The green struggle

Citation for published version (APA):Kirkels, A. F. (2016). The green struggle: development and trajectories of biomass gasification. Eindhoven:Technische Universiteit Eindhoven.

Document status and date:Published: 16/06/2016

Document Version:Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can beimportant differences between the submitted version and the official published version of record. Peopleinterested in the research are advised to contact the author for the final version of the publication, or visit theDOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and pagenumbers.Link to publication

General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, pleasefollow below link for the End User Agreement:www.tue.nl/taverne

Take down policyIf you believe that this document breaches copyright please contact us at:openaccess@tue.nlproviding details and we will investigate your claim.

Download date: 11. Jun. 2020

Arjan Kirkels

The Green StruggleDevelopments and trajectories of biomass gasification

The Green Struggle Developments and trajectories

of biomass gasification

Arjan Frank Kirkels

2016

This research has been made possible by: Eindhoven University of Technology, School of Innovation Sciences. A catalogue record is available from the Eindhoven University of Technology Library. ISBN: 978-90-386-4085-3 Copyright © 2016, Arjan Kirkels Cover design & printing: Proefschriftmaken.nl || Uitgeverij BOXPress Published by: Uitgeverij BOXPress, Vianen All rights reserved. No parts of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the author.

The Green Struggle Developments and trajectories

of biomass gasification

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de Technische Universiteit Eindhoven,

op gezag van de rector magnificus, prof.dr.ir. F.P.T. Baaijens, voor een commissie aangewezen door het College voor Promoties,

in het openbaar te verdedigen op donderdag 16 juni 2016 om 16.00 uur.

door

Arjan Frank Kirkels

geboren te Maarheeze

Dit proefschrift is goedgekeurd door de promotoren en de samenstelling van de promotiecommissie is als volgt: Voorzitter: prof.dr. I.E.J. Heynderickx 1e promotor: prof.dr.ir. G.P.J. Verbong 2e promotor: prof.dr. F. Alkemade leden: prof.dr. A.P.C Faaij (Rijksuniversiteit Groningen) prof.dr. W.J. Watson (University of Sussex) prof.dr.ir. A.M.C. Lemmens prof.dr. K. Frenken (Universiteit Utrecht) dr. H.A. Romijn Het onderzoek dat in dit proefschrift wordt beschreven is uitgevoerd in overeenstemming met de TU/e Gedragscode Wetenschapsbeoefening.

“It is not easy being green”

Kermit the Frog

“Vrees niet langzaam te gaan, vrees slechts stil te staan”

Acknowledgements My PhD career was a bit different compared to others. After working at the university for about 20 years, I finally have acquired my doctoral degree. Until 2007 I have been working at the center Technology for Sustainable Development (TDO) at Eindhoven University of Technology. Over these years I evolved in a multidisciplinary engineer with a focus on sustainability and energy issues, an expertise that I mainly applied in teaching. By 2007 it became apparent that the center would seize to exist and I started to look for new career opportunities. Despite my love for teaching, I felt that I was up for something new. The opportunity presented itself to take on a position that combined lecturing and PhD research in the group of Geert Verbong, a group with a focus on Innovation Science. Geert and I knew each other both from our work for TDO, as well as from my studies during which he was my master thesis supervisor. At the time I was mainly involved in studying renewables, energy from biomass and assessing the environmental impact and sustainability of technologies. I had joined study tours to Finland and Sweden, visiting several energy-from-biomass plants – including the well-known biomass gasification plant in Värnamo. On one of these tours Kees Daey Ouwens was a supervisor, a prominent Dutch expert in the field of biomass gasification. Later on he would become a close colleague. He successfully plugged his enthusiasm for energy from biomass and more specifically biomass gasification – to me and many others. When I decided to become a PhD, I had a strong preference for a topic in this field. This new career path of pursuing a PhD brought opportunities and challenges. It allowed me to deepen my insights in doing research and writing. It also gave me the opportunity to further broaden my already multidisciplinary background. And, of course, I could continue teaching, writing textbooks and experimenting with different didactical formats. A first challenge – one that I had anticipated on – was to fit in a research group on Innovation Science, while my background was mainly in industrial ecology and in feasibility studies. Especially in the first few years it took me a lot of effort to catch up. The second challenge was the (lack of) availability of time and energy: being a family man, doing a lot of teaching and doing research required juggling with time and tasks – and a patient boss that kept the hope on a successful ending alive. Life outside the office was at least equally challenging: trying to keep my family happy and positive. I certainly hope that this was a path dependent process: one in which we can learn from the past and from which we can benefit for the future. Of course my family

1

also contributed extensively to where I stand today. Carien, Jochem, Lieke: can’t do without your love and support. For this thesis I would like to specifically thank my three bosses over the years: Han van Kasteren, Lex Lemmens and Geert Verbong. I would characterize each of you as heterogeneous engineers that look beyond pure science. You all share with me a warm interest in sustainability and in teaching. And each of you provided me with the opportunity to explore my passions and shape my career. Floortje, thanks for your willingness to jump on the bandwagon and join my research in the last year. I appreciate your support and constructive feedback that helped me tying things together. Mieke, thanks for making me feel welcome; Hendry and Val for the practical support in the last stages. Thanks to the colleagues with whom I had the pleasure to closely cooperate with on teaching: Fred, Boukje, Jacob, Johan, Henny and Geert. And of course many thanks to all others: roommates, lunch groups, TIS group members, people from SHT, educational and didactical support, and certainly also our students. Arjan Kirkels Januari 2016

2

Abbreviations & Acronyms ATM Atmospheric BFB Bubbling Fluidized Bed BTL Biomass to Liquids BIG CC Biomass Integrated Gasification Combined Cycle CCS Carbon Capture and Storage CDM Clean Development Mechanism CFB Circulating Fluidized Bed CFD Computational Fluid Dynamics CHP Combined Heat and Power DME Di Methyl Ether EPO European Patent Office FICFB Fast Internally Circulating Fluidized Beds FT Fischer Tropsch GDP Gross Domestic Product GHG GreenHouse Gas HTU Hydro Thermal Upgrading HTW High Temperature Winkler IEA International Energy Agency IGCC Integrated Gasification Combined Cycle IPC International Patent Classification LEBEN Large European Biomass Energy Network MLP Multi Level Perspective Mtoe Mega ton oil equivalent MSW Municipal Solid Waste MW Mega (=106) Watt (kW kilo =103; gW Giga = 109) MWel capacity in MW electrical power MWth capacity in MW thermal power LCA Life Cycle Assessment odt/d oven dried tonne / day ORC Organic Rankine Cycle RD&D Research, Development and Demonstration RDF Refuse Derived Fuel RME Rapeseed Methyl Ester SNG Synthetic Natural Gas SNM Strategic Niche Management SOFC Solid Oxygen Fuel Cell

