∞/∞) analysis is a tool to determine the flow rates and compositions of the process streams, without requiring the number of stages, reflux or distillation column profiles (Petlyuk and Avet'yan [3], Bekiaris and Morari [4], Ulrich [5], Bonet et al. [6], Plesu [7]). ∞/∞ analysis is a framework that predicts possible product paths of a distillation column based on residue curve map information only and allows studying distillation systems, checking the feasibility, detecting multiple steady states and getting the influence of the distillate flow rate and recycle streams on purities in early stage of process design. A product path is obtained by the continuation of solutions found by varying, for example, the distillate flow rate D from 0 to its maximum, the feed flow rate F. In general, the product path can be generated by either a series of case studies using a distillation column model implemented in a commercial simulator such as AspenPlus, or by a “continuation” of solutions using a simplified model or a rigorous model as implemented in AspenPlus. ∞/∞ analysis is also useful to evaluate new alternatives for an existing process and to propose energy efficient alternatives by process integration. An approximation used in the literature is that the energy required in the reboiler is proportional to the distillate flow rate determined by the analysis. As mentioned previously, most of the energy consumed by the distillation columns is due to the reboiler; a vapour flow rate proportional to the reboiler duty and constant along the column (according to the McCabe-Thiele hypothesis) is collected in the condenser. Applying a mass balance at the condenser and using the definition of reflux ratio, the heat in the reboiler is proportional to the distillate flow rate, reflux, heat of vaporization of the mixture and the heat generated by the chemical reaction. As a first approximation, only the distillate flow rate is considered (Eq. 1). The heat of vaporization is not constant along the column and the reflux is also a relevant parameter, therefore the main results obtained from the ∞/∞ analysis must be followed by more rigorous calculations.

( ) DQrDQVQ RRreboiler ≈++⋅⋅=+⋅= 1λλ (1)

The heat in the reboiler can not be determined by the ∞/∞ analysis, therefore the rigorous simulation of the systems is required. A distillation column of one hundred stages with the feed plate in the middle is used to determine the minimum reflux: for the case of distillation columns with a large number of stages, the representation of reflux, number of stages and feed stages provides a flat minimum reflux zone (Bonet et al. [8]). The minimum number of stages is estimated by integrating the inverse of the modulus of the vector between the liquid and vapour composition in equilibrium, along the column profile composition. From the simulation results of the large distillation column, the reboiler duty is estimated according to an optimum reflux equal to 1.2 times the minimum reflux. The energy cost for each system is evaluated

260

considering the cost of a coal thermal plant with CO2 capture (Klemeš et al. [9]), which would be able to satisfy the energy demand for the selected process when it would be implemented at international scale. For the estimation of the process investment cost, the assumptions being used are formulated according to the optimum, so that the operational costs are similar to the investment costs and the investment costs are proportional to the quantity of steel needed to construct the distillation column, according to the procedure described by Bonet et al. [10]. As illustrative examples, the methyl acetate/methanol residual stream from the PVA industry (Bonet [11]) and the production of TAME by the catalytic distillation of methanol with isoamylenes (Plesu [7]) are chosen. The ∞/∞ analysis for the methyl acetate revalorization (Bonet et al. [12], Bonet et al. [13]) and the TAME synthesis by catalytic distillation (Plesu [7]) are used as a basis for the rigorous simulation of the reactive distillation columns and for the environmental and economical evaluation according to Klemeš et al. [9].

Methyl acetate revalorization case study The polyvinyl alcohol (PVA) is a polymer with environmental friendly properties: non-toxic, non-flammable and biodegradable by adapted micro-organisms (Chiellini et al. [14]). The PVA is the most widely produced water soluble and biodegradable polymer worldwide (Ozaki et al. [15]) and its market is in expansion. The quantity of PVA produced in 2010 is extrapolated that could be around 3.25 millions of Tm/year. The main limitation of the process is that 1.68 Tm of a residual azeotrope of methyl acetate and methanol is obtained for each 1 Tm of PVA. The MeAc/MeOH mixture is known industrially as MM80 because it contains about 80% mass of MeAc. Methanol is a raw material for the production for PVA, but the presence of the azeotrope makes its separation and recirculation difficult. The cost of the methanol is low (7 €/kmol) but, although the azeotrope could be broken, the methyl acetate is not well commercialized as a by-product, ethyl acetate being preferred. Hence, the methyl acetate from the PVA industry is a residue to be converted to other commercial product. Nowadays, the more widespread revalorization for the PVA residue is the hydrolysis to acetic acid (46 €/kmol). The value of the acetic acid is not high, but the process provides a higher recovery of methanol. Fuchigami [16] proposed the use of a reactive distillation column to improve the performance of the process. Another alternative comprises the reaction of the methyl acetate with ethanol (33 €/kmol) to produce ethyl acetate (115 €/kmol) (Steinigeweg and Gmehling [17], Chemical market report [18]), using several enhanced distillation strategies, such as extractive distillation, pressure swing and pervaporation. Generally, the MeAc transesterification with alcohols produces acetates of higher value than the acetic acid. The methyl acetate transesterification with bioethanol to produce directly ethyl acetate can be seen as a process intensification in comparison with the nowadays solution of hydrolyzing the methyl acetate to acetic acid and selling the acetic acid as raw material to produce ethyl acetate. For the case of reactive pressure swing system, a pressure of 8 bar is used for the reactive distillation column (Bonet et al. [19]). The processes are compared in basis of the CO2 emissions and costs.

261

TAME synthesis case study The applicability of the ∞/∞ analysis as a tool for an early oriented energy consumption minimisation is extended to the TAME synthesis, which is used as illustrative example due to its importance on providing environmental friendly gasoline additives. TAME is formed by the etherification of isoamylenes (2M1B and 2M2B) with methanol in the presence of inert components. The TAME reactions have been shown to be reversible and fairly exothermic. Several side-reactions can occur, among which the izomerization reaction between the two isoamylenes is the most important.

MeOH + IA TAME

r+1

r-1 (3) The computation procedure is similar to the previously analysed example and it is applied to a worldwide oil processing capacity of 1010 litres/day, 20% of it being transformed to gasoline subjected to further reformulation. It is considered that the end-of-pipe gasoline contains 11% TAME. The analysis is applied to different distillation column configurations both non-reactive and reactive (Figure 1). A traditional system consisting of reactor and distillation column, as well as a totally reactive and a hybrid reactive distillation columns are evaluated and compared in terms of energy savings, with the aim to collect pure TAME at the end of the process and to recycle the non-reacted components (Figure 1a). Stoichiometric and non-stoichiometric feeds are investigated at 1 and 10 bar. The distillate flow rate is calculated easily from the ∞/∞ analysis, but the heat in the reboiler can not be determined, therefore the simulation of the systems is required. The rigorous simulation of these situations, at their optimal distillate flow rates obtained from the ∞/∞ analysis, is performed to rank the results according to the energy required by the column reboiler.

(a) (b) (c) Figure 1. Traditional system of reactor and distillation column (a); entire reactive

distillation column (b); hybrid reactive distillation column (c).

Results for hydrolysis of methyl acetate to acetic acid The basic case which will be used as reference is the hydrolysis of methyl acetate performed in a reactor followed by the separation of products using a train of distillation columns (Figure 2a). For the first column, the azeotrope is recovered in the distillate and is recycled, the acetic acid obtained is collected in the bottoms of the second distillation column, whereas methanol is obtained in the distillate of the third distillation column; water is recovered in the bottom stream B3 and is recycled to the reactor. The table next to the Figure 1a shows the ratio of column distillate flow rate versus the residual stream flow rate, minimum number of stages, minimum reflux and the reboiler duty corresponding to 1.2 times the minimum reflux for each column. The energy required for the entire residual stream produced worldwide requires a total of 1630 MW. If it is assumed that this quantity of energy is produced in a coal fired plant with CO2 capture, the corresponding energetic costs associated to a CO2

262

capture of 95% will be of 370 MUSD/year according to the techno-economic modelling presented by Klemeš et al [9].

Column ID

D/Fc rmin Nmin

Q reboilers (MW)

1 3.59 0.86 10 698 2 3.68 0.69 12 713 3 1.00 1.03 17 219

Figure 2a. Hydrolysis of methyl acetate. Reference system.

Considering the approach which specifies the energetic requirements and the investment requirements equal for the reference case, then the total costs involved will be two times the energetic costs: 740 MUSD/year. The benefits produced by the methanol recovery and acetic acid obtained are around 4850 MUSD/year, with a positive balance of 4110 MUSD/year. There are two more alternative processes to the reference system: (1) recovering the acetic acid in the bottoms of the first column, the azeotrope in the distillate of the second column and the third column remaining unchanged or (2) collecting the acetic acid in the bottoms of the first column, the water in the bottoms of the second column, the methanol in the bottoms of the third column and the azeotrope in the distillate of the third column. Assuming the energy cost proportional to the entire system vapour flow rate in the condenser and the column investment cost proportional to the vapour flow rate, minimum number of stages and operating pressure, then the total investment costs for each of the above-mentioned systems analysed are calculated taking the reference system as a basis. In these cases, the reflux of the first distillation column was found to be higher than for the reference process and the overall distillate flow rates were obtained higher as well. These results lead to the conclusion that these system alternatives are less profitable, with a positive balance smaller than the reference system: 3860 and 1650 MUSD/year respectively. Fuchigami (1990) [16] proposed the use of a reactive distillation column at total reflux to decrease the recycle flow rates. The methyl acetate reacts completely inside the reactive distillation column and a mixture of the other components is obtained in the bottoms of the column. The result has been verified by rigorous simulation, using a high excess of water (5.5 times the stoichiometric quantity). In this case, there are less distillation columns required in comparison with the reference system of reactor followed by distillation columns. The benefits obtained for this alternative are 64% higher than the ones provided by the reference system. Figure 2b illustrates the process of methyl acetate hydrolysis using the reactive distillation column. The methyl acetate/methanol azeotrope can be broken as well by several other enhanced distillation strategies, such as extractive distillation, pressure swing and pervaporation. These alternatives consume lower amounts of energy than the reference system. For instance, the energy consumption for the extractive distillation using butyl acetate as extractive agent is around 1300 MW, whereas the pressure swing consumes 400

263

MW, while the use of a pervaporation membrane for the feed stream requires 300 MW compared to the 1630 MW for the reference system. This option is not used industrially because the quantity of methanol recovered is small and the methyl acetate remaining has not a good market. The income obtained by selling the products is of 309 MUSD/year compared with the 4847 MUSD/year for the hydrolysis. This fact leads to no benefits obtained from the enhanced distillation processes.

Column ID

D/Fc

rmin Nmin Q reboilers (MW)

1 0 Total reflux 44 418 2 1.00 5.58 19 645 3 3.50 0.84 15 716

Figure 2b. Methyl acetate hydrolysis by reactive distillation (Fuchigami et al.[16]).

Results for transesterification of methyl acetate to ethyl acetate The synthesis of ethyl acetate offers higher potential benefits than the transformation of methyl acetate to acetic acid. Ethyl acetate is a valuable industrial solvent and its synthesis from a residual stream and bioethanol represents an environmentally friendly alternative. The main difficulty associated is due to the equilibrium-type reaction with a low conversion to products. Moreover, a mixture difficult to separate, containing four azeotropes, is collected from the reactor: methanol/methyl acetate, ethanol/ethyl acetate, methanol/ethyl acetate and ethanol/methyl acetate. The investments in large trains of distillation columns and the high energy requirements make this alternative clearly rejectable.

Column ID

D/Fc

rmin

Nmin

Q reboilers

(MW) 1 3.12 1.48 4 856 2 1.65 1.16 13 334 3 1.00 0.46 15 170 4 1.93 14.78 24 2405 5 1.94 1.29 15 415

Figure 2c. Transesterification of methyl acetate with ethanol using enhanced distillation followed by reactive distillation.

The reactive distillation seems a good choice for this system, because the chemical equilibrium limitations are overcome and some of the azeotropes formed by reacting components disappear: ethanol/ethyl acetate and ethanol/methyl acetate. A first alternative is the use of the extractive distillation to break the methyl acetate/methanol azeotrope; the methyl acetate obtained is fed to a reactive distillation column to collect ethyl acetate in the bottoms and the azeotrope ethanol/methyl acetate in the distillate (Figure 2c). This process reports benefits, but lower than for the reference process. A high reflux of 14.78 and a tall column are necessary to obtain a purity of 0.9999 for methyl acetate in the distillate of the fourth column. The total energy requirements are 4180 MW for this process, which leads to energetic costs of 800 MUSD/year using a 95% CO2 capture. The investment costs

264

are estimated to be around 4300 MUSD/year, value which leads to high total costs of 5100 MUSD/year. The high profitability of the ethyl acetate synthesis (7930 MUSD/year) is consequently diminished, the final income being of 2830 MUSD/year. The series of distillation columns used to break the azeotrope can be substituted by other enhanced distillation strategy. The extractive distillation process uses high amounts of extractive agent, which would dilute the reaction media if the reaction is implemented in the separation process. The pressure swing distillation does not require the use of external components to the system and it is feasible to introduce the reaction directly within the enhanced distillation system. This procedure is more direct than breaking first the azeotrope and feed it to a reactive distillation column. The combination of the reaction within the pressure swing process requires only two columns for breaking all the four azeotropes: two of them are eliminated due to the reaction and the others form a boundary line which is crossed by the pressure swing. Figure 2d illustrates the system configuration for the pressure swing coupled with a reactive distillation column. The vapour-liquid equilibrium data for the ethyl acetate/ethanol mixture implemented in the AspenPlus simulator database influence the results, but however, the reactive distillation column requires a high number of stages and a high reflux, which make the process economically not viable.

Column ID

D/Fc

rmin

Nmin

Q reboilers

(MW) 1 3.35 12.12 58 3303 2 2.35 1.58 15 428

Figure 2d. Transesterification by reactive pressure swing distillation.

Results for TAME synthesis case study The reactor and distillation columns characteristics obtained from the rigorous simulation of the analysed configurations are presented in Table 1. The entirely reactive distillation column (Figure 1b) proved to be unable to collect pure TAME because it decomposes according to the chemical equilibrium reaction; on the other hand, the simulation results showed that the use of a hybrid column with non reactive stages in the stripping section overcomes this limitation (Figure 1c). The increase of the pressure in the traditional system of reactor and column does not show any advantage: it requires a compressor, thicker walls and more stages in the distillation column and higher recycle flow rate. This corresponds to higher energy demand in the reboiler and higher costs. At 1 bar, the stoichiometric feed to the reactor leads to a lower energy consumption than the non-stoichiometric one, but it requires the usage of a higher number of stages. The process intensification achieved by the hybrid reactive distillation column demands a shorter column, and in the same time, the recycle streams are minimized or even avoided. The heat of the reaction is used directly in the column, decreasing consequently the reboiler duty. The reactive distillation is able to overcome the chemical equilibrium limitations and it is the most advantageous alternative.

Table 1. Results for the TAME synthesis

265

Distillation System

P (bar)

D/Fc

rmin N Q reboilers (MW)

Energy cost (MUSD/year)

Total cost (MUSD/year)

Stoichiometric feed

1 0.4032 0.54 42 359 68 137

Non stoichiometric

feed

1 0.4118 0.73 19 365 69 105

Stoichiometric feed

10 1.2267 2.16 36 1620 281 15347

Non stoichiometric

feed

10 1.1638 0.70 38 800 143 4274

Hybrid reactive

distillation

1 0.0644 14.63 15 328 63 100

Conclusions A simple method is used to evaluate several alternatives for two illustrative examples. The process intensification can provide more advantageous solutions than the classical process structures of reactor followed by distillation columns. However, the process in which less distillation columns are used is not necessarily the most profitable. The transesterification of methyl acetate with ethanol is a clear example in this respect: breaking the azeotrope first in an enhanced distillation process and feeding the pure methyl acetate to a reactive distillation column provides more economical alternatives and more energy saving opportunities than other simpler systems in which the reaction is directly integrated within the enhanced distillation column. The TAME synthesis example shows that an entirely reactive distillation column is not suitable due to the TAME decomposition, but if a non reactive stripping section is used, then the resulting hybrid reactive distillation column is the most advantageous alternative. From the results obtained by rigorous simulation, it can be concluded that the recommended configuration among the systems analysed for TAME synthesis is the hybrid reactive distillation column, which provides an energy saving in the reboiler of 10%, in comparison with the traditional system at 1 bar from which the azeotrope methanol – isoamylene is collected in the distillate.

References [1] Klemeš J, Dhole VR, Raissi K, Perry SJ, Puigjaner L. Targeting and design methodology for reduction of fuel, power and CO2 on total sites. Applied Thermal Engineering 1997;17(8–10):993–1003. [2] Varbanov P, Perry S, Klemeš J, Smith R. Synthesis of industrial utility systems: cost-effective de-carbonisation. Applied Thermal Engineering 2005;25(7):985–1001. [3] Petlyuk F, Avet'yan, V. Investigation of the Rectification of Three-Component Mixtures with Infinite Reflux, Theor. Found. Chem. Eng. 1971;5(4):499-507. [4] Bekiaris N, Morari M. Multiple Steady States in Distillation: infinite/infinite Predictions, Extension, and Implications for Design, Synthesis, and Simulation. Ind. Eng. Chem. Res. 1996;35:4264-4280. [5] Ulrich J. Operation and control of azeotropic distillation column sequences. PhD Thesis. ETH No. 14890, Zurich, Switzerland, 2002.

266

[6] Bonet J, Thery R, Meyer XM, Meyer M, Reneaume JM, Galan, MI, Costa J. Infinite/infinite analysis as a tool for an early oriented synthesis of a reactive pressure swing distillation. Computers & Chemical Engineering 2007;31(5-6):487-495. [7] Plesu AE. Modeling and Simulation of Continous Catalytic Distillation Processes. Case Study: Tert-Amyl-Methyl-Ether (TAME) Synthesis, PhD Report, University POLITEHNICA of Bucharest, 2008, romdoc.upb.ro/record/512, accessed 20 February 2008. [8] Bonet J, Galan M-I, Costa J, Meyer, X-M, Meyer M. Multicomponent rectification: Representation of number of stages as function of reflux ratio. ECCE-6 Book of Abstracts, 2007; 2:553-554. [9] Klemeš J, Bulatov I, Cockerill T. Techno-economic modelling and cost functions of CO2 capture processes. Computers and Chemical Engineering 2007;31:445–455. [10] Bonet J, Galan M I, Costa J, Thery R, Meyer X, Meyer M, Reneaume J M. Pressure optimisation of an original system of pressure swing with a reactive column. Institution of Chemical Engineers Symposium Series 2006; 152: 344-352. [11] Bonet J. Contribution à l’étude de la transestérification de l’acétate de méthyle par distillation réactive, PhD Thesis, University of Barcelona, Barcelona, Spain, 2006. www.tesisenxarxa.net/TDX-0618107-122057/, accessed 20 February 2008. [12] Bonet J, Thery R, Meyer X, Meyer M, Gerbaud V, Costa J, Galan M. Discrimination of reactive distillation processes via residue curve maps: application to a transesterification reaction. ESCAPE-14: European Symponium on Computer Aided Process Engineering - Lisbon, Portugal, May 16-19, 2004. [13] Bonet J, Thery R, Meyer X-M, Meyer, M, Gerbaud V, Galan M-I, Costa J. The revalorization of the residual methanol/methyl acetate azeotrope produced by the polyvinyl alcohol industry. 1st international conference on engineering for waste treatment, 2005; 1:185. [14] Chiellini E, Corti A, D’Antone S, Solaro R. Biodegradation of poly(vinyl alcohol) based materials. Progress in Polymer Science 2003; 28:963-1014. [15] Ozaki SK, Monteiro MBB, Yano H, Imamura Y, Souza MF. Biodegradable composites from waste wood and poly(vinyl alcohol). Polymer Degradation and Stability 2005; 87:293-299. [16] Fuchigami Y. Hydrolysis of methyl acetate in distillation column packed with reactive packing of ion exchange resin. Journal of Chemical Engineering of Japan 1990; 23(3):354-359. [17] Steinigeweg S, Gmehling J. Transesterification processes by combination of reactive distillation and pervaporation. Chemical Engineering and Processing 2004; 43: 447-456. [18] Chemical Market Reporter, Prices tables. Week ending January 7, 2005 [19] Bonet J, Galan M-I, Costa J, Thery R, Meyer X-M, Meyer M, Reneaume J-M. Pressure optimisation of an original system of pressure swing with a reactive column by a modified boundary value method. Proceedings of the 7th World Congress of Chemical Engineering, 2005: 231.


267

Reliability and maintenance software in energy generation and saving

László Sikos, Jiří Klemeš

EC Marie Curie Chair (EXC) “INEMAGLOW”, Department of Computer Science FIT, University of Pannonia, Egyetem u. 10, Veszprém, H-8200, Hungary


Abstract The paper outlines several important issues related with energy generation and potential saving – availability, reliability and maintenance. Information about various software tools have been gathered and a brief overview drawn. The present offer of various software tools has been rather wide. However, the most of the information available has been driven by marketing and sales features. More detailed analysis with possible testing has been planned for the near future and this should provide deeper and more applicable recommendations.

Introduction Many research papers have been devoted to tasks of optimisation energy genera-tion, conversion, transmission, exchange, integration and utilisation. However, they are some additional issues which have important influence on quality and quantity of the energy supply. One of them comes from the family of issues as availability, reli-ability and maintenance. The paper has explored up to date software tools which could help to assess and improve all those issues. The study is still going on both in the scope, details and testing and will be presented later this year.

Software evaluation – an overview Numerous software packages are available which provide similar features and capabilities. In several cases demo versions are available upon request (after filling a form or via e-mail). The installation of some trial versions are password protected. The demo or trial version of a reliability software is either available on a demo CD, can be downloaded from the internet or accessed online as a web-based application. For some tools are available live web demonstrations (tours) and brochures. They include even the main features of the package. Some features (e.g. saving, printing) are often disabled. There is another demo type where the product is fully featured but there are some limitations on the number of spares or the value of different parameters. The day limits 14, 30, 45 or 60 are common for trial versions. This is enough to make a good decision to choose the appropriate reliability software for our needs.

268

Software assessment Before taking a closer look at the features of reliability software packages, software types should be differentiated. There is a wide variety and although special programs designed for only a specific problem is quite rare, the main functions can be obviously separated. The only exception is the class of complex reliability software that combine a wide variety of functions within a modular structure. The main groups of specific reliability software packages are availability, reliability, life cycle cost modelling, computerized maintenance management, failure analysis, maintenance audit, failure mode effects and criticality analysis, safety and environmental mainte-nance, shutdown/turnaround management, spare parts analysis and optimisation, and tool control software. Definition of availability and maintenance – their role in energy generation Availability is the probability of a system’s successful operation in a determined period of time, can be calculated by the ratio between life time and total time between failures of the equipment [1].

= =life time life time

total time life time + repair timeA (1)

Maintenance means those activities undertaken after a system is in the field to keep it operational or restore it to operational condition after a failure occurred [2].

Availability software

The availability depends on many features, including design, the way of operation, fuel etc. Availability, reliability and life cycle cost modelling can be treated together as they are influenced by each other. The availability software has a long track history and first industrially successful applications are dated as early as early 1980s [3]. They have been based on the in-house software developed by ICI plc (UK).

This is a common approach in many software packages, e.g. AAP Reliability Tool-box, ACE, ALTA, Analyst, APT-MAINTENANCE, ASENT, CARMS, D-LCC, Fault-Tree+, FMEA-Pro. Table 1 made an attempt to summarise the main features of some of the widely used software tools.

Shutdown/turnaround management software Plant maintenance shutdowns require careful planning, scheduling, and control. Shutdown management software can help this process with multiple shutdown scenarios, shutdown job plans and job steps creation, automatic scheduling, shut-down resource management etc. Turnaround management software tools are plant-wide information systems for out-ages and shutdowns. As examples can serve: APT-SCHEDULE, ATC Professional,

269

Cobra, Contractor Cost Tracking System, DataPath, RationPlan, RB.eye, Turnaround Manager.

Table 1. Some availability and related software Software/ package

Company or vendor

Demo accessibility/ restrictions of demo Website Demo. Time limit. ASENT Reliass www.reliability-safety-software.com Demo. Project size and save limitations. Availability

Workbench Isograph Ltd.

www.isograph-software.com Demo. Project size and save limitations. AvSim+ Isograph Ltd. www.isograph-software.com Demo (30 days). CAME BQR Reliability

Engineering Ltd. www.bqr.com No demo. COMPARE Maintenance 2000 www.m2k.com No demo. ENDAT SPG Media PLC www.power-technology.com Demo or online demonstration. MEADEP SoHaR Corporation www.sohar.com No demo. PLASMA Maintenance 2000 www.m2k.com Demo. Time limit. RAMP Reliass www.reliability-safety-software.com

Reliability software Reliability is beside the design and manufacturing inhered features strongly influ-enced by operating conditions. These parameters can be handled by specific soft-ware tools. Reliability software suites contain prediction or analytical modules, or both. They have standardized critical functions and capability to model complex system scenarios (e.g., parallel, standby, bridge, network), quantify risk and ensure safety, control corrective actions, build FMEAs, determine optional system parameters for maximum performance, control all analyses with the centralized navigator, support database connectivity (Microsoft Jet Engine, MSDE, Oracle, SQL Server etc.). Their level of integration is high, containing widely used standards and methodologies to analyze systems or components and increase safety. Modern reliability analysis tools offer multiple system analysis, project file sharing, results comparison, component or system drag and drop between projects, as well as quick new project creation by reusing components from other systems. GUIs make it easier and faster to use the components or the whole system, hierarchy trees, charts or graphs, automatic chart generation through a wide variety of powerful wizards. Table 2 presents some features of several software packages.

270

Table 2. Specific reliability and related software tools Software/ package

Company or vendor

Demo accessibility/ restrictions of demo Website Demo. ALTA ReliaSoft Corp. www.reliasoft.com Cannot save files, maximum of 10 components. Analyst Powertechnic Pty

Ltd www.powertechnic.com.au Demo. Project size and save limitations. AvSim+ Isograph Ltd. www.isograph-software.com To download a demo, a license must be obtained from ReliaSoft. Validation limited, some features (e.g., printing) not available – see screen shot in Fig 1

BlockSim

ReliaSoft Corp.

www.reliasoft.com Demo (30 days). CAME-LCC BQR Reliability

Engineering Ltd. www.bqr.com Demo (30 days). CARE-RBD BQR Reliability

Engineering Ltd. www.bqr.com Demo. D-LCC Reliass www.reliability-safety-software.com Demo. Project size and save limitations. FaultTree+ Isograph Ltd. www.isograph-software.com Demo. A sales engineer will demonstrate it. FMEA-Pro Dyadem

International Ltd. www.dyadem.com Demo. Unlimited version, valid for 30 days. GoldSim

Reliability Module

GoldSim Technology Group

www.goldsim.com

Demo. The number of components is limited – see screen shot in Fig 3

ITEM ToolKit

ITEM software

www.itemuk.com Demo. MainTain ITEM software www.itemsoft.com Demo. Project size and save limitations. MKV Isograph Ltd. www.isograph-software.com Demo. Limited to 4 years of historical data entry. PERDEC OMDEC www.oliver-group.com Demo. Number or records held and data exporting.

PMO2000 Operations & Maint. Consulting Services www.reliabilityassurance.com

Demo. Register max. 30 provisions for 30 days. Protecs The Lance Group www.lance-safety.com Demo. RAMP Reliass www.reliability-safety-software.com Freeware. No restrictions. RAPTOR ARINC www.raptorplus.com Demo. Project size and save limitations. RCMCost Isograph Ltd. www.isograph-software.com

271

Demo. Limited capacity. RelCalc for Win.

T-Cubed Sys., Inc. www.t-cubed.com

Demo. No restrictions. RELCODE OMDEC www.oliver-group.com Fully functional, limit placed on components or assemblies.

Reliability Studio

Relex Software Corporation

www.relex.com Demo. No project save, project size limitations. RiskVu Isograph Ltd. www.isograph-software.com Demo. Project size and save limitations. Reliability

Workbench Isograph Ltd.

www.isograph-software.com Demo. 60 days trial version. SAPHIRE

[4,5] Idaho National Laboratory https://saphire.inel.gov

Demo. Impossible to save. SIMFIA APSYS www.apsys.eads.net Freeware (requires Weibull++). SimuMatic ReliaSoft Corp. www.weibull.com Demo. Contains a data subset representative of the actual data contained in the full version.

SPIDR System Reliability Center

http://src.alionscience.com Demo. Project size and save limitations. Weibull Pro Isograph Ltd. www.isograph-software.com Demo. Weibull ++ ReliaSoft Corp. www.reliasoft.com

Figure 1. ReliaSoft BlockSim

272

Some reliability software tools are capable to solve specific problems only. However, there are complex packages that contain modules providing various availability, reliability and maintenance related solutions in one integrated program, including:

• Availability Simulation, availability predictions; • Failure Mode Effects and Criticality Analysis (FMECA); • Reliability Block Diagram (RBD); • Fault Tree Analysis (FTA); • Event Tree Analysis (ETA); • Markov Analysis (MKV); • Life Cycle Cost Analysis; • Reliability prediction; • Maintainability prediction; • Visualization with different distributions (normal, Weibull, exponential ...).