3

SPLC Search Path Link Count SPNP Search Path Node Pair SRC Short Rotation Coppice SRF Short Rotation Forestry ST Science and Technology (indicators) STIG Steam Injected Gasturbines t/a ton per annum TPES Total Primary Energy Supply TIS Technological Innovation Systems TWh Tera Watt hour USPTO United States Patent and Trademark Office VTT Technical research centre of Finland WIPO World Intellectual Property Organization

4

Brief contents Acknowledgements 1 Abbreviations & Acronyms 3 Brief contents / Table of content 5 / 6 1. Introduction 9

2. Discursive shifts in energy from biomass: a 30 year European overview 21

A discourse analysis of dominant drivers, feedstock, technology and applications

3. Biomass gasification: still promising? A 30 year global overview 45 Dynamics in and status of coal and biomass gasification

4. Punctuated continuity: the technological trajectory of advanced biomass gasifiers 71 Reconstruction of emerging trajectories based on shared expectations.

5. Biomass boom or bubble? A longitudinal study on expectation dynamics 99 Longitudinal dynamics in expectations and hype-disappointment cycles

6. Embedding emerging of technological trajectories: a patent analysis of biomass gasification 133 Reconstruction of emerging trajectories by a patent-citation approach, embedded in an overview of real-time developments and spillover effects.

7. Synthesis: conclusions and discussions 163 References 177 Appendix A Gasifier types 195 Appendix B Demo plants 199 Summary 201 Curriculum Vitae 207

5

Table of content Acknowledgements 1 Abbreviations & Acronyms 3 Brief contents 5 Table of content 6 1. Introduction 9 1.1 Scope of this research 1.2 Biomass gasification: a promising technology 1.3 Theoretical background 1.4 Research question and methodology 1.5 Structure of this thesis 2. Discursive shifts in energy from biomass: a 30 year European overview 21 2.1 Introduction 2.2 Methodology 2.3 European Biomass Conference 2.4 Literature study based discourse 2.5 Quantitative analysis discourse 2.6 Discussion and conclusions 2.7 Future of energy from biomass 3. Biomass gasification: still promising? A 30 year global overview 45 3.1 Introduction 3.2 Theoretical framework and methodology 3.3 Gasification 3.4 Biomass gasification 3.5 Science and technology indicators 3.6 Conclusions and discussion

6

4. Punctuated continuity: the technological trajectory of advanced biomass gasifiers 71 4.1 Introduction 4.2 Conceptual framework 4.3 Methodology 4.4 1970s and 1980s: methanol as transport fuel 4.5 1990s to 2004: IGCC for high-efficiency power generation 4.6 After 2000: biofuels 4.7 Conclusions and discussion 5. Biomass boom or bubble? A longitudinal study on expectation dynamics 99 5.1 Introduction 5.2 Theoretical background 5.3 Methodology and sources 5.4 Indicator trends 5.5 Late 1970s – early 1980s: methanol production 5.6 1990s: IGCC power production 5.7 After 2000: fuel production 5.8 Conclusions and discussion 6. Embedding emerging technological trajectories: a patent analysis of biomass gasification 133 6.1 Introduction 6.2 Uncovering patent citation networks 6.3 Methodology 6.4 Developments in fluidized bed gasification 6.5 Interrelated technological developments 6.6 Technological trajectories 6.7 Conclusions and discussion 7. Conclusions 163 7.1 How did biomass gasification develop over time? 7.2 Did this result in emerging trajectories? 7.3 Discussion

7

References 177 Appendix A Gasifier types 195 Appendix B Demo plants 199 Summary 201 Curriculum Vitae 207

8

Chapter 1

Introduction This study traces the emerging technological trajectories of biomass gasification. I start by describing the relevance of this topic then explain the empirical and theoretical background in sections 1.2 and 1.3. Subsequently I present my research questions and defend my approach to answer these questions (section 1.4). And finally I outline the structure of this thesis by briefly discussing the upcoming chapters and their relevance (section 1.5).

9

1.1 Scope of this research Energy is omnipresent in our society: among other things, for producing goods and food, transporting people and cargo, heating and cooling buildings, and exchanging information. As such, energy use is closely linked to our economy and our well-being. Over the 20th century, developed economies increasingly started to use fossil fuels like natural gas, oil, and coal, which now dominate their energy portfolios (Niele, 2005; Lagendijk and Verbong, 2012; OECD/IEA, 2014; Smil, 1999). This dependency on fossil fuels has become a public and political concern, especially regarding the issues of depletion, geo-physical dependency, global warming and other environmental impacts (UNDP, 2004; WCED, 1987). To ensure the future supply of energy while simultaneously addressing these concerns requires the development and deployment of new technological options based on renewables. Worldwide initiatives are taking up the challenge. Many technologies have been developed and start to find application (REN21, 2014). However, market shares often lag behind targets and the general expectation is that society will have to rely on fossil fuels for at least some more decades (Verbong and Loorbach, 2012; World Energy Council, 2013). A faster uptake of renewables is required to avoid the worst effects of climate change (IEA, 2015). We need knowledge, not only about individual technological options but also on the general process which will enable new technological options to emerge and develop. Innovation scholars study the development of technologies to better understand the innovation process and determine the factors that influence successful development and diffusion. These scholars argue that technological developments are highly selective: very few of all the possible technological options end up being selected and explored. Such developments are often improvement sequences of a technology, characterized by incremental innovation and a strong focus on specific technologies to overcome specific problems. Engineers in a technological field share routines and therefore work in more or less the same direction. This leads to technological continuity–also called technological trajectories. Discontinuities are associated with radical innovations that happen less frequently (Dosi, 1982; Nelson and Winter, 1982). Other scholars maintain that this is a too narrow and instrumental perspective on technological development, one that puts too much emphasis on the role of engineers. They argue that agency by a variety of actors involved in the innovation process plays a crucial role, see Kemp, Rip and Schot (2001), Garud and Karnøe (2001) and Geels (2002). To emphasize this change in perspective, they focus on path creation, the process enabling new technological trajectories to emerge. Various methodologies are used for studying path dependency and path creation. In my view, in order to better understand emerging technological trajectories, we need