Some of these integrated platforms offer their modules to be used individually.

Figure 2. Relex Reliability Studio The Relex Reliability Studio has been used as an example and a deeper test has been undertaken. It consists of several modules, including fault tree and event tree analysis, FMEA/ FMECA, FRACAS corrective action, human factors risk analysis, LCC (life cycle cost

273

analysis), maintainability prediction, optimisation and simulation, reliability block diagram, reliability prediction, and Weibull analysis. An example of the work screen of this tool has been presented in Fig 2. Another example of a complex software package which has been tested is the Isograph Reliability Workbench. It includes different analyzing techniques, including reliability prediction, maintainabil-ity prediction, FMECA, reliability block diagram analysis, reliability allocation, fault tree analysis, event tree analysis, and Markov analysis. These modules are inter-connected – they can work together. The links between the modules are maintained and data can be automatically updated in one module due to changes made in another module. Data can be easily transferred between the different parts of the software tool by automatic data transfer or copy/paste. They are Access and Excel compatible.

Figure 3. ITEM Toolkit

274

Safety and environmental maintenance software There is a wide variety of industrial plants and power generators, which have at least one feature in common: they are dependent on high quality maintenance. Managing optimum maintenance is a broad, complicated and time-consuming task. Maintenance software tool track everything to let maintenance managers know exactly what kind of maintenance is needed and when. With the help of the right software tool it is possible to prevent problems from occurring. Software tools offer additional features as well. They include the database of users and vendors, a built-in invoicing system, or a scheduled task system. Some examples: AMMS, AUDITWorks, CIRSMA, SafePro, Tracker.

Failure Analysis software Failure analysis requires collecting and analysing data to determine the cause of a failure and how to prevent it from happening again. These tools can be used in the development of new products and for the improving of existing ones. Examples of Failure Analysis software tools are: AAP606, ASENT FRACAS, AvSim+, CAME-FDA, FaultTree+, FMEA-Pro, Maintenance Analytics, Meridium, MKV, QTMS, RDA Utility, Ttree.

FMECA software Failure Mode Effects and Criticality Analysis are performed as design processes which should eliminate or even reduce failure modes with high probability and severity. This analysis is complex task, which includes several techniques, e.g. mechanical reliability predictions, determining the effects of system and equipment failures. Examples of MFECA software tools are: FMECA Plus, FMECA Processor for Windows, Maintenance Analytics, MEAnalyst, Manifer, Meridium.

Computerized Maintenance Management System (CMMS) software CMMS software packages (also known as Enterprise Asset Management) maintain a computer database of information about the company maintenance operations. This information may be capable to improve the effectiveness of the work of maintenance engineers. It can be also helpful to make informed decisions and deal with third parties. Some examples of CMMS software tools are: AMPRO, MPulse, Maintenance Connection, MAPCON, Bigfoot, eMaint, MicroMain Maintenance Management, OPRA, Maintenance Coordinator.

275

Maintenance audit software If maintenance plan is required, the process of performing the routine actions, which keep the device in the working order should be audited. Most maintenance audit software tools are aimed at work order entry, equipment history, and accounting. Some of them may help to even measure current productivity, set productivity goals and track the process toward those goals. Examples of such programs are Aware.MPS, Aware.MNT, Easy Audit, infoRouter, ON KEY Auditor, PEMMS, UmtAudit, xmEXEC.

Spare parts analysis and optimisation software Spare parts (also referred as service parts or spares) are the extra parts available in the proximity of the item of a system for which they might be used. The analysis and optimisation of spares is the main concept of the strategic service management. It can be used to ensure that right spares are at the appropriate place at the right time. As examples can be listed software tools APT-SPARES, AvSim+, CAME-S2A, CARE, FMEA-Pro, Integrated Spares and Logistics Evaluator, VMetric.

Tool control software Tool control software tools track vital assets and tools, providing a detailed audit trail for check-in/checkout, minimums and maximums for ordering, and calibration tracking with detailed history, issue and return functions, inventory and ordering, re-work, kit building etc. Examples are: Automated Tool Inventory Control and Tracking System, collectiveTool Crib, CribMaster Inventory Management System, TLC32 Professional, ToolHound, ToolManager.

Conclusions It has been a wide verity of software tools available on the market. To solve availability, reliability and maintenance tasks a proper choice should be made. The choice influences the quality of the results. The available marketing and sales documentation is rather comprehensive and widely supported by web based information. However, they mostly feature only their strong and selling points and very few weaknesses and/or missing features are mentioned. For those reasons more overview and testing combined with benchmark case studies is need to provide at least basic suggestions and recommendations. In this direction the future work is targeted.

276

Acknowledgements The financial supports from the EC projects EMINENT2 – TREN/05/FP6EN/ S07.56209/019886 and Marie Curie Chair (EXC) MEXC-CT-2003-042618 are gratefully acknowledged. References [1] de Castro HF, Cavalca KL. Maintenance resources optimization applied to a manufacturing system. Reliability Engineering and System Safety 2006; 91:415. [2] Ireson WG, Coombs CF, Moss RY. Handbook of reliability engineering and management. McGraw-Hill 1996; Chapter 15.3 [3] Klemeš J, Krus A: Computer-Aided Availability Analysis of Chemical Plants. Computers in Industry 1985;6:265-272. [4] Faghihi F, Ramezani E, Yousefpour F, Mirvakili SM. Level-1 probability safety assessment of the Iranian heavy water reactor using SAPHIRE software. Reliability Engineering and System Safety (in press) [5] Smith C, Knudsen J, Kvarfordt K, Wood T. Key attributes of the SAPHIRE risk and reliability analysis software for risk-informed probabilistic applications. Reliability Engineering and System Safety 2008; 93:1151-1164.


277

CO2 Capture on Zeolites for Sustainable Energy Production

Gheorghe Bumbac1, Aurelia Bolma2, Anca Dumitrescu1, Vasile Bologa1 1University POLITEHNICA of Bucharest, Centre for Technology Transfer in the

Process Industries, 1, Polizu Street, Building A, Room A056, Sector 1, RO-011061, Bucharest, Romania, Phone: +40-21-4023916, Fax: +40-21-3185900, email:

[email protected] 2ICEMENERG S.A, 8 Energeticienilor Blvd., Bucharest, Sector 3, Romania, email:

[email protected]

Abstract In the context of the global warming, which is unanimously considered to be determined by the increasing of greenhouse gases emission, CO2 capture and sequestration is one of the possible ways to lower emissions resulted in fossil fuel combustion for industrial energy production. Most of the emissions of CO2 to the atmosphere from the electricity generation and industrial sectors are currently in the form of flue gas from combustion, in which the CO2 concentration is typically 4-14 % by volume, although CO2 is produced at high concentrations by a few industrial processes. Different technologies for CO2 capture are currently under research and development: physical/chemical absorption, adsorption, membrane separation. Today, chemical absorption is a typically process, with an energy cost ranging from 280 up to 960 kcal/kg CO2, depending on the sorbent type and the process configuration. Adsorption processes were first not really considered as competitive for CO2 capture when compared to absorption. In this work we study the adsorption on zeolites as alternative for CO2 separation from a variety of process streams, especially from flue gases resulted from the energy industry. Different types of zeolites with low Si/Al ratio, i.e., high aluminium content (zeolites NaA and NaX), high Si/Al ratio, i.e. low aluminium content (zeolite ZSM-5) and mesoporous silica MCM-41, respectively have been synthesized to check their CO2 adsorption capacity. Some data regarding zeolites synthesis, their characterisation, the experimental setup description and CO2 adsorption results will be presented in this paper.

Introduction The burning of fossil fuels for energy generation for industrial and domestic use accounts for the emission of huge quantities of CO2 in the atmosphere, which lead to a high degree of pollution by supporting the greenhouse effect. An evaluation of the CO2 concentration in the environment has concluded that 70 % of the emissions result from the energy industry mainly because of fossil fuels combustion. In order to reduce CO2 emissions in the atmosphere the present tendencies are the following: the development and the implementation of feasible economic technologies to capture CO2. There are three main directions for CO2 emissions reduction:

278

• Increasing of processes efficiency; • Replacement of present energy sources with others with less carbon or new

technologies (unconventional energy sources); • CO2 capture and sequestration.

The sorbents materials designated to be used in CO2 capturing are the synthetic zeolites. Gas adsorption on zeolites gains remarkable attention in the nanotechnology era, since it has importance in many industrial processes (Ruthven [1]; Delmas et al. [2]; Mac Dougall et al [3]; D. W. Breck [4]). The post-combustion technology for capturing the carbon dioxide has as main goal a selective mass transfer of one or more components to a stationary layer of solid porous materials, the adsorption process taking place simultaneously by chemical reactions or by forming physical bonds between the captured component and the adsorbent solid surface. In this way the gases are purified, the harmful components – in this case, the carbon dioxide, being eliminated.

Experiments and results

Industrial flue gas characteristics In our case study flue gas used to prove CO2 capture on zeolites was collected from an industrial power plant, located nearly Black Sea Coast refinery and petrochemical plant named Midia Navodari (a Rompetrol Society Subsidiary). Midia Power Plant produces power for national energy system and steam for the refinery and petrochemical total site and uses for combustion feed also fuel oil and also natural gas or both of them. In order to characterise the flue gas composition generated by burning of mentioned fuel laboratory tests and measurements have been conducted in a steam boiler of the plant, with a 105-t/h capacity of steam generated, adequate adapted to supply the needed information and data for our studies. The flue gas resulted from the combustion process are evacuated in atmosphere through the funnels for dispersion. The tests and experiments have been conducted for different charges of the boiler: between 70 and 90 t/h by using fuel oil and natural gases in different proportions of volumetric flow rates. The analysis of component concentration in the flue gas which was evacuated in the vent has been carried out with a high sensitive sensor (TESTO 350 XL). Table 1 presents the flue gas component concentrations recorded at their evacuation in the atmosphere when the two types of fuel were used.

Syntheses of zeolite adsorbents (molecular sieves type) for carbon dioxide capture

The structure of zeolite we had to be synthesized was of molecular sieve type and selective for CO2 and it had to have the following characteristics: large pore structure, medium pore structure, small pore zeolite structure and ordered mesoporous material in order to establish the relation between adsorption process performances and zeolite structure.

1. Different types of zeolites with low Si/Al ratio, i.e., high aluminum content (NaA and NaX zeolites), high Si/Al ratio, i.e. low aluminum content (ZSM-5 zeolite) and mesoporous MCM-41 silica, respectively have been synthesized to check

279

their CO2 adsorption capacity (C. Zhang et al, 2003, [5]; J. Weitkamp, 2000, [6]; J. S. Beck et al., 1992, [7]).

2. Generally, the synthesis of zeolites involves a few elementary steps in which a

mixture of silica and alumina precursors, alkaline cations and water is converted into microporous crystalline aluminosilicates R. M. Barrer, 1987, [8]). Zeolites A and X were synthesized in an inorganic system: SiO2-Al2O3-Na2O-H2O. The high silica containing ZSM-5 zeolite was synthesized using 1, 6-diaminohexane (HDA) as structure directing agent (template). The synthesis of mesoporous MCM-41 silica involves the use of a surfactant aqueous solution. The cationic surfactant assisted synthesis of the MCM-41 (sample 07ZEC6) has been performed according to reported procedures using cetyl trimethylammonium bromide (C16TMABr) as surfactant (G. D. Stucky et al., 1994, [9]). Sodium silicate, tetramethylammonium hydroxide and fumed silica were used as reagents.

Table1. Flue gas component concentration intervals, identified from the experiments

Component Units minim maxim minim maximexcess of

combustion air % vol. 35.6 47.3 12.9 14.3tga

o C

O2 % vol. 5.51 6.74 2.4 2.63CO2. % vol. 8.08 8.78 14.96 15.15CO mg/Nm3 0 2 2 2NOx mg/Nm3 213 214 487 505

SO2 mg/Nm3 0 0 1832 1926

157-160 186-189

Natural gas Fuel oil

For all zeolite syntheses the crystalline solid materials were recovered by successive operations of filtration, washing with distilled water to pH of 7.5-8.0 followed by drying in air at 100oC. The preparation process of the MFI-type material included several successive hydrothermal post synthesis stages to obtain the hydrogen form of ZSM-5 zeolite as follows:

• Calcination in air at 600oC for 6 hours to remove the organic template • Ion exchange with 1M NH4NO3 solution at 90oC, followed by washing with

distilled water and drying at 100oC • Calcination at 550oC to convert NH4-ZSM-5 into H-ZSM-5

Table 2 gathers the synthesis conditions of sodium form of zeolites A and X and ZSM-5 zeolite, respectively and mesoporous MCM-41 silica.

280

Structural characterization of adsorbents The synthesis of the designed molecular sieves was checked by X-ray diffraction (XRD). The samples were recorded on a computer-controlled DRON DART UM2 diffraction meter equipped with a graphite monochromatized Cu anode. The XRD patterns confirm the formation of the desired zeolite structure: A zeolite, X zeolite and ZSM-5 zeolite and of ordered mesoporous MCM-41 respectively. Figures 1 (a,b) and 2 (a) display the XRD patterns of well-crystallized zeolites exhibiting the crystalline structure of zeolite A (framework type LTA), zeolite X (framework type FAU) and ZSM-5 (framework type MFI), respectively (M.M.J. Treacy et al., 2001, [10]).

Table 2. Syntheses of zeolites A, X and ZSM-5 and mesoporous silica MCM-41

Synthesis type Gel - molar composition

Hydrothermal synthesis

NaA

07ZEC1 Zeolite

2.0SiO2 · 1.0Al2O3· 2.2Na2O· 70H2O

Dynamic system, 1 L Parr reactor, pH=13, 80oC, 4 h

NaX

07ZEC5 Zeolite

1.0SiO2· 1.0Al2O3· 3.5Na2O· 120H2O

Static system, 10 mL autoclave, pH=13, 100oC, 4 h

Na-ZSM-5

07ZEC3 Zeolite

1100.0SiO2· 1.0Al2O3· 1.2Na2O· 0.22HDA· 60H2O

Dynamic conditions,

1 L Parr reactor

pH 12.2,170oC, 48 h

MCM-41

07ZEC6

SiO2· 0.07Na2O· 0.08TMAOH · 0.15C16TMABr · 60H2O

Dynamic conditions

1 L Parr reactor

pH 12,100oC, 48 h

It is necessary to stress that the diffraction pattern of faujasite zeolites showed that the zeolite synthesized in our laboratory (sample 07ZEC5) is structurally similar to commercial 13X-UOP zeolite.

0 10 20 30

I (a.

u.)

2 θ C uK α

(a) As-synthesised A-zeolite specimen

(07ZEC1)

0 10 2 0 30 40 5 0 60

I (a.

u.)

2 θ C u K α

(b) XRD pattern of as-synthesised

X-zeolite specimen (07ZEC5)

Figure 1. XRD patterns for A- and X- zeolites

281

Note that the LTA and FAU frameworks types described structures build from three-dimensional channels intersecting large cavities. The channels are of 0.41x0.41 nm for LTA framework type and of 0.74x0.74 nm for FAU framework type, respectively. The MFI framework has a particular structure characterized by a void space of three-dimensional interconnecting channels of 10-rings of 0.51x0.55 nm and 0.54x0.56 nm respectively (Ch. Baerlocher, et al., 2001, [11]). The MCM-41 structure is not a genuine crystalline structure because the atomic arrangement of walls is not crystalline.

0 5 1 0 1 5 2 0 2 5 3 0 35 4 0 4 5

2 θ C u K α

I (a.

u.)

(a) As-synthesized ZSM-5 zeolite

specimen (07ZEC3)

0 5 1 0

2 θ C u K α

I (a.

u.)

(b) As-synthesized ordered mesoporous

MCM-41 specimen (07ZEC6)

Figure 2. XRD patterns for ZSM and MCM zeolites The typical XRD pattern exhibits in the wide-angle region only a broad peak around 2θ=23o, indicative for amorphous silica including the intrinsic contribution of wall disorder structure (F. Schűth, 1996, [12]) (not shown here). In the low-angle region 4 to 5 Bragg reflections appears revealing the regular disposition of pore systems and can be indexed according to a 2D hexagonal p6mm symmetry (fig. 2b). The MCM-41 structure is an assemblage of 3-5 nm size parallel hexagonal channels developing a so-called “honeycomb” motif.

Textural characterization of adsorbents Textural characteristics of the samples including specific surface area, pore volume, pore size distribution and average pore diameter were evaluated from nitrogen adsorption measurements at 77K using an Automatic Volumetric Sorption Analyzer (Autosorb MP-1, Quantachrome). The following samples were analyzed to evaluate the pore structure parameters: commercial 13X zeolite, NaX (07ZEC5) zeolite, H-ZSM-5 (07ZEC3) zeolite and calcined MCM-41 (07ZEC6), respectively. All samples were degassed at 300oC for 4 hours at 10-5 torr prior to nitrogen adsorption measurements. The adsorption isotherms corresponding to microporous materials are of type I having relatively small external surface in accordance to IUPAC classification. Some differences between the zeolite samples can be related to different pore structure accessibility. For all samples adsorption isotherms showed a fast increase in the nitrogen amount adsorbed in the low pressure region (p/p < 0.1) followed by a long flat region at higher pressures. The difference appeared in the higher-pressure region (p/p0>0.9) leads to the following order of large pore size: NaX (07ZEC5) zeolite > commercial 13X zeolite > H-ZSM-5 zeolite. This result can be related to different factors as type

282

framework structure, cationic composition, synthesis conditions, etc. Nitrogen adsorption-desorption isotherm of MCM-41 sample is of type IV according to IUPAC classification. The shape of the isotherm and the hystheresis loop associated with capillary condensation in mesopores are characteristic for cylindrical mesopores, which indeed form the MCM-41 structure.

Table 3. The textural characteristics of adsorbents

Surface area, m2/g

Pore volume, cm3/g

Sample

BETa micropore total micropore

Average pore

diameter, nm

13X UOP Zeolite 906 719 0.667 0.210 1.41 NaX-07ZEC5 Zeolite 692 501 1.035 0.164 1.42

H-ZSM-5-07ZEC3 Zeolite

410.5 255 0.323 0.102 1.41

MCM-41 - 07ZEC6 1113 - 1.437 0.37 2.52

Table 3 shows the textural data of different adsorbents. The surface area was calculated using the BET equation. The total pore volume was obtained by converting the amount adsorbed at a relative pressure of 0.999 to the volume of adsorbate liquid. The micro pore area and micro pore volume was evaluated by the t-plot method. The pore size distribution was determined by the BJH method using the desorption branch of the isotherm. In general the values obtained for X zeolite and ZSM-5 zeolite samples are in accordance to reported data. Data presented in table 3 shows that FAU-type structure offers some advantages in terms of surface area and pore volume, which can provide more space for adsorbate molecules to accumulate and adsorb inside the cage structure of diameter>0.11 nm. In general, high surface areas and large pore volumes are important parameters for the selection of a zeolite adsorbent. The MCM-41 sample has high specific surface area, high pore volume and large pore diameter that are typically for a highly ordered pore structure. In connection with reported data it can presume that good MCM-41 based adsorbents for carbon dioxide capture could be obtain by both exploiting the MCM-41 mesoporosity and applying a post synthesis chemically modified treatment (A. Macario et al., 2005, [13]; R. S. Franchi et al., 2006, [14]).

Adsorption process experiments and modelling The equilibrium dates from adsorption CO2 on zeolites were obtained with thermo gravimetric instrument Du Pont 951. The laboratory set-up consists of thermo gravimetric balance (TG) to determine the amount of CO2 at a wide range of temperatures and pressures, and a modified differential scanning calorimeter. These determinations may represent a quantitative criterion for the pre-selection of synthesized zeolites, which will be used in the CO2 capture technology. The set-up is carried out as a closed system in which the sample is placed in an atmosphere of pure vapor of carbon dioxide. No other gas is present in the test cell during the measurements. The system is designed to work at a temperature range for the sample from 0 to 600 oC. For each temperature step, one equilibrium point consisting of temperature, pressure and adsorbed amount mass of the sample is

283

needed. The sample is completely desorbed as soon as its weight does not show any further change. During an adsorption process step and also desorption process step all equilibrium points are reached and the hysteresis diagrams are records.

Table 4. CO2 adsorption capacity of several sorbents

Maximum adsorption capacity in g CO2/ 100 g adsorbent, at 27 oC (nearly liquefaction pressure of pure CO2 at this temperature)

Zeolite

99.99% CO2 purity CO2 with traces of H2O Commercial zeolites

13X- UOP 19.52 -*

07- UOP 19.37 - 13XE 14.39 9.71

5A bile 13.31 - A 15 5.4 -

Synthesized zeolites 07ZEC 5 20.02 23.6 07ZEC 3 6.16 -

07ZEC 3E 4.6 - 07ZEC 1 4.2 - 07ZEC 1.6 - A7 TUF 5.35 - A5 TUF 12.7 - CLN-Cu 5.43 -

CLN-Cu-La 4.1 - * “- “ no experimented

Figure 4. Equilibrium data of system CO2 gas-solid phase of ZECASIN 07 ZEC 5

zeolite, at different temperatures

284

Table 4 presents the adsorption capacity of the commercial and our synthesized zeolites, and figure 4 gives adsorption isotherm resulted from experiment data interpretation.

Post-combustion CO2 Capture process topology Several operation techniques of the adsorption process are known: thermal swing adsorption (TSA) and pressure swing adsorption (PSA). In both techniques, components adsorption and desorption are carried out through an adequate temperature change or pressure change in the adsorption system. Based on experimental data (especially adsorption equilibrium and adsorption capacity) we used process simulation to synthesize a process for post-combustion CO2 capture. In figure 5 is given process flow diagram of the synthesized unit. This process unit corresponds to an experimental micro pilot.

Figure 5. Process flow sheet for post-combustion CO2 capture

Flue gas stream characteristics are presented in figure 6. This stream leaves steam boilers of the power plant unit (the energy converter) and enters in a solid separator in order to remove its content of solid particles. After a succession of operations of cooling and condensed water advanced removing flue gas is dried in a molecular sieves process unit. Then, dried flue gas enters in the adsorption unit, equipped with ZECASIN 07 ZEC 5 zeolite, after a pre-compression at 6 atm pressure. Superheated steam is used to regenerate zeolite sorbent in the adsorption unit. Poor in CO2 flue gas, in adequate quantities, at an adequate temperature, is used to regenerate molecular sieve drier unit. Figure 7 gives characteristics of poor gas discharged in the vent.

285

Figure 6 Flue gas characteristics at outlet

of boilers unit Figure 7 Flue gas discharged in vent

Adsorption unit considered in flow sheet simulation was a PSA device with 2 cylinders. This PSA system was modelled and simulated by using MATLAB, according process data given in flow sheet scheme in figure 5. 200 kg was the quantity of zeolite required to fit the mass balance of the process flow sheet.

Conclusions CO2 on zeolites post-combustion capture process was studied to estimate adsorption capacities and to synthesize and economic evaluate the process. A NaX type zeolite (ZECASIN 07 ZEC 5) was prepared and characterized. There have been also synthesized and morphological and structural characterized a series of different type of zeolites LTA, FAU and MFI. From all the synthesized zeolites, the 07 ZEC C5 adsorbent has shown a structural and morphological similitude with the commercial zeolites 13 X-UOP and 07 UOP. The 07 ZEC C5 type synthetic zeolite has indicated the higher CO2 adsorption capacity after the conducted tests by using the thermo desorbtion technique.

Table 5. Cost evaluation for various equipment items No. Equipment for the CO2 capture installation Cost Value 1 Cyclon X -100 $ 4 281 2 Heat exchanger E 100 $ 1 200 3 Cooler $ 1 200 4 Phase separator V 100 $ 2 676 5 Drying unit $ 3 224 6 Drying unit packing $ 9 166 7 Blower / compressor $ 213 816 8 Heat exchanger E 101 $ 1 200 9 Adsorption / desorption PSA unit $ 3 376

10 Packing for the adsorption unit (Zeolite) $ 24 444 11 Total $ 264 577 12 Total Euro 175 000

286

Based on this zeolite adsorption experiments were developed in order to modeling the process from thermodynamic and kinetic point of view. A post-combustion capture of CO2 unit was then modeled and simulated in order to evaluate the technical-economic characteristics of a pilot plant. Table 5 illustrates the main apparatus and equipment necessary to the pilot, designed to process 90 Nm3/h burned gases. We present also the cost evaluations made with the software instrument CAPCOST (Turton, R. (1997).[15]). It was estimated that the 200 kg of zeolite are capable to capture 5 kg/h CO2 from the flow rate of approximate 108 kg/h burned gasses containing initially a concentration of 14.3%(wt) CO2. The operating costs for this capture were estimated to correspond to a theoretical consumption of 7 kW electric energy and 2 kW thermal energy (superheated steam). The pilot installation mentioned is in process of realization and testing.

Acknowledgements The financial support from the Romanian ”MCT CEEX”- Program Project 107/2005 is gratefully acknowledged.

References 1. Ruthven D. M, Microporous Materials, 22, 1998, 537 2. Delmas M. P. F & Ruthven D. M, Microporous Materials, 3, 1995, 581 3. Mac Dougall and D. M. Ruthven, Adsorption, 4, 1999, 369 4. D. W. Breck, “Zeolite Molecular Sieves: Structure, Chemistry and Use, 1984, R.

E. Krieger Publishing Company, Malabar, Florida, 172 5. C. Zhang, Q. Liu, Z. Xu and K. Van, Microporous and Mesoporous Materials, 61.

2003, 157 6. J. Weitkamp, in Zeolites and Catalisis, Solid State Ionic, 131, 2000, 175 7. J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz et al, J. Am. Chem. Soc,

114, 1992, 10814 8. R. M. Barrer, Hydrothermal Chemistry of Zeolites, London Academic Press, 1987 9. G. D. Stucky, A. Monnier, F. Schuth, Q. Huo et al, Mol. Cryst. Liq. Cryst, 240,

1994, 187 10. M.M.J. Treacy and J. B. Higgins “Collection of Simulated XRD Powder Patterns

for Zeolites, Fourth Revised Edition, Elsevier, 2001 11. Ch. Baerlocher, W. M. Meier, D. H. Olson, “Atlas of Zeolite Framework Types”,

Fifth Revised Edition, Elsevier, 2001 12. F. Schűth, Ber.Bunsenges. Phys. Chem., 99, 1995,1306. 13. A. Macario, A. Katovic, G. Giordano et al, Microporous and Mesoporous

Materials, 81, 2005, 139 14. R. S. Franchi, P. J. E. Harlick and A. Sayari, Ind. Eng. Chem. Res, 44, 2006,

8007; 15. Turton, R., Bailie, R. C., Whiting, W. B., & Shaeiwitz, J. A. (1997). Analysis,

synthesis and design of chemical processes. New Jersey: Prentice Hall.


287

New Energy-Saving Technologies in the Chemical Industry Vladimir Panarin, Valery Meshalkin

D. Mendeleyev University of Chemical Technology of Russia Russia, Moscow, 125047, Miusskaya Square, 9,

Tel: + 7 499 9783164, e-mail: [email protected]

Abstract Some of energy-saving technologies used in the chemical industry in Russia are described which enable considerable energy- and material savings, prove to be environmentally friendly and increase the output of high-quality production with reduced requirements of operational and capital costs [1].

Vortical photochemical thermocatalytic reactor The design of new vortical photochemical thermocatalytic reactor for cleaning of emission gases is given below (Fig 1 and Fig 2). It provides flameless burning of gaseous process and ventilating emissions and can be used for sanitary cleaning the gases in chemical, paint and varnish and other industries [2].

Figure 1. Vortical photochemical thermocatalytic reactor: 1-cylindrical case; 2 -

distributive chamber; 3, 5 - joints; 4 - cleaned gas chamber; 6, 7 - lattices; 8 - tubes; 9 - thermocatalytic lattice; 10 - screw twisting unit

The developed design provides identical conditions in the process of gas neutralization in all the volume of a reactor simultaneously maintaining effective

288

cooling of the catalyst that enables to increase efficiency and reliability of the reactor described. The reactor operates as follows: emission gases containing hydrocarbons enters through the reception unit in the distributive chamber from which it passes through screw channels of twisting units as the twirled jets and interacts with thermocatalytic elements and passes them on ring channels between catalyst lattices and sources of infra-red radiation. During high-speed flow of the twirled stream of gas on the ring channel of thermocatalytic elements it is exposed to an infra-red irradiation, radiant energy is absorbed selectively only by molecules of hydrocarbon connections which thus activate.