10

to analyze the factors that lead to path creation as well as the factors that contribute to further development of technological trajectories. One of these factors that I explicitly take into account is interrelated development: some technological developments build on earlier technologies, while others provide synergies or competition that explain the emergence or discontinuation of technological trajectories. I will analyze technological developments in detail by studying the case of biomass gasification. Biomass is currently the most applied renewable and its use may further increase for biofuels and a bio-based economy, which in recent years have attracted significant attention (Asveld et al., 2011; EuropaBio, 2008-2011; Faaij, 2007; Hoogwijk, 2004; OECD/IEA, 2014; World Energy Council, 2013). Biomass gasification is a conversion technology that facilitates large-scale high-end applications. It too has attracted attention over the past four decades, but with shifts in foreseen application and technology (Knoef, 2005a). The technology seems to have developed in interaction with biomass combustion and coal gasification (Heerman, Schwager and Whiting, 2001). This enables me to study the early stages of technological development as well as the interaction between technologies. By taking a longitudinal perspective, I can study product sequences to reveal upcoming trajectories as well as unfolding innovation processes.

1.2 Biomass gasification: a promising technology Biomass can be converted to energy by a variety of conversion processes as shown in figure 1.1. Which process is preferred depends on the feedstock type and the final energy service to be delivered (heat, power or fuels). Biomass gasification is a thermochemical process: the biomass breaks down under high temperatures (and sometimes high pressure) into a producer gas or syngas mainly consisting of hydrogen (H2) and carbon monoxide (CO). This gas can be used in various ways such as small-scale power production by engine-generator sets and larger-scale advanced applications like biofuel production. Gasification has a long history (Harmsen, 2000; Knoef, 2005a). It was used for production of town gas in the 19th and 20th centuries. Coal gasification was explored in situations of oil shortages (Germany in World War II, South Africa facing an international oil boycott under apartheid) and in response to the oil crises of the 1970s. Several basic designs of gasifiers exist that serve specific applications, see textbox 1.1 for a short introduction and appendix A for a more detailed overview.

11

Figure 1.1 Conversion routes of energy-from-biomass (based on Turkenburg, 2000). Thermochemical routes operate under high temperatures (and pressure) and perform best using dry lingo-cellulosic feedstock (wood, straw, etc.). Biochemical conversion processes

depend on enzymes and bacteria that require modest temperatures and wet conditions. In general, manure and agricultural crops are the preferred feedstock. Extraction requires oil-

containing seeds.

Textbox 1.1 The characteristics of gasification technologies

• Fixed bed technology: a fixed bed of feedstock is gasified using a gasification medium, generally air at low velocity. The main subtypes are downdraft and updraft gasifiers, which are usually applied on a smaller scale.

• Fluid bed technology: a small fraction of feedstock is added to a much larger fraction of bed material, which is fluidized by a gasification medium (air, oxygen, steam) that flows through the bed at a sufficiently high speed. The main subtypes are the bubbling and circulating fluidized bed, usually applied for biomass on a medium scale.

• Entrained flow gasification: small droplets or particles of feedstock are ‘entrained’ in a flow of gasifying medium – generally oxygen or steam. Also referred to as suspension flow or dust cloud gasifiers and applied on a larger scale for coal and petroleum based feedstock.

Combustion Gasification Pyrolysis

Liquefaction HTU

Digestion Fermentation Extraction (oil seeds)

T h e r m o c h e m i c a l c o n v e r s i o n B i o c h e m i c a l conversion

H e a t E l e c t r i c i t y F u e l s

Steam

Steam turbine

Gas turbine Combined cycle

Engine

Methanol / hydrocarbon /

hydrogen synthesis

Gas Gas Oil Charcoal Biogas

Upgrading Distillation Esterification

Ethanol Biodiesel

Gas engine

Diesel Fuel cell

12

A biomass gasification plant has several process units, as illustrated in figure 1.2. The biomass is stored and sometimes pre-treated (dried, cut into pieces), then fed into the gasifier that converts the biomass into producer gas. This gas often needs cleaning and conditioning, after which the clean gas can be converted into end products (power, heat, fuels). For such a configurational technology, the process units are interdependent: a gasifier that produces relatively clean gas will require limited gas cleaning, whereas a cheap and rigorous gas cleaning unit offers the opportunity to consider gasifiers that produce dirty gas but are cheaper or less complex. Also, if one of these process units faces persistent problems, it might delay overall development.

Figure 1.2 Biomass gasification is a configurational technology. At a biomass gasification plant, the biomass has to go through several process steps to produce the end product. Each step operates under specific input and output conditions with a certain efficiency, and might

face particular problems. Thereby each step influences the other process steps and overall performance.

1.3 Theoretical background The concept of technological trajectories is based on an evolutionary perspective of technological development. Evolutionary economics addresses the role of technological innovation by looking at economic and technological change. This type of economics is evolutionary in that it draws on the concepts of variation and selection, explaining both the diversity of technologies as well as focused or directed development. The development of technologies shows an internal logic that is described by the concept of technological trajectories (Dosi, 1982; Nelson and Winter, 1982). It is the result of path dependency (history matters, lock-in) and technological