Figure 2. Thermocatalytic element of a vortical reactor (Longitudinal section)

1 - damping chamber; 2 (3) - entrance branch tube; 4 - nozzle; 5 - diametrical crossing plates; 6 - console sites; 7 - shaped (bent) plates

Figure 3. The device with disperser and damping the chamber to increase of selection and cleanliness of divided distillate products

The field of centrifugal forces in the twirled stream creates conditions for radial separation of heavy hydrocarbon connections to a wall of the channel, and (which is experimentally established) the optimum size of length of the ring channel provides high uniformity of distribution of hydrocarbons on all the length of the channel. Besides the uniform warming of the catalyst on all the surface is used. It allows

289

defining the optimum length of the irradiation zone and the general length of thermocatalytic element.

1 (10) – entrance branch tube; 2-nozzle; 3 - damping the chamber; 4 (5) - the first (second) group of shaped plates; 6 - crossing diametrical plates; 8 - console sites; 9 - shaped plates; 11, 12 - activators.

Figure 4. The device with disperser as crossing plates to increase the selection and cleanliness divided distillate products

Two dispersive devices (Fig 3 and Fig 4) have been developed to increase selection and cleanness of distillate products for processing oil-containing raw materials and decrease energy consumption in processes of raw material separation:

• the device with dispergator and snubber chamber. • the device with dispergator as crisscross plates.

To reduce the fuel and power consumption it is necessary to use the following design solutions operations which are taking into account mass and heat exchange and of processes: - Design and optimisation of reactors for thermal processes in combination with small-sized heat exchange devices providing full use of the heat of chemical reactions, with optimum temperature approach of the process, in particular, development of units with the organized boiling layer with a nozzle from heat exchange components; - Design and optimisation of small-sized reactors for processes with maximum use of utility consumption which can replace of expensive type of fuel, i.e. oil and gas, coal or slates; - Development of highly effective mixing units at the minimum costs of energy for mixing streams a liquid - liquid and liquid – solid phases; - Development of high intensity absorptive and stripping units for gas energy utilisation (Ventury dispersing liquids with the weighed layer of a liquid such as “foamy devices”; packed columns with a mode of inversion of phases, gas lift devices);

290

- Development new thermal- and corrosion resistant materials with the low temperature sensitivity differences with the purpose to increase the working temperature of processes and their efficiency; - Reduction of direct heat and energy losses during the transportation and in pipeline systems by improvement of quality heat insulation surfaces of the equipment and pipelines, creation of the equipment with re-circulating utilities, and also by choosing optimum diameters and characteristics of durability of the pipes; it is necessary to rationally choose types of utilities for various chemical manufactures; - To carry out replacement of ventilating installations by manufactures, by installing encapsulated the equipment and the equipment; - Creation of economically efficient installations providing use of low potential heat process streams in chemical plants.

Conclusions The developed devices will allow to use more effectively energy at processing oil-containing raw material and to lower power consumption of processes of raw material division.

References [1] Meshalkin V, Tovazhnjanskij L. Fundamentals of the theory energy-saving the integrated chemical-technological systems. Kharkov, 2006. [2] Meshalkin V. Increase of efficiency of perfection of processes and devices of chemical manufactures. Kharkov, 1985.


291

Innovative Master Course on Computer-Aided Synthesis of Energy-Saving Process Systems at Mendeleev University

Valery Meshalkin, Gleb Zakhodiakin

Mendeleev University of Chemical Technology of Russia, International Institute of Resource-saving Logistics and Technology Innovation

125047, Miusskaya sq., 9, e-mail [email protected]

Abstract In this paper a course on computer-aided synthesis of energy saving systems is presented. The course is aimed to improve the skills of Mendeleev University’s specialist and master degree students in mathematical modeling of process systems and computer-aided process synthesis methods. The course is intended for students of High Technology Department, Process Engineering Department, and International Institute of Resource-Saving Logistics and Technology Innovation. The course syllabus, prerequisites and goals are provided.

Introduction Chemical plants process huge amounts of raw materials and products and they are also intensive consumers of energy and water used by processes. Operating chemical plants generate emissions to the atmosphere, to the water and to the soil, and this is what makes environmental management a major factor these days. Since 1970 a lot of research of computer-aided process system synthesis and optimization techniques has been carried out. The synthesis of process systems is a complex creative task involving experts both in chemical technology, equipment design as well as plant economics. A traditional approach to chemical process design is evolutionary: from selecting reaction path and reactive subsystem design towards design of separation and energy subsystems. The results rely heavily on previous experience with process design and knowledge of process chemistry and physics, innovation, heuristic search and economical assessment of technology. The degree of freedom while designing the process system’s structure is very high. For instance, while choosing the best structure of a heat exchanger network for a refinery plant with 25 hot and cold streams one must consider about 10117 possible combinations, which is obviously infeasible. Automated process synthesis methods are aimed at solving the problem of complexity. They can help engineers to select best designs, at the same time providing clear economical evaluation of systems developed. Every engineer involved into chemical process design should be aware of these methods.

Course syllabus More than 30 years ago, a course on computer-aided process synthesis has been established at Mendeleev University by Meshalkin [1]. Today, chemical engineering students from the Higher College of Resource- and Energy-saving and High Technologies Department of Mendeleev University attend this course. The syllabus

292

includes problems of economic analysis of process technology design and retrofit, selection of optimization criteria, process system modeling, synthesis algorithms for several subsystems such as separation, energy and utility subsystems. The syllabus is summarized in Table 1. Table 1. Computer-aided synthesis of energy-saving process systems: the syllabus

3. Number of acad. hours 1. 2. Section/topic

Lectures Practice 1. Introduction, course structure, goals and

resources 2

2. Process system synthesis hierarchy. Phases of process system design. Techno-economical analysis of process system design

4

3. Modeling and analysis of process systems. Model verification using topological methods 10 8

4. Process system synthesis problems and methods. Design of resource-saving process systems.

20 12

4. 5. Total academic hours: 6. 36 7. 20

The course starts with a brief introductory lecture that poses main problems and tasks in the automated process synthesis area. The next major section of this course contains topics on technical and economical analysis for process design, including the description of process system design hierarchy, various costs associated with process plant construction, retrofit or operation, as well as selection of criteria for process optimization. Automated synthesis of process systems is based on mathematical models of such systems. As a consequence, the course covers structural modeling of process systems using graph models and principles of material and energy balances using process operators and flow graphs. Several methods for verifying and solving balance systems equations using mathematical software packages are presented. The main part of the course covers application of process-synthesis methods and includes the following topics [1]:

• Purpose of automated synthesis and an overview of process synthesis methods. • Structure of optimization problems. Choosing optimization criteria. • Synthesis of process systems and combinatorial complexity. Choosing between

superstructure, constrained search (branch and bound) and heuristic methods. • Synthesis of multistage separation systems using heuristic and constrained

search methods. • Synthesis of heat-exchange networks using heuristic and constrained search

methods. • Synthesis of heat-exchange networks using pinch-analysis [2-4].

Resource-saving and environmental management methods for cleaner production The energy system is an important activity from the economic and environmental perspective. This course therefore includes a detailed analysis of the energy systems.

293

Techniques to reduce the energy consumption of a process or activity are addressed. For example, selection of a right fuel or utility system integration method will reduce not only SO2 emissions, but also particulates (including metals) and NOx. All measures to reduce energy consumption will result in a reduction of all emissions to air including CO2.

Major attention within this course is drawn to theory and principles of resource-saving for chemical process design and retrofit. The applicable resource-saving techniques have been structured in four blocks. The first one includes the energy management systems, including general techniques to reduce energy consumption. The second one includes techniques to consider in the selection and cleaning of fuels that can be used for producing energy. The third block contains the technologies for energy production that can provide a good environmental performance using the different type of fuels as well as the utilities needed to run those techniques. The last block includes the abatement techniques to control air emissions that are applicable to the energy systems.

All process units, systems and activities are typically operated in an integrated way, aimed to optimize the production in an economic, sustainable and acceptable to society way. This requires a well managed approach in the execution and planning of all activities. This aspect has great implications in the impact of the chemical plant into the environment. Thanks to the TEMPUS-TACIS project CD-JEP 26045-2005, new topics on integrated pollution prevention and control have been added. In 2007-2008, 4 new textbooks have been developed [5-8], including teaching materials based on Reference document «Best Available Techniques for Mineral Oil and Gas Refineries». These materials help to understand focus processes as a whole, as well as give insight on cost efficiency, pollution and emission levels feasible.

We plan to produce some new materials for training of chemical engineering students in electronic form and deploy them in International Institute of Resource-saving Logistics and Technology Innovation e-learning system.

Due to a long-term collaboration between Mendeleev University, The University Manchester, Genoa University, Catalan University and National Technical University ‘‘Kharkiv Polytechnic Institute’’ in the framework of trans-European academic and research projects (TEMPUS-TACIS and INCO-COPERNICUS2 programs) there was an opportunity to significantly upgrade the contents of the Computer-Aided Process Synthesis course with modern and efficient synthesis methods (such as pinch analysis) as well as to introduce specialized software for process synthesis.

Acknowledgements This paper has been partially supported by TEMPUS-TACIS CD-JEP 26045-2005 project «Ecological and Resource-Saving Engineering» — ECORSE.

References [1] Kafarov V, Meshalkin V. Analysis and Synthesis of Chemical Process Systems, Moscow. 1992. [2] Puigjaner L, Smith R, Dovi V, Meshalkin V. Fundamentals of Process Integration and Environmental Economics. Moscow 1999. [3] Puigjaner L, Smith R, Dovi V, Meshalkin V. Advanced concepts of Process Integration and Environmental Economics. Moscow 2000. [4] Meshalkin V, Tovazhnansky L, Kapustenko P. Theory of integrated resource-saving process systems. Moscow 2006. [5] Meshalkin V, Gnauck A. Environmental informatics fundamentals. Moscow, 2008.

294

[6] Meshalkin V. Energy-saving and environment-friendly oil refining processes. Moscow 2008. [7] Meshalkin V, Butusov O, Gnauck A. Ecological systems mathematical modeling fundamentals. Moscow 2008. [8] Meshalkin V, Dovi V, Grinevich V, Koifman O, Bubnov A, Rybkin V, Gustchin A. High energy chemistry methods for environmental protection. Moscow 2008.


295

New Efficient Technology for Stabilisation of Humid Biomass in View of Large-Scale Supply

Giuliano Grassi, Stephane Senechal

EUBIA European Biomass Industry Association Rue d’Arlon 63-65, 1040 Brussels, Belgium

Phone: +32-2-4001020, Fax: +32-2-24001021, E-mail: [email protected]

Abstract There is a general consensus that the deployment of bio-energy on a global scale will bring many significant energy-security environmental, socio-economic benefits. However as far as the vast potential offered by cellulosic biomass conversion and utilisation is concerned, variation in the entire production system for biofuels supply (feedstock production, pre-treatment, conversion, utilisation) will have to take into account the differences in available potential resources, priority needs and economies, country by country. The new “Agro-pellets technology” is a basic pre-treatment technology, for stabilisation of humid biomasses. Taking into account its innovative performances, its high energy efficiency and relative low cost, this new technology is opening wide (short-term) perspectives for large-scale exploitation of all types of cellulosic biomasses (agro-forestry residues, organic wastes, energy crops) and for all sectorial energy markets (heat, power, transport) and some basic chemicals markets (Bio-H2, Bio-methanol etc.).

Co-firing and new pelletisation technology Beyond the decentralised bioenergy deployment for heat, power production, in a more distant future large-scale supply of biomass resources to the utilisation/ conversion plants will be required. For example :

• District heating; • Cofiring (coal-biomass) plants; • BTL conversion plants.

These types of plants may require 0.5 up to 2 mio t.d.biomass/year. This new Technology (Agro-pellets) opens competitive realistic opportunities for large- scale delivery:

• Reducing the logistics and costs of supply (storage- transport- handling); • By the homogenisation of the feedstock, for supply all the year around; • Blocking the biodegradation, of humid biomasses (when the moisture level is

above 10%). For large-scale Cofiring Operations, the positive characteristics of this new technology has been confirmed in Europe (UK) on the largest European Coal Power Plant (4,000 MWe)- that is collaborating with EUBIA on this issue since 2005.

296

The attractive performances of this new Commercial Technology can be summarised as here below:

• Capacity range:1 →10 t/h; • Reasonable specific investment: ≅ 30 €/t pellets; • Reasonable processing cost: 35 - 45 €/t pellets (depending of the type of

feedstock); • Mechanic drying of the feedstock with total low-energy consumption of: ≅

70→220 KWhe depending of the moisture content and quality of biomass. Thus the H.V. of humid biomasses (≅ 2,000 Kcal/Kg) can be increased to ≅ 4,000 Kcal/Kg with an energy cost of about ≅ 500 Kcal/Kg;

• The production cost of “agro-pellets” (from biomass residues at 50 €/ d.t) results thus at ≅ 90€/t pellets therefore competitive with the Natural Gas Spot - market price of 8 $/MMBTU (≅ 320 €/TOE) and also with the large-scale gas import natural price of 280 $/1,000 m3 (≅ 280 €/TOE);

• All types of biomass mixtures can be pelletised (Agro-pellets) without use of chemical binding compounds;

Conclusions

• More than 80 different types of biomass feedstock had tested so far. • Multi stages units for the production of Aro-pellets from very high moisture

content (~ 60%) should be available commercially in 2008; • Due to the compactness of this innovative technology (absence of thermal

pre-drying) also Mobile Units are being considered for construction.


297

Commercial Production of Bio-H2 from Agro-Forestry Residues

Giuliano Grassi, Stephane Senechal EUBIA European Biomass Industry Association

Rue d’Arlon 63-65, 1040 Brussels, Belgium Phone: +32-2-4001020, Fax: +32-2-24001021, E-mail: [email protected]

Abstract Intensive worldwide R&D activity is under way now for the efficient production of Biohydrogen in view in particular of its utilisation in vehicles. The commercial production of Bio-H2 by a 4 steps process could be initiated. Before the arrival of the H2 vehicles-market (~ year 2020), there are already huge potential markets available :

• Bio-H2 enrichment (10%) of the natural gas in pipelines; • Petroleum refining; at present ~ 40 mio t/y; • High quality steel production (~ 1.2 billion t steel/y world production).

With the Carbon-credits this production cost of Bio-H2 will be competitive with the present most used H2 production cost. EUBIA has initiated discussion with important potential industrial partners, for implementing in cooperation with Russia Federation, a first commercial demonstration Bio-H2 production unit (3-4 t H2/day), which could represents also a possible future Bio-H2 vehicles Refuelling Station.