biomass growing

harvesting transport

storage pre-treatment

transport

biomass feeding

gasifier producing

syngas gas cleaning

conversion syngas to

end-product

local use or in market

13

paradigms (restricted search spaces for solutions). Nelson and Winter (1982) consider the directed development of technology in the variation phase a sort of coordination within a population of firms in an industry or sector. Because of shared routines, engineers in a technological field work in more or less the same direction. Hence, “sequences of minor variations … add up to global technological trajectories that proceed in particular directions” (Geels and Schot, 2010, p.37). Dosi (1982) comes to a similar conclusion with the concept of technological paradigms. Dosi sees continuity in technological development, or development that adds up to a technological trajectory, as a pattern of normal problem-solving within a technological paradigm, while discontinuities are associated with the emergence of a new paradigm. Looking at technological trajectories is not equivalent to focusing on a single technology. On the contrary, evolutionary theory assumes that firms will be using different technologies, characterized by different suitability expressed in profits and performance. Only over time will this result in dominance, as market selection singles out the best performing technology and firm. One of the most famous examples of a technological trajectory is Moore’s law–the steady increase in the speed of microprocessors (Dosi, Orsenigo and Sylos-Labini, 2005). Other examples include the exploitation of latent economies of scale in chemical process industries and power generation or the improvements to the Douglas DC-3 over subsequent designs, resulting in a faster plane with a longer range and more comfort (Nelson and Winter, 1982, p258/259). In the case of the airplane, a technological trajectory is conceptualized as a development over multiple technological and economic dimensions, with progress showing “the movement of multi-dimensional trade-offs among the technological variables which the paradigm defines as relevant” (Dosi, 1982, p154). Scholars in this field often make use of quantitative approaches and models in order to recognize patterns in larger datasets. We see a recent operationalization of the concept of technological trajectories in Verspagen (2007). He identifies a dominant path in a patent-citation network, which is a series of subsequent patents related by citations. Other scholars argue that Dosi, Nelson and Winter put too much emphasis on embedding routines in the minds of engineers, see Kemp, Rip and Schot (2001), Garud and Karnøe (2001), and Geels (2002). Researchers and RD&D departments do take into account what they consider promising technologies based on in-lab performance or merely expectations, as well as the perceived future socio-economic context in which the technology will have to perform. According to these theories, the way a technology develops is (partially) a social construct. This quasi-evolutionary perspective builds on and incorporates sociological and historical theories of

14

innovation. It focusses more on stakeholders–for goals, problem agendas and expectations. Another literature stream, called ‘sociology of expectations’, deals more extensively with how expectations influence innovation processes–that is to say their performative nature. Scholars in these fields often make use of qualitative approaches such as interviews, literature study, and case studies, which allow them to study a broader variety of explanatory factors and a more contextualized development. Applying a quasi-evolutionary perspective to the concept of technological trajectories, Geels and Raven (2006) conclude that shared cognitive rules and expectations create stable trajectories of technological development, while changes in the technological trajectories’ direction depend on changes in the content of cognitive rules and expectations. The emerging technological trajectory can be found at the level of communities of practice or emerging fields. To summarize, the evolutionary perspective emphasizes path dependency, while the quasi-evolutionary perspective has a strong focus on the agency and shared expectations that result in path creation. I argue, that somewhere both perspectives come together: that an emerging trajectory starts to be subjected to self-reinforcing mechanisms and to show lock-in. To better understand this process, I will reflect on both approaches. 1.4 Research questions and methodology The aim of this thesis is to understand the development of biomass gasification and the dynamics of this process. The research questions I pose are:

1. a) How did the technology of biomass gasification develop? b) How can we explain these developments?

2. a) Did this result in emerging technological trajectories? b) What are the contributions from a path creation and path dependency perspective to the reconstruction of these emerging trajectories?

In this study I draw on the economic theories of innovation (path dependency, as in Verspagen, 2007) as well as the sociological and historical theories of innovation (path creation, as in Geels and Raven, 2006) to compare both approaches. To reconstruct the technological trajectory of biomass gasifiers, I highlight the notions underlying technological trajectories: dominant design, incremental innovation, product sequences, shared expectations and/or routines, and self-reinforcing processes that contribute to lock-in.

15

I will apply a mixed-method approach, drawing on the strengths of both qualitative and quantitative methodologies. The methodological advantages are that it allows for cross-examination and complementarity but also might reveal differences and lead to contestation that can trigger a critical reflection on theories and methodologies, see Johnson and Onwuegbuzie (2004), Johnson, Onwuegbuzie and Turner (2007) and Sandelowski (2013). Technologies typically develop over different levels of aggregation, ranging from technological fields to specific technologies. Technology studies often focus on one specific level. However I argue that an integrated study over these different levels can bring specific benefits. I see a parallel with approaches in related fields: van Lente (1993) who introduced micro, meso and macro levels to study dynamics in expectations; and the Multi-Level Perspective (MLP) to understand dynamics in socio-technical development and transitions (see Geels and Schot (2010) for more details). Depending on the specific research question and approach, the differentiation between levels can bring several advantages: it can add detail to the analysis; each of the levels might come with specific characteristics and dynamics that can be addressed; and the interaction between the levels can add to the understanding of the overall dynamics. I explicitly conceptualize different levels (see figure 1.3). General technology categories are depicted on the left. Towards the right, technologies become more specific and are somewhat similar: they are likely to share a technology and knowledge base, involve the same actors (scholars, companies), and share networks and institutions, also called technological, social and institutional proximity. This is relevant for shared learning and bridging between different technological paths (Sandén, 2004). To reconstruct the technological developments of biomass gasification in detail, I draw on science and technology indicators, patents, and the literature. The exposure of developments in biomass gasification in academic journals has been limited, especially early developments in the 1970s and 1980s. However, a great deal has been documented in conference proceedings, revealing an inside-out perspective. I add to that an outside-in perspective emphasizing the performance of biomass gasification compared to other (renewable) energy technologies. I traced this perspective by studying positioning papers on biomass gasification (conference summaries, overview articles), broader literature and statistics on renewables, and specific literature to follow up specific leads such as what was the expectation for methanol fuel in the late 1970s?

16

Figure 1.3 The development of a specific gasifier technology is influenced by drivers and by

synergies and competition at different levels Reconstructing a technological trajectory requires a longitudinal study to capture multiple product-sequences. The somewhat similar technologies of fluidized bed combustion and coal gasifiers developed over decades (Watson, 1997; Harmsen, 2000). Therefore I will reconstruct developments from the mid-1970s onwards, at the start of upcoming interest in response to the oil crisis. This is an important contribution because hardly any longitudinal studies have been performed in the field of biomass gasification. I know of two: firstly Negro (2007) who studied developments of energy from biomass in the Netherlands, including biomass gasification; secondly Hellsmark (2010) who studied biomass gasifiers in four main European countries over the past decade. Both scholars take a national perspective, as does a significant body of literature for IEA Biomass in so called country reports on the status and developments for biomass gasifiers within specific time periods. However, I argue that the technological development has been carried by a transnational field and as such should be studied from a global perspective: research and development are supported by international conferences; the manufacturers involved typically serve an international market. Global coverage is ensured with an extensive literature study focusing on global developments, as well as by studying science and technology indicators covering international developments.

Renewables Fossil energy

Nuclear energy Biomass Wind Sun Etc.