Efficient biohydrogen production by using agropellet from new technology

Intensive worldwide R&D activity is under way now for the efficient production of Biohydrogen in view in particular of its utilisation in vehicles.

Figure 1. H2 productivity.

However, with the arrival on the market of a new very efficient, reasonable cost technology for stabilisation of humid biomasses by mechanical drying and pelletisation, the commercial production of Bio-H2 by a 4 steps process (all the involved technologies are commercially available) could be initiated. The anticipated

298

production cost is ~ 2.000 €/t (biomass at ~ 50€/d.t.) with a minimum yield of ~ 55 Kg H2/t pellets (via carbonisation) up to ~ 70KgH2/t pellets (via torrefication) this last value is now under experimental verification (i.e. Figure 1). Assuming that the Carbon-credits will be available (~ 7tCO2/t Bio-H2) this production cost of Bio-H2 will be competitive with the present most used H2 production cost (steam-Reforming of Natural Gas) (~ 1800 €/t for natural gas at ~ 8$/MMBTU). Before the arrival of the H2 vehicles-market (~ year 2020), there are already huge potential markets available: Bio-H2 enrichment (10%) of the natural gas in pipelines; Petroleum refining; at present ~ 40 mio t/y (more and more H2 is needed for the degradation of oil-quality ad for environmental fuels constraints); High quality steel production (world production is ~ 1.2 billion t steel/y). EUBIA in collaboration with Russia Federation Organisations (Boreskov Institute of Catalysis, Centre for the Development of Innovative Technology) with the EC-INTAS support is implementing a strategy study for optimising the 4 steps process of Bio-H2 production from Forestry Residues in view of its injection in pipelines for Natural Gas enrichment (Figure 2).

Conclusions EUBIA has initiated discussion with important potential industrial partners, for implementing in cooperation with Russia Federation, a first commercial demonstration Bio-H2 production unit (3-4 t H2/day), which could represents also a possible future Bio-H2 vehicles Refuelling Station.


299

Innovative Low-Energy Distillation Technology for Bio-Ethanol

Giuliano Grassi, Stéphane Sénéchal EUBIA European Biomass Industry Association

Rue d’Arlon 63-65, 1040 Brussels, Belgium Phone: +32-2-4001020, Fax: +32-2-24001021, E-mail: [email protected]

Michail Gougel

Globlive International AB ; Johan Skyttes väg 200, S-125 34 Älvsjö (SE).

Abstract Actual Industrial bioethanol processing requires significant investment and energy costs (~ 100-120 Kg OE/t bioethanol). A new technology (still under development) that shows very promising perspectives in terms of:

• Expected Energy Cost reduction for distillation : ~ 1/5 • Investment cost reduction : ~ 1/3

will be presented with main experimental results obtained.

Innovative bioethanol processing This new technology is based on the utilisation of “absorption/heat regeneration synthetic crystal hydrated” powder, of low cost (~ 5€/Kg), absorbing preferentially the ETOH molecules from a hydrous bioethanol solution (beer). The absorbing powder is submitted to frequent thermal regeneration cycling for re-evaporation of ETOH molecules from the crystals followed by condensation.

Scientific concept This synthetic crystal absorb water and ethanol (preferentially ethanol) up to 150 times its own weight. Its specific weight is 0,5 Kg/l. A solution of ethanol and water is released from crystals to a higher percentage of ethanol than from traditional distillation by evaporation (using low temperature fluid ~ 80°C) and recovered by condensation. By one step the ETOH concentration can be increased from 10% to 50%. Today the process of distillations need between 37 and 43 steps. GIAB with this technology need between 5 and 7 steps.

Bioethanol distillation operation The GIAB-technology can be used in two days. The first one, mash is transformed to a stable alcohol solution, which can be stored and used in a later process for producing ethanol; It could be also possible to produce fuel-ethanol direct from the mash. During the second day, the filtered mash is pumped into a container, passing through a bed of adsorbing hydrated crystal compounds (i.e. Fig. 1). The adsorbed

300

ETOH is then recovered by crystal regeneration and the surplus mash is evacuated and recycled in 2 steps: a. Bio-ethanol is adsorbed by the crystal compounds (~ 5 minutes); b. then it is recovered by low temperature heating (partial vacuum evaporation) and condensation (~ 4 minutes). The first results are already available, two years ago the process efficiency was for the 1st step : 25-30% of ethanol and 40-50% of ethanol for the second step. In November 2006 small scale test made of wine with 10% of alcohol give 40-50% of alcohol after the first step.

Figure 1: Distillation Process Scheme

Conclusion The world Bioethanol production is reaching now an high level (~ 40 mio m3/year). A large increase is expected in future (~ 75 mio m3/year for the year 2020). The incidence of energy costs on total bioethanol processing costs is considerable and of the order of 10%-25% of the total cost. From this the related interest for low cost distillation. International ethanol actors have recognised the possible potential of the GIAB-technology and they want to see more systematic research and development work concerning the technology. Focus of the development is to optimise the process. Globelive International AB owns the patents for ethanol production and is the owner of technology for producing of adsorbent.


301

Technological, Economic and Organisational Innovations in the Field of Energy-Saving in Industry Teaching Methods

Maksim Dli, Tatiana Kakatunova

Smolensk branch of Moscow Power Engineering Institute, Energeticheskyi proezd 1, Smolensk 214013, Russian Federation, E-mail: [email protected]

Abstract Best practice analysis shows that successful implementation of energy-saving steps in the industry is based on innovative character of the underlying solutions. This requires appropriate training of specialists and masters by providing their curricula with modules related to energy-saving including efficiency rating of innovative projects. All this defines appropriateness of introduction of an additional course at technical universities “Innovation management in the field of resource- and energy-saving”.

Industrial Parks and Innovative Management Teaching Best practice analysis shows that successful implementation of energy-saving steps in the industry is based on innovative character of the underlying solutions. This requires appropriate training of specialists and masters by providing their curricula with modules related to energy-saving including efficiency rating of innovative projects. All this defines appropriateness of introduction of an additional course at technical universities “Innovation management in the field of resource- and energy-saving”. Specific features of energy-saving processes in power-intensive industries define appropriateness of separation and investigation of separate categories of innovations providing economic, technical, environmental and social effect. It is obvious, that while estimating effects of the two latter types, extensive use of mathematical modelling methods, expert judgement methods, and in some cases - information intelligent software products is required. The resulting model of economic, environmental and social indices of innovative projects in the field of energy-saving can be applied to estimate energy-saving measures at the preliminary analysis stage. In the last several years, we can observe a shift in the field of energy-saving innovative management from federal to regional level. Mostly it is caused by the necessity of taking specific features of the region into account when defining priorities of investment policy directed to forming and implementation of Russian Federation regions’ innovative potential in the field of energy-saving. At the same time, the efficiency of energy-saving innovative programs is defined by the overall effectiveness innovative system, with industrial parks as its key elements. As world practice shows, an industrial park is also used as one of the basic instruments for state and regional innovative policy implementation [1].

302

Generally industrial park investigates perspective developments and provides scientists, innovators and inventors with financial, judicial, material, technical, consulting and information support. It usually doesn’t organise mass production, but helps the concept to evolve into the development of a new product or technology. It is obvious that regional industrial parks should be created considering specifics of existing innovative infrastructure. For example, if a region possesses a sufficient human resource potential for scientific research, then it is not necessary to staff the industrial park with full time employees who are dedicated to perform such research activities. More advisable in this case is to form temporary research groups for a specific problem. In general industrial park should include up-to-date laboratorial basis, marketing department and innovation dissemination department; also it should have multifunctional staff, etc. [2]. Considering innovative projects development and implementation capacities, related to support of energy-saving at regional level, it should be mentioned that in some Russian regions establishing of complex industrial parks which are meant to carry out the full cycle of innovation on their own, would clash with the lack of resources or just will not be reasonable. At the same time, orientation of the industrial park’s activities both on implementation of the whole innovative cycle and on implementation of its separate stages, makes it possible to increase the flexibility of the industrial park’s organisational management by the up-to-date network services. The industrial park, based on network services, can organise effective innovative process, based on efficient usage of innovative capacities and resources of the region by the arranging data exchange between them. Depending on amount and quality of innovative resources in the region, different number of enterprises and organisations can participate in implementation of innovative cycle, distributing specific functions, possible risks and future profits between themselves. In this case, innovative industrial park manager’s goal is to organise data cooperation and to coordinate data flows, which will connect all the participants of the innovative process in the region. This will make it possible to coordinate previously uncoordinated innovation potential at the regional level of Russian Federation in several cases. Also it is possible to employ foreign experts and to use foreign research expertise [3]. Depending on orientation on different functions and considering the region’s specifics, the following types of the industrial parks can be suggested, whose characteristics are given in Table 1.

303

Table 1 - Types of industrial parks, considering specificity of regions

Type Main Functions Type of Region Innovative industrial park

Development of innovative programs for a region together with government bodies; Development of innovative ideas and projects; Training experts in the field of innovative activity.

Has enough production and financial potential for unassisted creation and commercialisation of innovative products.

Marketing industrial park

Certification, patenting and promotion of innovative product; support in novelty commercialization.

Prevalence of structures, which generate innovative ideas and are able to create finished prototypes of innovative products.

Investment industrial park

Financing innovative project’s stages; search for innovation funding source; support in development of investment projects.

Prevalence of scientific and industrial structures, independent creative teams.

Manufactu-ring industrial park

Implementation of experimental development, creation prototypes of innovative production.

Prevalence of innovative activity subjects, which do not have trial production basis.

Distributed (virtual) industrial park

Arrangement of informational cooperation and coordination of innovative process’ participants in the region.

Has fundamental distributed innovative resources, which can be integrated into united innovative infrastructure.

Complex industrial park

Implementation of the whole innovative cycle.

Region with poorly developed innovative potential.

Firstly, the industrial park’s network organisation enables involvement of various participants for specific innovative program implementation. Secondly, availability of common information network allows participants of innovative process to carry out data exchange through direct channels in the course of project implementation. Another aspect of further progress of regional industrial parks can be their integration with neighbouring CIS countries’ innovative structures. All this defines appropriateness of studying at technical universities the innovation management topics at the regional level within the scope of disciplines “Regional economics” and “Innovations management”.

Conclusions The establishment of industrial parks in Russian Federation aimed at development and implementation of innovative energy-saving solutions means that introduction of an additional course at technical universities “Innovation management in the field of resource- and energy-saving” is justified and appropriate. The structure and concept of the course “Innovation management in the field of resource- and energy-saving” is discussed. Within the course specific features of innovations in the field, approaches to mathematical modelling results of resource- and energy-saving measures, and

304

also innovative projects development method on implementation of energy-saving programs of an industrial enterprise should be considered.

Acknowledgements The financial support of the European Community via the FP6 project TEMPUS-TACIS JEP_26045_2005, ECORSE "Ecological and Resource Saving Engineering" is gratefully acknowledged.

References [1] Golichenko OG. National innovation system of Russia: the position and way of development. Moscow: Science, 2006. [2] Shukshunov VE, Varuha AM. Technoparks: definitions, signs, indexes of activity. Innovation 1998:2-3. [3] Dli M, Kakatunova T. Innovative activity: regional aspects. Smolensk: Smolensk CNTI, 2007.


305

A successful heat integration at Lube oil Refinery Unit in Danube Refinery

Zoltán Varga; Tibor Karmacsi; Klára Kubovics Stocz; Alexandra Szűcs

MOL Plc, Százhalombatta, Hungary, email: [email protected] Improvement of the energy utilization in the crude oil refining industry is inevitable from the following reasons. One hand the cost of the energy is continuously increasing, on the other hand the fears from the global warming have led to legislative actions for decreasing in the green house gas emission. Objects of our study were to improve the energy efficiency in the Lube Oil Refinery Unit of the Danube Refinery as well as decrease in the fouling of the water cooling heat exchangers being caused the high temperature of the hydrocarbon streams. The unit was modeled with commercial type process simulation software (Hysys) on the basis of plant data. The Lube Oil Refinery process was divided into three parts namely feed preheating, product cooling and heat exchangers between the columns. Study of the feed preheating part showed that 548 kW of hot and 548 kW of cold utility can be saved by rerouting of the stream being applied for the preheating the feed. Investigation of the product cooling line showed that 621 kW of hot and 621 kW of cold utility saving is possible with implementing new heat exchangers and rerouting the product stream. Energy integration of the heat exchanger network of heat exchangers between columns was done with the Supertarget software. Results showed that 1200 kW of hot and 1200 kW of cold utility saving can be reached by restructuring the heat exchanger networks. As a result of the new heat exchanger network the hydrocarbon streams enters into the water coolers such a temperature level that decreases the formation of scaling, too. On the basis of our study the reconstruction of the heat exchanger network of the unit was implemented in the last year. Results of the test run being made after the start up has confirmed the calculated data.


306

Mapping of Key Problems in Energy Saving Research

Andrzej Kraslawski, Yuri Avramenko Lappeenranta University of Technology, Finland, [email protected]

Identification of the most important literature references and gaining a “panoramic view” of the discipline is a commonly encountered situation when a research is initiated in a given field. Usually a literature research takes a long time and requires a considerable effort. The growing amount of information makes this process even more difficult. Moreover, there is an increasing pressure in academia and industry to shorten time of research and development cycle of new products. A “panoramic view” of the discipline is built basing on an analysis of major research problems and not the most cited authors or references, as it is done at the moment. Therefore a new approach, for mapping a discipline, is needed. This paper proposes a method of problems mapping in a given technical or scientific discipline. The presented approach is a first attempt to combine the problems and research papers to show the structure of the given discipline as the existing methods concentrate exclusively on the analysis of word frequencies and co-citation analysis. The process is composed of three steps. First, there is performed an identification of the research problems in a given field. And next, the classification of those particular problems into the major research issues is done. In the second step, the visualization of the intensity of research in the specific fields is realized. The third step consists in the identification or the core papers which most strongly influence the research of the previously identified major issues in the given discipline. The presented approach is applied to the identification of the major research issues and obtaining the panoramic view of in the research on energy saving. The dynamics of the changes in the research priorities in 1997-2007 has been presented. The applied approach shows a clear structure of the reference papers in the field, showing papers related to the given research problem category. The final result are two maps – a diagram illustrating the change in the studied problems in a specified period of time and a figure showing the relations between the major papers and groups of the problems.