Thermochemical conversion Biochemical conversion

Gasification Pyrolysis

Combustion Fluidized bed Entrained flow

Fixed bed

Also for fossil fuels

Context specific: Feedstock x, technology y, product z, by manufacturer a in country b at time c

17

1.5 Structure of this thesis In chapter 21 I trace the shifting focus on different energy-from-biomass technologies and within renewables. This allows me to compare developments in biomass gasification with those in the closely related technologies of biomass combustion and biomass pyrolysis, thereby reconstructing a timeline that includes key drivers. To this end, I apply a discourse analysis for European RD&D on energy from biomass. My main source is the European Biomass Conference, the only leading conference in the field of energy-from-biomass held over the full time-span of this study. I consulted additional literature for cross-examination and to place developments in energy-from-biomass technologies in the context of other energy technologies. In chapter 3 I reconstruct and contrast global developments in coal and biomass gasification, both for small-scale and large-scale gasifiers. My aim is to identify the dominant technologies and applications, manufacturers and countries, the research and development dynamics, and to ascertain whether the developments in coal and biomass gasification are closely linked. This chapter is based on a literature overview and trends in science and technology indicators. I conclude that advanced biomass gasifiers received strong transnational support, unlike small-scale fixed-bed gasifiers, which are not well represented by patents and scientific articles. Thus in chapter 4 I focus on advanced biomass gasifiers, reconstructing their technological trajectory using the Geels and Raven (2006) conceptualization of trajectories emerging from shared expectations. I consulted a broad body of literature and constructed an overview of demo plants. In chapter 5 I study the dynamics in expectations in more detail. I examine three dominant developments: biomass gasifiers for methanol production in the late 1970s and early 1980s, Integrated Gasification Combined Cycles (IGCC) for high-efficiency power production in the 1990s, and biofuel production after 2003. The main focus is the rationale of these promises, the actor groups involved, and the dynamics in the field, as the underlying self-reinforcing mechanisms contributing to temporary structuring and path creation. I studied the literature as well as science and technology indicators. In chapter 6 I examine the technological development of advanced biomass gasifiers from a path dependency perspective using Verspagen’s (2007) patent-citation approach. Studying the patent-citation networks related to fluid bed gasification

1 Chapters 2 to 5 have been published in peer reviewed scientific journals.

18

enabled me to identify trends in patents over time (real-time perspective) and the interrelatedness of developments. By linking these findings to the top paths representing technological trajectories, I show how different developments contribute to or compete with these top paths. Finally, in chapter 7 I draw conclusions, answer the research questions, and reflect on the approach taken.

19

20

Chapter 2

Discursive shifts in energy from biomass: a 30 year European overview2

2 Kirkels, A.F. Discursive shifts in energy from biomass: A 30 year European Overview. Renewable and Sustainable Energy Reviews 16 (2012) 4105-4115.

21

2.1 Introduction Over the past decades energy from biomass has been on the forefront of promises and developments in renewable energy. Its large potential and flexibility regarding feedstock, conversion technologies and end-products certainly contributed to that. As such, it has been widely reported upon. But, surprisingly, this strand of the literature has been mainly focusing on either specific technologies or national developments (e.g. see Kaltschmitt, Rösch and Dinkelbach, 1998; Kwant and Knoef, 2004). One might argue that this reflects the strong role of national policies and developments that have proven to be of major importance (Faaij, 2006b). However, this neglects the international effort, e.g. the EU efforts on biomass Integrated Gasification Combined Cycles in the 1990s, and more recently the EU directive on biofuels. This focus is especially surprising as the general promises on energy from biomass are truly international ones: fuels from biomass as response to the international oil crises of the 1970s; energy from biomass to reduce greenhouse gas emissions since the 1990s; and more recently the promise of a bio-based economy. To live up to each of these promises requires international and large-scale application, as the burden of oil dependency is not levitated by a small local project, and the same hold with respect to reducing greenhouse gasses. The involved technologies ultimately serve a global market. Our argument is not that local or national developments are not relevant; our argument is that the supranational level does matter. Our focus is on Western Europe, a region in which there has been a broad and intensive continuous research, development and demonstration (RD&D) effort on energy from biomass over a long period. This focus is given in by practical reasons: availability of the literature and the need to reconstruct a comprehensive story line. However, we are aware of the importance of other regions: the USA and Canada that played a leading role in the 1970s and early 1980s (see e.g. Bio-energy Council, 1980; Klass, 1985, 1987 ); and more recently the developments in the USA and Asia and more specific China (see e.g. Office of the Biomass Programme, 2009; Wu, Yin, Yuan, Zhou and Zhuang, 2010). Although limited, several scholars did take on the international perspective on developments in Western Europe. Hall (1982) is an editorial that describes to upcoming of interest in Europe in energy from biomass. His later papers (Hall and House, 1995; Hall, 1997) both describe the status at that time and ongoing developments in order to assess future potential of energy from biomass in Western Europe. In his well-known paper, Faaij (2006b) follows a somewhat similar approach covering the 1995-2005 period. We will take a somewhat different perspective, by performing a discourse analysis over a longer time horizon, 1980-2010. Our main interests and contributions are (a)

22

how attention for specific promises and practices have developed over time and (b) the structure of the debate. As such it shows how the scenery in the energy from biomass community shifts with respect to applications and socio-economic and political context, but also with respect to feedstock and conversion technologies considered - see textbox 2.1. This can lead to deepening of the understanding of the ongoing developments, which is of support for policy in this field. 1980 is chosen as starting point, as it is still well covered by the literature and reflects the period in which the concerns on oil supply resulted in a renewed interest in energy from biomass that has been continued ever since. After discussing the followed methodology, we will introduce the European Biomass Conference, the main source that we will be relying on. After that we will represent the discourses as reconstructed by two different approaches. Finally, we will discuss the results and draw conclusions.

Textbox 2.1 Conversion routes energy from biomass (as presented by Turkenburg, 2000). Thermochemical routes operate under high temperature (and pressure) and perform best on

dry lingo-cellulosic feedstock (wood, straw, etc.). Biochemical conversion processes are depending on the use of enzymes and bacteria that require modest temperatures and wet

conditions. In general, manure and agricultural crops are the preferred feedstock. Extraction requires oil containing seeds.