307

Brent Field 3D Reservoir Simulation

Julius M. Tollas Formerly of Exxon Houston, USA

The Brent Field consists of two separate reservoirs, and multiple pressure zones, with variable fluid properties and condensate rich gas caps. These were developed simultaneously by common facilities, presenting many reservoir engineering challenges that cannot be met by "classical" reservoir engineering tools alone. Reservoir simulation has thus become an integral and essential part of reservoir studies. The results of the third generation 3D model simulation studies confirmed the viability of the base development plan. They, however, also quantified the benefits of alternate development schemes for the final period of field depletion. The follow up reservoir simulation runs indicated that large amount of additional gas and oil volumes could be produced by a controlled depressurisation of the reservoirs rather than abandoning them at high pressure as it was previewed in the original development plan. Shell and Exxon spent $2 billion in the early 1990-es to redevelop the Brent field for depressurisation. Actual performance of the Brent field under depressurisation during the past 15 years was slightly better than the simulation indicated.


308

Energy Issues in the Restructuring of EU Sugar Industry

Krzysztof Urbaniec, Jacek Wernik CERED Centre of Excellence, Warsaw University of Technology, Płock Campus,

Płock, Poland The European beet sugar sector is currently subject to a deep restructuring stimulated by the reform of EU sugar market. The sugar production quotas are decreasing, the beet growing area is shrinking, and the number of sugar factories is going down. In a number of EU countries, the production of sugar has already been stopped, or will be stopped in coming years. The planned outcome of the restructuring process is that only the most competitive manufacturing companies continue beet sugar production in the EU. Apart from factors relating to the raw material, the competitive position of a beet sugar factory depends on its energy efficiency and ability to reduce the consumption of fossil fuels. With the current sugar manufacturing process set at obtaining high-quality white sugar as the main product, many sugar factories in the EU are operated close to the thermodynamic limit of energy efficiency and therefore the potential for improvement is very small or non-existing. However, there is a possibility to follow the path of development of the cane sugar industry in South America, that is, to produce sugar and ethanol as two main products. This may open new prospects for sustainable processing of sugar beet in Europe but the changed production process and its energy requirements still require a closer investigation aimed at the identification of optimum solutions. Similarly, new concepts for biorefinery processing of sugar beet need to be investigated taking energy requirements into account. The consumption of fossil fuels for energy supply to a sugar factory depends not only on the energy efficiency but also on the factory’s ability to include renewables in the energy mix. A possible renewable fuel co-produced with sugar is the beet pulp, that is, organic substance remaining after sugar extraction from beet tissue. Although the current price levels stimulate the use of beet pulp as animal feed, the situation may change in the future. As the pulp discharged from extraction equipment has a high moisture content, various processing routes are envisaged for converting it into material suited for energy generation. After drying, the pulp can be burned or ad-mixed to solid fuels. Alternatively, wet pulp can be used for biogas production. The needs for research aimed at working out engineering solutions for sustainable processing of sugar beet in the EU have been investigated by the participants in EU FP6 project “Towards Sustainable Sugar Industry in Europe - TOSSIE”. The project results confirm that energy issues are of key importance to the future of EU sugar industry.


309

On large point sources for CO2 and the emerging challenge relating to coal-based power generation in a global perspective

Jens Hetland

Deptartment of Thermal Energy, SINTEF Energiforskning AS, Trondheim, Norway Apparently there is a subset of major concerns that appear contradictory by nature – such as the issue of security of energy supply and the global warming issue. Although there is a quest for new energy technologies, it seems evident that no real option to fossil fuels is (so far) identified that is capable of supplying sufficiently large quantities of power on demand at an acceptable cost. Hence, the harnessing of fossil fuels has become a geopolitical issue of growing concern. Hence, it is interesting, relevant and necessary to analyse the impacts and drivers for large power generation technologies, in order to understand the future – also when dealing with technologies at smaller scale (EMINENT-2). The presentation intends to give a survey of the global dynamics relating to large-scale power generation – particularly the use of coal. Emphasis will be placed on the largest point sources for CO2 in the world by country and plant. Basis will be information retrieved from the IEA GHG CO2 Emissions Database, and the Sino-European project COACH that has been set to pave the ground for carbon capture and storage in China (CCS) under the auspices of the European Commission. Furthermore, the presentation will also deal with opportunities pertaining to emerging mitigating technologies in a reasonable time scale.


310

Sugar beet based bio-refinery concepts - the results of sector wide process synthesis

Gernot Gwehenberger, Michael Narodoslawsky Institute for Process Engineering, TU-Graz, Graz, Austria

The European sugar sector is currently faced with two challenges that may well influence the structure of sugar industries in the future: sugar production is exposed to an increasing global sugar competition as well as the pressures associated with the increased demand of bioenergy in general and biofuels in particular, leading to competition for raw materials, too. The current work looked first into the environmental performance of the sugar sector. This analyses revealed that the European sugar sector (based on sugar beet) is already well on its way to become sustainable but that there is still a wide margin for improvement. In a second part the structural challenges of the European sugar sector were investigated with the tools of process synthesis, employing the combinatorial P-graph method combined with a branched and bound MINLP optimisation. The key results of the process synthesis are the following:

• At current price structures the demand for ethanol as an alternative fuel makes the ethanol production more competitive than the original sugar production. This may in the long term lead to a major re-alignment of the sugar sector;

• Employing new technologies such as the “Green Biorefinery” to utilise agricultural by-products (e.g. beet leafs) to produce basic chemicals (lactic acid, amino acids, possibly fibres) is becoming increasingly competitive as crude oil prices soar;

• Utilising energy from the Green Biorefinery as well as from sugar beet chips via biogas will increase competitiveness of sugar production (and/or ethanol production from sugar beets), especially by providing electricity as a by-product. There will however remain an energy gap which must be covered by external energy sources.

The results indicate that sugar beet is a very interesting crop as a base for industrial processes. Given the existing installations, the long term relation with the farming sector and the established logistics for sugar beet, the sugar sector is well positioned to play a pivotal role as a key sector for the development of a sustainable European industry.

ENERGY FOR SUSTAINABLE FUTURE SUPPLLEMENT Edited by Petar Varbanov, Jiří Klemeš, Igor Bulatov University of Pannonia, Veszprém, Hungary, 5-6 May 2008 Copyright © Manuscript Authors, ISBN 978-963-9696-38-9

311

Local learning energy-efficiency networks – a mean of dissemination for knowledge about new energy technologies

Martin Jakob, 1Eberhard Jochem

CEPE, ETH Zurich, Zürichbergstr, 18, CH 8032 Zurich, Switzerland, e-mail: [email protected]; fax:++41 44 632 1622

Abstract Energy consultants report large unused potentials for energy cost savings. These potentials represent not only new and sophisticated technical solutions, but often small and very profitable investments or organisational measures. One of the main obstacles are high transaction costs for search of information of possible solutions and for the decision processes involved. To reduce transaction costs for energy-efficient solutions, locally organised learning networks of companies have been established in Switzerland, in Germany, and other European countries. The companies of these efficiency networks realise new energy technology at least twice as fast as other companies quickly taking up new energy technologies by mutual exchange of experiences. Presently, a network management system for initiating and operating the efficiency networks at high quality level is being developed. This system could integrate the EMINENT platform, a database and evaluation tool of early stage energy technologies, into its communication system. The results of EMINENT can be particularly used by the moderator and consulting engineers of those networks and the energy managers of larger companies and also by speakers of new technologies who are invited by those networks to report on new technologies.

Introduction Energy analysts and intermediaries often deplore that large potentials for energy cost savings are not used and that new technologies are disseminated into the market place only to a quite low pace. One of the main obstacles are high transaction costs for search of information of possible solutions and for the decision processes involved. Among others, there are two types of promising policy instruments that aim at reducing such transaction costs. One of them is the EMINENT project that aims at accelerating the market introduction of promising early stage technologies by means of a early stage technology database and assessment tool. The other promising instrument are local learning energy networks that aim at reducing energy costs and emissions by exchanging experiences and by committing to (voluntary) targets. This paper explores whether there are synergies between these two policy instruments. More specifically it is examined to which extent local learning networks could make use of the EMINENT database and tool.

ENERGY FOR SUSTAINABLE FUTURE SUPPLLEMENT

312

Locally organised learning networks of companies have been established in Switzerland, in Germany, and other European countries. Such networks consist of groups of ten to twenty companies and their goal is to exchange information and experiences on technical and organisational measures to reduce energy costs and to implement such measures. Local learning networks on energy efficiency are dating back on a voluntary initiative of some energy-intensive companies in the mid 1980s in the region of Zurich, Switzerland, who were afraid of increasing electricity prices and even rationing. The instrument was so successful in reducing specific energy demand of the participating companies (Bürki 1999 [1], Graf 1996 [2], Kristof et al. 1999 [3], Konersmann 2002 [4]) that the principle was included into the Swiss legislation on CO2 mitigation and that one of the authors transferred the concept to Germany in 2002 initiating a first demonstration project in the North-East of Baden-Württemberg within regional “learning networks” of about 15 companies each. The EMINENT project was launched by DG TREN as a node in the OPET network, with the main purpose of identifying and accelerating the introduction and implementation of leading edge European energy and environmental technology into the worldwide market place. Within EMINENT a database and software evaluation tool was being developed. It is capable of identifying and evaluating the (techno-economic) potential of early stage technologies. One goal of EMINENT was to shorten the lead time from technology development into commercialisation by generating reliable and professionally evaluated information about the latest research and development that is easily accessible (Jansen et al., 2004 [6a], Klemeš et al. 2005 [6b]). Next to the development of a database and tool, dissemination activities were an integrative part of the EMINENT project. Such activities include the development of Technology Assessment Reports, the organisation of thematic workshops, an interactive internet site and other kinds of technology transfer (Klemeš et al. 2005 [6b]). Within the follow-up project EMINENT2 it is the goal to disseminate information about new and promising energy technologies and to foster their use in companies, particularly in energy-intensive industries. Within EMINENT it is explored to which extent local learning network of such companies could serve as a platform for such a dissemination.

Experiences with local learning networks during the last 20 years The first 20 local learning efficiency networks have been initiated in Switzerland between 1986 and 2000, while the development in Germany started in 2002 reaching eight operating networks by the end of 2007 and at least further 8 networks starting in 2008. Major insights will be reported in the following sub-sections. The process of networking between 10 to 15 companies can be characterised as follows:


313

• collection of energy-related data within the companies according to a checklist, prior to an energy audit by an experienced engineer in each participating company,

• regular moderated meetings about three to four times per year with presentations by external energy experts and exchange of experiences among the participants,

• additional meetings in small groups on special technical areas or solutions,

• common target setting of energy savings and CO2 reduction within a defined time period such as three to four years,

• yearly monitoring of measures and investments undertaken and of energy consumption and CO2 emissions with computer-based models, observation of progress made and targets achieved taking into account structural changes, change of production, weather conditions, etc.

In several cases, scientific evaluation of the initiation and operation of learning networks was performed by qualitative social research methods by the authors. Jochem/Gruber. 2007 [5], demonstrating the effectiveness of the approach.

Experiences in Switzerland Local learning networks on energy efficiency are dating back on a voluntary initiative of some energy-intensive companies in 1986 in the region of Zurich, Switzerland, who were afraid of increasing electricity prices and even rationing due to a scarcity of supply. The principle was then further developed within the framework of Energy 2000, an energy-efficiency and renewable energy promotion programme of the Swiss Federal Office of Energy. Due to the success in reducing specific energy demand of the participating companies (Bürki 1999 [1], Graf 1996 [2], Kristof et al. 1999 [3], Konersmann 2002 [4]), the principle was included in the energy legislation of the canton of Zurich and into the Swiss legislation on CO2 mitigation which allowed for and an additional push. The Swiss CO2 law allows the exemption of the CO2 surcharge imposed on all energy consuming sectors if companies have individual efficiency and CO2 reduction targets that are accepted by the government. The implementation for industry and the service sector was handed over to the Swiss Energy Agency for Industry (EnAW) that is acting as an intermediary to negotiate target agreements on CO2 reduction between companies and the federal government and to monitor the audits of the performance of each company. Companies which reduce energy-related CO2 emissions by a negotiated target and accept a yearly evaluation can be exempted from a surcharge on fossil fuels of presently 12 CHF per ton CO2 which was approved by the Swiss Parliament in 2006. Hence, the number of local learning efficiency networks boosted from 20 in the year 2000 to presently 70 so-called “Energy Tables” with about 720 companies involved in this scheme, representing almost 3 million tonnes of CO2 or one quarter of the total CO2 emissions of the Swiss industry and service sectors. The target agreements are mostly based on energy efficiency improvements until 2010 or substitution options for fossil fuels such as industrial organic wastes, renewables, or electricity (which is almost CO2 free due to 60 % hydro power and 35 % nuclear power generation in Switzerland) The target agreements amount to 0.7 million tonnes of CO2 by 2010


314

where about 280,000 tonnes can be attributed to the energy efficiency networks and two other Programmes for small companies of the Agency (BAFU 2007 [7]). The efficiency networks were partially subsidised by the Swiss government in the first few years of their existence before 2000 and are now indirectly subsidised by federal money allocated to the Energy Agency for Industry by an average of 600 € per company and year. Today, the moderators and consulting engineers are mostly financed by the participating companies with contributions of some 6,000 to 30,000 CHF each per year, depending on the size or the energy costs of a participating company. The average annual energy cost savings were on average 165,000 CHF (or 100,000 Euro per year) per company (ranging from 13,000 € to 1.5 Mill. €). Important motivating factors for the Swiss companies to join one of the networks were the exemption from the CO2 surcharge, the reduction of energy cost, existing trustful relations between the companies and the local Chamber of Commerce or an other economic platform, and surprisingly a 10 % reduction of the electricity price for those companies that signed and fulfil the target agreement. There is some evidence that the efficiency successes diminish after some 7 to 10 years of operating such a local learning network which one might expect. However, the oldest local efficiency network in Zurich has celebrated its 20th anniversary in 2006 looking back to extremely positive and continuous results which one would not have expected during such a long period assuming that the participating companies run out of profitable efficiency potentials. During the last six years, this oldest network still improved its energy efficiency by 3 % per year (the average improvement of the German industry during the last ten years was only 0.9 % per year).

Observations and results learnt from German efficiency networks The initiative was so successful in reducing specific energy demand of the participating companies (Bürki 1999 [1], Graf 1996 [2], Kristof et al. 1999 [3], Konersmann 2002 [4]) that the principle was included in the energy legislation of the canton of Zurich and into the Swiss legislation on CO2 mitigation and that one of the authors transferred the concept to Germany in 2002 initiating a first demonstration project in the North-East of Baden-Württemberg within regional “learning networks” of about 15 companies each. Starting from the positive Swiss experiences, a first learning energy efficiency network was launched in the Hohenlohe region by the government of Baden-Württemberg in mid 2002 and a second one in the city of Ulm in 2005. Less than a year later, the large utility EnBW launched two additional networks, one in Baden-Württemberg, the "home" distribution area of the utility, but also one in middle-east Germany, the distribution area of the two large competitors, e.on and Vattenfall. By the end of 2007, more than 10 learning efficiency networks were operating. The targets for the networks are decided upon by the energy managers of the participating companies after the results of the initial consulting became available (see Table 1): The targets on energy efficiency and specific CO2 emission reduction are usually set for four years and broken down to a yearly target path with lower interim targets in the first two years and higher interim targets in the following years.


315

The meeting of the yearly targets was – and still is - monitored by the authors achieving interesting results for the first two networks in Germany (see Table 1):

• In the case of Hohenlohe with initial 17 companies and 20 sites, both targets were slightly surpassed;

• however, the larger nine companies of this network who continued the cooperation after the end of public financial support already had better results during the first four years and achieved an efficiency increase of more than 18% within five yeas - i.e. an unexpected high improvement of 4 % per year in the average. and almost a similar success for CO2 emission reduction.

• In the case of a network with eight companies in Ulm (excluding two participating energy converting companies, a municipality producing and distributing electricity and district heat, and a second university based company) one had substantial success with regard to energy efficiency improvement by 2.7 % per year, whereas the reduction target of the specific CO2 emissions of some 1.9 % per year was not fully reached.

Table 1: Overview of the targets and of meeting the targets for energy efficiency

and specific CO2 emissions of the local learning networks Hohenlohe and Ulm

Efficiency network Targets set Targets reached energy-

efficiency specific CO2-

reduction energy-

efficiency specific CO2-

reduction Hohenlohe - 17 companies, (4 years)

2001-2005

7 %

2001 – 2005

8 %

2001 – 2005

7.8 %

2001 – 2005

10 % - 9 companies1), (5 years)

2001 – 2006 8.6 %

2001 – 2006 10 %

2001 – 2006 18.3 %1)

2001 – 2006 17.7 %1)

Ulm - 8 companies (with out utilities)

2004 – 2008

7 %

2004 – 2008

8 %

2004 – 2006

5.3 %3)

2004 – 2006

3.8 %3)

- 9 companies (including utilities)

s.o.2) s.o.2) - 3.1 %4) 13.9 %4)

1) Die neun verbleibenden Unternehmen hatten in der Periode 2001 – 2005 höhere Effizienzgewinne und spezifische CO2-Minderungen als diejenigen Unternehmen, die mit Ende 2005 ausschieden.

2) nicht besonders genau die speziellen Bedingungen der Energiebetriebe a priori bedacht 3) bei guter Kapazitätsauslastung in 2006 durch gute Konjunktur 4) Verschlechterung der Energieeffizienz und sehr große CO2-Emissionsverbesserung durch in

Betriebnahme einer Holzhackschnitzel-KWK-Anlage Sources: Jochem et al. 2006 [8], Jochem and Gruber 2007 [5]

• The energy efficiency network Ulm also includes two energy converting and distributing companies which dominated the energy demand, but also the changes of the whole network. So it became quite clear that a substitution of a cogeneration plant in 2006 fired by wood chips instead of heating oil would dominate the outcome of the yearly monitoring process: The change from a liquid fuel to wood chips decreased the efficiency of the whole group (leading to less energy


316

efficiency of the whole group by 3 %, but dominated the reduction of specific CO2 emissions by almost 14 % or about 7% per year (see Table 1).

Almost all companies had already implemented some energy efficient solutions before they joined the network. However, time restrictions and lacking specialised staff hindered the intensification of activities. The companies were very interested in more know-how and reliable information as well as in a systematic review of weak spots and efficiency potentials in their buildings, factories, and processes. According to the results of the energy audits the participants concentrated during the first two years on more efficient electricity use, e. g. in the area of lighting, ventilation, compressed air, cooling, or pumps, as well as small measures of fuel savings in heat production and use, such as insulation of pipes, improvement of control systems, etc. Many of the measures were organisational and low-cost measures, which could be implemented in a short-term period. Larger investments, mainly in fuel and heat efficiency refer to areas with long-term re-investment cycles and were postponed until the re-investments will be made. However premature investments were sometimes taken into account. Finally after four years, 60 % of the measures recommended by the energy consultant were implemented or at least planned in detail. Again, it was interesting to note that the monitoring of the yearly undertaken investments and organisational measures concluded that about 50 measures were realised per year, but additional 30 measures were additionally identified. These new measures have been identified due to several reasons; new technical solutions developed by the investment goods industries, cost decreases because of learning and economy of scale effects, new technologies coming to the market, new integrative and intelligent solutions of already existing technologies, and finally higher energy prices that made technical potentials profitable. One interesting observation was the fact that 10 companies (mostly smaller ones with less than 150,000 € yearly energy cost) who left the network of Hohenlohe after two years; immediately fell back to the industrial average of efficiency improvement of about 1 % per year. An indicator of the success of the local learning efficiency networks was the fact that almost all the participants continued their co-operation in the network after the public funding ended paying a yearly contribution between 4,000 € and 20,000 € for moderation of the meetings and monitoring of the yearly performance – according to the company's size. There was little fluctuation in the German networks and almost none in the Swiss networks after to CO2 law came into practice after 2000. The evaluation of the first German energy efficiency network in Hohenlohe was so convincing that the Ministry of Environment and Transportation in Stuttgart set up a financial incentive programme providing efficiency learning networks with one third of the total cost (including a scientific evaluation) for two years. Under this new programme, the network in Ulm started operating in April 2005 and in East-Württemberg in mid 2007.

Economic results


317

In absolute monetary terms, energy cost savings due to efficiency improvements after the first four years amounted to an average of about 100,000 € (of which two thirds to 80% were electricity cost savings) with large variations depending on the total energy cost of the companies and the efficiency improvement achieved. The measures taken were all highly profitable ranging between 15 and 50 % internal rate of return, and the average net profits (after deduction of the taxes and the cost for participating in the network) amounted to 10 to 20 € per tonne of CO2.

In all cases it was no doubt that the saved energy cost would refinance the capital cost for the investments or to pay for the organisational cost of labour emerging from organisational measures. The results confirm the substantial potential of immediate organisational measures mostly realised within the first two years of the networks. Investments in fuel and heat efficiency are more often related to the production processes or heating equipment with very long-term reinvestment cycles that cannot be changed if stranded investments are to be avoided.

Accompanying monetary benefits, e.g. better utilisation of equipment (i.e. reduced capital cost), less rejects, higher product quality, less need for air conditioning and ventilation, etc. are often not taken into account because it is difficult to describe them in monetary terms, but they are important arguments, e.g. for energy managers in discussion with their company managers. There was no measurement of the reduction of the transaction cost induced by sharing the experiences during the meetings and also by bi-lateral exchange of information and experiences among the energy managers of the participating companies, because the scientific evaluation had not planned to work with control groups to identify the differences of the transaction cost. First estimates of the moderators involved were between three quarters and 90 % reduction of the transaction cost due to the local learning efficiency network. In any case, empirically based figures on transaction cost would be weak; more importantly is the aspect that the exchange of knowledge and experiences among the energy managers increases the number of cases they can handle in a given limited time of their work that increases their efficiency between a factor of two to four relative to the industrial average of energy efficiency improvements. And the observations of the network in Zurich suggest that this acceleration of the impact of the energy managers can last for 20 years. It has proved in Germany that utilities can develop new networks faster than other institutions (like economic platforms, chambers of commerce, or consulting engineers), as they have close connections with their customers and they start realizing that the networks and their successes to save energy cost induce customer binding effects. They may also have better consulting engineers and more ready to use the EMINENT data base as one source of information. If the operation of the networks is not co-ordinated by one institution, there is a risk that individual institutions or private consultants initiate those local learning networks on their own operating them at low quality standards. In order to avoid this risk, five partners, an economic platform, two research institutes, a large electricity company (EnBW), and one large manufacturer, Siemens, have started a research and development project in June 2007 to generate a management system for Local


318

Learning Energy Efficiency Networks (LEEN). This will be the basis for professional training of moderators and consulting engineers and a related certification in Germany, and probably later in many European countries. This management system would guarantee a minimum quality standard of future networks an could include the EMINENT data base as one major source of knowledge on new energy technologies.

Perspectives using EMINENT as important source of information and dissemination

The authors have started contacting relevant persons in Switzerland, the EnAW as the principal agent of the Swiss economy that builds the bridge between the companies involved in the 70 local learning networks and the Swiss federal government in the context of the CO2 law. One author has also close contact to about 10 existing learning networks on energy efficiency in Germany. The observations and consideration of the planned dissemination via existing local learning networks lead so far to the following insights and first conclusions:

(1) The simple hint to the web side of EMINENT by written or oral information will not induce sufficient interest for most companies to visit the site at all. This is the main reason that the energy managers of the participating companies are unlikely to visit the site of EMINENT.

(2) Instead, the moderator of each local learning network (and/or the consulting engineer of the network) will have to present major aspects, benefits, and opportunities of the new energy technologies described in EMINENT suited to the needs of the participating companies. They should also select the technologies according to the needs and opportunities of the participating companies.

(3) The information given to the participating companies should contain a rough technical description with major features, pre-conditions, and economics. Essential is also the description of a real application with the address of the company applying the new technology and of the manufacturing company. This is still a challenge of the data bank of EMINENT.

(4) Larger and innovative participating companies of the learning networks are likely to invest into the technology first, because the understanding and knowledge of the energy managers of the larger companies is more developed alleviating the introduction of new technologies as first movers. Larger companies can also take larger financial risks when the financial benefits of the new energy technology do not realize.

(5) A presentation by a salesman of the manufacturers of the new technology may be touchy regarding the acceptance of the new technology. Also a consulting engineer who is not familiar with the network should not be asked to present the new technologies.

(6) A list of two or three consulting engineers being familiar with the new energy technology and the integration in existing or new plants may be appreciated by


319

the companies in case they want to follow the new idea. He should also be able to assist in the concrete planning process.

(7) In order to speed up the diffusion of the new technology pushing it from classification 4 to 5, the generation of new efficiency networks in several European countries becomes important.

(8) Providing the planned e-based distance and face-to-face training for energy agencies and industrial associations becomes important now. e-Training courses should be offered in autumn 2008 to interested moderators. consulting engineers, and interested companies.

Conclusions In almost all companies, there is a high potential of profitable energy efficiency measures. Even energy managers from large companies did not know about all the new innovations of efficient energy use and were able to inform themselves about the new options faster than was the case without the learning efficiency network. The intensity of sharing of experiences increases with the operating time of the network as the trust within the group of the participating companies increases. Even companies of the same industrial branch co-operate with regard to cross-cutting technologies as long as they are not immediate competitors. The participants confirmed a substantial reduction of transaction costs and faster implementation of measures as a result of the meetings, sharing experiences and know-how, and the experts invited. Some of the participating companies also expect an improved image from the energy efficiency networks. The existing 70 networks ("Energy Models") in Switzerland give some indication of the potential of this instrument in Germany and Europe - or in industrialised countries. The experiences in Switzerland and Germany suggest so far that at least 300 efficiency networks in Germany would be feasible within the next five to eight years, and more than 1'000 in Europe. The authors and the five partners of LEEN suggest a first step of 30 efficiency networks in Germany as a demonstration project, and a first survey how the EMINENT data bank is used and accepted by the relevant users described above. For this purpose, the planned e-based distance and face-to-face training of EMINENT become important, to study this option of dissemination of knowledge on new energy technologies.

Within this context it has to be stressed that there is a missing field of new energy efficient solutions at the useful energy level in EMINENT. Energy managers are most excited if they can reduce losses at the useful energy level by their investments as they can reduce the size of energy converting equipment or have extra capacity of energy conversion for production growth without being forced to invest in new energy converting and distributing technologies and installations. It is also recommended to consider the instrument of the LEEN at the level of the European Commission and the governments of the member states as a potentially


320

very effective instrument for picking the "low hanging fruits" of efficient energy use in the first phase and developing a new culture of efficiency attention in the longer term in Europe and elsewhere.

Acknowledgements The authors have been able to evaluate this interesting option of dissemination of new energy technologies via the EMINENT data base by combining research funds of the European Commission of the EMINENT project and research funds by the German Environmental Foundation, and the two Federal States of Baden-Württemberg and Hessen allowing to develop the network management system for local learning efficiency networks.

References [1] Bürki, T.: Das Energie-Modell Schweiz als Erfahrungsfaktor für Schweizer

Unternehmen. Bundesamt für Energie: Energie 2000, Ressort Industrie. Benglen 1999.

[2] Graf, E.: Evaluation des Energie-Modells Schweiz. Bern: Bundesamt für Energie: 1996.

[3] Kristof, K. et al.: Evaluation der Wirkung des Energie-Modells Schweiz auf die Umsetzung von Maßnahmen zur Steigerung der Energieeffizienz in der Industrie und seiner strategischen energiepolitischen Bedeutung. Bern: Bundesamt für Energie: 1999.

[4] Konersmann, L.: Energy efficiency in the economy – Evaluation of the Energy Model Switzerland and Conception of a multi-agent model. Master Thesis (in German). ETH Zurich, 2002.

[5] Jochem, E., Gruber, E.: Local learning networks on energy efficiency in industry – Successful initiative in Germany. Applied energy 84 (2007): 806-816

[6a] Jansen, P., Koppejan, J., Hetland, J., Klemeš, J., Phuengphaeng, T., Pipio, A.: EMINENT accelerates market introduction of promising early stage technologies for transport and energy, 2004.

[6b] Klemeš, J., Zhang, N., Bulatov, I., Jansen, P., Koppejan, J.,: Novel Energy Saving Technologies Assessment by EMINENT Evaluation Tool. In: CHEMICAL ENGINEERING TRANSACTIONS, Volume 7, 2005.

[7] Bundesamt für Umwelt (BAFU): Anhang zur Vollzugsweisung: Verpflichtungen und Zielvereinbarungen – Beschreibung der Zielvereinbarungsmodelle, Berichter-stattung. Ittingen 2. Juli 2007

[8] Jochem, E., Gruber, E., Weissenbach, K., Westdickenberg, J., Feihl, M., Ott, V.: Modellvorhaben Energieeffizienz-Initiative Region Hohenlohe zur Reduzierung der CO2-Emissionen 2002 – 2006. Schlussbericht für das Umweltministerium Baden-Württemberg. Karlsruhe/Waldenburg, 2006.