Combustion Gasification Pyrolysis

Liquifaction HTU

Digestion Fermentation Extraction (oil seeds)

Thermochemica l conv ers ion B iochemica l conversion

H e a t E l e c t r i c i t y F u e l s

Steam

Steam turbine

Gas turbine Combined

cycle E i

Methanol / hydrocarbon /

hydrogen synthesis

Gas Gas Oil Charcoal Biogas

Upgrading Distillation Esterification

Ethanol Biodiesel

Gas engine

Diesel Fuel cell

23

2.2 Methodology 2.2.1 Discourse analysis Discourse analysis finds wide application in the social sciences. Object of analysis is the discussion and the argumentative structure of the discussion. This is based on the ideas of Foucault. By discourse Foucault meant “a group of statements which provide a language for talking about - a way of representing the knowledge about - a particular topic at a particular historical moment” (Hall, 2001, p72). This implies a degree of repetition, in the sense that the discourse - or discourses for that matter - are shared by groups of people and show some stability over time. Several scholars emphasize that it is not a pure ‘linguistic’ concept, but that it is about ‘language and practice’, what you say and do, ‘identifiable set of practices’ (Hajer and Versteeg, 2005; Hall, 2001). But as Fairclough (2003, p124) arguments, “discourses not only represent the world as it is (or rather is seen to be), they are also projective, imaginaries, representing possible worlds which are different from the actual world, and tied in to projects to change the world in particular directions”. This only adds to the importance of understanding the discourse. Discourse analysis in the field of energy from biomass seems to have drawn significant attention recently: Zschache, von Cramon-Taubadel and Theuvsen (2010) have been studying the mass media discourses in leading German newspapers; Sengers, Raven and Van Venrooij (2010) followed a similar strategy, but focused on the biofuels debate in the Netherlands; Lehrer (2010) also studied biofuels, but she focused on the US and takes a policy perspective; Huttunen (2009) studied the discourses in rural non-wood bioenergy production in Finland; and ly Asveld, Est and Stemerding (2011) focused on the debate on a bio-economy in the Netherlands. Two underlying characteristics that fuelled this attention seem to be that bioenergy production has recently been widely debated worldwide and that the issue is ‘multilayered and characterized by a great variety of competing interests, opinions and perceptions’, as Zschache et al. (2010) phrases it. Most of the discourse analysis above focus on the structure of the public or policy debate and as such involve the identification of actors and advocacy coalitions involved and their position in the debate. We will take a route less travelled by, studying the discourse in the RD&D community and its involved policy. This way we hope to highlight shifts in arguments and policy and how they relate to preferences for specific feedstock, conversion technology and application - a mix of policy and technology discourse. As such, the actors will remain more or less invisible - although one can image that a focus on agriculture crops relates to farmers; wood relates to forestry and pulp and paper industry; a focus on thermochemical conversion will require process technology, for example by boiler manufactures; a new, innovative, high-tech application will require support by RD&D institutions; etc.

24

2.2.2 Empirical approach To reconstruct the European discourse on energy from biomass is not a straightforward thing to do. For starters, there is a lack of overview articles, as was already discussed. Also the academic literature on energy from biomass is limited prior to 1991 - at least as covered by Web of Science (Thomson Reuters, 2014). This does not mean that there was limited activity on this field: proceedings of conferences and policy meetings suggest the opposite. Some of the more prominent were the international Bioenergy conference of 1980 and 1984; the IEA conferences on thermochemical biomass conversion; the VTT conferences on power production from biomass in the 1990s; and last but not least the European Biomass Conference. The European Biomass Conference is a large international conference mainly supported by the European Commission. It has been held since 1980, mostly annually or bi-annually. It encompasses the whole field of energy from biomass, from biomass production and harvesting, to processing and conversion, to final application, market formation, environmental impacts and policy. From the start it did not only include research and policy contributions, but tried to involve industry as well. The European Biomass Conference seems to be a good starting point for reconstructing the discourse. First, the conference covers the complete period. Second, it contains a significant number of overview papers, both with respect to technology development and with respect to EU policy. And third, due to its open character and large scale, it can be seen as one of the larger European platforms for people working in RD&D on energy from biomass or policy areas related to that and as such as a representation of the energy from biomass community. For reconstructing the discourse, we followed a two path approach. First of all, we studied conference introductions, summaries, positioning papers and overview papers over the period 1980-2010 in order to reconstruct a storyline. This storyline represents the trends as seen by leading experts and by European policy makers. For cross-examination and to reduce single source bias, we also studied some of the overview papers as presented on other conferences, as well as open literature and data sources - for example with respect to the production of energy from biomass and RD&D budgets. This led to refinement and better contextualization, and as such has been included in the storyline of the first approach. The results of this approach are described in section 2.4. Second, we followed a quantitative approach by doing a text analysis on all contributions to the conference, the results of which will be discussed in section 2.5. The text analysis was performed using the T-lab software (see Lancia 2010a, 2010b – on which the description below is also based). T-lab is a linguistic and statistical tool

25

for text analysis. T-lab offers multiple functionalities based on counts of occurrence and co-occurrence of words in a body of text that is fed into the software. The researcher can label the different text inputs (e.g. by year of publication, source, technology, etcetera). This allows to analyse subsets of text as well as to perform comparative analysis. Strictly speaking, this sort of software tools support content analysis. However, we studied developments over time that we contextualized. Many scholars (e.g. see Hardy, Harley and Philips, 2004; Neuendorf, 2004) believe that both methodologies - content analysis and discourse analysis - overlap and can be combined, as also Sengers et al. (2010) have shown in their recent paper on the media discourse of biofuels. The quantitative text analysis that we performed covers over 7000 titles, which are all titles of contributions to the European Biomass Conference over the period 1980-2010. As such it is a very powerful tool and an indication of the discourse of the energy-from-biomass community as present at the conferences. For the analysis we differentiated to five periods that are made explicit in table 2.1. Each period is covering about five years and is more or less in line with typical time periods as distinguished by the first approach. This way it became easier to analyze developments. And as each period covers 3-4 conferences, it reduces the overrepresentation of the organizing country - that would bias a year-to-year comparison. We used T-lab under standard settings. In this modus T-lab makes an automatic selection of key words to be included in the analysis. For this selection T-lab first performs lemmatization: the reduction of the set of words to their respective headwords (i.e. lemmas) that define a set of words with the same lexical root. It entails that verb forms are taken back to the base form, nouns to the singular form, and so on. A second selection that T-lab performs is the exclusion of stop words. Stop words are words that don’t have any specific and/or significant content: e.g. ‘all’, ‘the’, ‘do’, ‘very’, ‘before’, ‘that’, ‘they’, ‘what’, etc. We used T-lab’s specificity analysis to compare the occurrence of lemmas per restricted time period (the subsets of about 5 years) to the occurrence over the whole period (30 years) (so called part/whole analysis of typical lexical units). T-lab does so by calculating Chi-square values and the associated probability. This checks whether the frequency values of lemmas in the restricted time periods are significantly different from the theoretical ones – based on the occurrence of these lemmas over the whole period. This provided a profile of over-used and under-used lemmas over different periods. Only lemmas are included for which the probability that the frequency distribution can be explained by the distribution in the complete set is smaller than 5%. The output table needed further interpretation, as it was quite long: 109-151 lemmas of over-used words and a bit less for under-used words. It also still

26

contained general and not very relevant terms, like process, production, conversion, unit, move, etc. We made a selection of most relevant lemmas (indicative of developments, meaningful) and categorized them. Over the years the total number of contributions to the conference tripled. As a consequence, one would expect a wider variety of topics and terminology, resulting in longer lists of over-used words. However, the opposite was true: the number of over-used words decreased over time (respectively 151-188-120-109-151 words). Apparently over time some alignment of research topics and terminology has been taken place. The interpretation required some extra care: lemmas that were completely not used in a period are not represented, nor are lemmas that are dominant over the 30 years but do not stand out in a specific period. Also, results show the time dependency of use of synonyms. In the 1980s ethanol was produced, after 2000 bioethanol. This was part of a much broader tendency in naming everything ‘bio’, including words like biodiesel, biofuels, bioenergy, etc. T-lab does allow correcting for that, by building a dictionary in which synonyms are labeled with the same lemma. But this takes a significant effort and presents new dilemmas: bioethanol and ethanol can be considered synonyms, but should we also cluster it thematically with fermentation; and CFB, referring to circulating fluidized beds, might sometimes be synonym for fluidized bed, while in other cases it is a distinctive subset. Therefore we used the software under standard settings without defining a dictionary, checking on relevant terms using additional test where required. This approach is justified as it is used to cross-examine the literature-based-reconstruction of the discourse and any anomalies will stand out and can be checked upon. To come to the storyline of the discourse, we used the lemmas identified by T-lab and checked upon their use in titles. This provided details of the context they were used in. In addition, we checked some lemmas on co-occurrence - for example to check to what lemmas ‘market’ was related to over the period 2007-2010. 2.3 European Biomass Conference We already introduced shortly the European Biomass Conference, as a large international conference covering the whole field of energy from biomass. The conferences have been held annually or bi-annually, mainly by the support of the European Commission and the organizing countries. The general focus has been on the European scene. Both the 2000 conference in Sevilla and the 2004 conference were named ‘World conference’. However, the number of international contributions did not differ from later conferences. In addition, both conferences were held in and

27

were dominated by Europe. As such, we will treat them just like the other EU conferences - as also the organization did: the proceedings of both World conferences were styled the same as the EU conferences, and in the numbering of subsequent EU conferences both World conferences are included. There remains the question which Europe is represented by the conference, as over the considered period the EU has been gradually expanding with mainly Scandinavian and Eastern European countries. And more importantly, are the dominant countries in energy from biomass represented well over time? First dominant European countries in the field were identified, both with respect to resource potential and application of bio-energy. Especially France and Germany have both large wood resources and agricultural output. Sweden and Finland also have large wood resources, while the latter also has large domestic peat resources. Exactly these four countries also excel in the use of energy from biomass (Beurskens and Hekkenberg, 2011; EurObserv’ER, 2008; Eurostat, 2009; FAO, 2006; Paappanen, Leinonen and Hillebrand, 2006). Next we checked for the contributions to the conference. In the 1980s France is dominating. From the smaller countries Belgium has a relatively high share. Contributions show a strong focus on EC. For the period 1989-1996 no detailed data are available. From 1998 onwards, Italy and Germany have been the main contributing countries. The importance of smaller countries increases, especially for 1998-2002: Sweden, the Netherlands, Austria and Denmark. In general, EU contributions continue to dominate. Since the 1990s there are some contributions by Eastern European countries, but at a limited scale. Overall, one can state that the conference does represent well EU and especially West European developments. Major misrepresentation is the lack of contributions by Sweden and Finland in the 1980s. These countries excel both in resources and application of bio-energy and both countries held large relevant RD&D programs in this period, focused on combustion, gasification and liquefaction of wood, peat and black liquor (Asplund, Sahrman and Solantausta, 1980; Kurkela, 1989; Palmberger, 1980; Pettersson, 1980; Sipilä, Kurkela and Solantausta, 1993/1994; Ström, Liinanki. Sjöström, Rensfelt, Waldheim and Blackadder, 1985). A first characterization of the success of the conference over the years is given in figure 2.1. Both contributions and participants showed a strong increase during the late 1990s. This was due to a change in focus of the conference, enlarging the circle of participants (Scheer, 1995), as well as a growth of interest in energy from biomass, especially over the 1994-2000 period, as we will see.

28

Figure 2.1 European biomass conference, number of contributions and participants 2.4 Literature study based discourse The early years, 1980-1985, are characterized by major concerns regarding high oil prices and attempts to decrease dependency on oil. Policy focuses on improving energy efficiency and alternative fuels, especially coal and nuclear (and later also natural gas) for power production. It is a period of increasing awareness on energy from biomass, in which a diversity of resources and conversion technologies is being explored - with highest priority for production of a liquid fuel for the gasoline market and displacing oil for heating purposes in rural areas. Given this diversity, neither the overall impact biomass can make nor its strategic importance is widely recognized. And although there is a need for short term application, there is also the recognition that developing the energy from biomass option will probably take a long time (Chartier, 1983; Chartier and Hall, 1981; Chartier and Palz, 1980; Clarke, Strub, Ghose and Berger, 1981; Strub, Chartier and Schleser, 1983b; Williams, 1981). Biomass per unit land turns out to be a sensitive cost factor and considerable effort is given to improve plant productivity. Combustion, digestion and fermentation are widely used technologies. Special attention is given to demonstration of farm scale biogas production by digesters - although at the end of the 1980s hopes are dashed. Combustion also receives significant attention, mainly the combustion of agricultural wastes (straw), industrial waste and RDF. Gasifiers are much less applied and mature. However, significant research focusses on production of transportation fuels, and methanol synthesis via gasification of wood is given main priority. Ethanol is initially not considered as a promising liquid fuel based on energy balance economics and competition for land. During the 1980s it finds new interest, for example as co-solvent in biofuel blends, and by cellulolysis for the breakdown of lignocellulose to cellulose and subsequent fermentation, which becomes an area of very active RD&D (Fabry

0

200

400

600

800

1000

1200

1400

1600

Contributions

Participants

29

and Ferrero, 1990; Fabry and Goudeau, 1987; Grassi and Pirrwitz, 1983; Hall, 1983; Strub, 1984, 1985). Lignocellulosic biomass (wood, straw, etc.) are considered most versatile, as world production is large, competition with food is limited and they can be stored and allow for steady state feeding (Chartier, 1983; Chartier and Hall, 1981). In the subsequent period, 1985-1987, oil prices stabilize at lower prices and no longer provide the sense of urgency that characterized the early 1980s. The context shifts: on the one hand to dealing with waste and residues - a low cost feedstock that is topic of increasing environmental concern; and on the other hand to dealing with agricultural surpluses in the EU, with energy crops giving a possibility to use set aside land and provide additional income, and as such support rural areas and communities. This results in an increase in attention for short rotation coppice. Digestion and combustion receive even more attention in demonstration projects. With respect to liquid fuels attention shifts to ethanol production by fermentation of agricultural crops: it is considered the most suited biofuel when agricultural lands are available and receives strong support by sugar beet farmers and sugar industries - that try to deal with a depressed sugar world market in the early 1980s (Chartier, 1983; Fabry and Ferrero, 1990; Fabry and Goudeau, 1987; Grassi, Delmon, Molle and Zibetta, 1987b; Hall and Coombs, 1985; Hall and Palz, 1985; Strub, Chartier and Scheser, 1983b). Over 1989-1993, government expenditures on energy from biomass RD&D are at a low point (OECD/IEA, 2011) - according to Grassi (1992) budgets in the European biomass program over 1989-1991 drop by 40%. This is reflected in a limited number of conference contributions and visitors. In 1991 for the first time researchers and experts of Eastern Europe participate at the conference (Chartier, 1992; Grassi, Collina and Zibetta, 1992b). The Gulf crisis is a reminder of the dependency of oil - but the impact on oil prices and energy policies remains limited (Chartier, 1992). The European network has achieved a high level of scientific competence, for example in enzyme hydrolysis of cellulose and short rotation coppicing. The industrial applications of these breakthroughs are yet to emerge, with limited interest from the energy market due to the low oil prices (Chartier, 1990). Apart from energy production from waste, development of biomass technologies is only justifiable if external costs are internalized. Emphasis is given to lignocellulosic biomass (wood, agricultural residues like straw, etc.) (Allgeier, Caratti and Sandberg, 1995; Auken, 1996; Chartier, 1994; Grassi, 1993). A major driver becomes the concern about environmental problems: limiting emissions, effects of intensive land use, dealing with (the volume of) waste, and the upcoming concern on global warming (Auken, 1996; Chartier, 1990, 1992, 1994; Grassi, 1988, 1990, 1992, 1993, 1994b; Grassi, Gosse and dos Santos, 1990b; Grassi et

30

al., 1992b; Hall, Grassi and Scheer, 1994b; Scheer, 1995). The Earth Summit of Rio gives impetus that sustainability will be of growing concern and results in EU commitments to stabilize CO2 emissions (Allgeier et al., 1995; Hall et al., 1994b). Another driver is rural redevelopment and support of the agricultural sector - including job opportunities. At that time it is expected that a further increase in competition from Eastern Europe and possibly from developing countries (due to the unavoidable reforms in agricultural policy) will result in more arable land set aside (Allgeier et al., 1995; Grassi, 1992, 1993, 1994b). There is a start to positively reframe the contribution of renewables to the energy supply. And energy from biomass is considered the option to deliver at the short and medium term, being one of the most cost-effective renewables that also is widely available and widely spread. This would finally be formalized in the 1997 White Paper on renewable energy, with a leading role for biomass. As such it is quoted and defended till well after 2000. Main focus is still on biomass as an indigenous resource, which requires large integrated pilot projects on a regional scale (Allgeier et al., 1995; Auken, 1996; Beenackers, 2001; Bridgwater, 1995; Chartier, 1994; European Commission, 1997; Grassi, 1988, 1992, 1994b; Hall, 1997; Hall and House, 1995; Maniatis, Guiu and Riesgo, 2003; Scheer, 1995; Schmidt, 1992). In the late 1980s and early 1990s emphasis is given to pyrolysis: its developments seem promising, with Europe leading world developments and the liquid produced has a strategic value: it has a high energy density; it is easy to transport; it is easy to apply in heating and power applications; and it can possibly be applied as fuel blend or for the production of high value chemicals (Bridgwater, 1990b; Bridgwater and Beenacker, 1990; Grassi, 1988, 1992, 1993, 1994). In the 1990s a new sector has come into focus, electricity production, with possible contributions coming from advanced technologies like fast pyrolysis and gasification using combined cycle generators. The power production sector has found itself under pressure, as after the Chernobyl accident there is little support for expanding nuclear power, and concerns regarding global warming limit the support for coal based power generation (Allgeier et al., 1995; Chartier, 1992, 1994; Grassi, 1992, 1993, 1994b). Energy from biomass gets wider acceptance (Chartier, 1992). As Grassi (1992) formulated it: “We are also witnessing greater agreement on the significance of this sector and the valuable contributions that it can make.” However, the period is one of transition and reassessing the role of energy from biomass, as reflected in the number of positioning papers, and the process shows ambivalence toward energy from biomass (Bettini, 1994; Colombo, 1994; Grassi, 1994a). As de Sampaio Nunes (1995) phrases it ‘Energy from biomass does not enjoy the same administrative and political support as other traditional forms of energy do’.

31

In the period 1994-1996 there is growing commitment to reduce greenhouse gas emissions. There is a growth in research, which shows in the budgets and number of contributions and visitors to the conference. The latter also increases due to broadening with more active participation from agricultural interest groups, government agencies and technical journals (OECD/IEA, 2011; Scheer, 1995). These come from increasingly more countries, with the EU expanding and with a growing international attention for energy from biomass (Chartier, Ferrero, Henius, Hultberg, Sachau, Wiinblad, 1996b). Parallel to the 1996 conference also the first technology exhibition was held. From this period on growing emphasis is given to bringing bioenergy to the market place, with more attention for commercial or near commercial demonstration activities, economy, environment, non-technical barriers and political means for accelerating the use of energy from biomass (Beenackers, 2001; Bridgwater, 1995; Chartier et al., 1996b; McCormick and Kaberger, 2007; Mitchell, Bridgwater, Stevens, Toft and Watters, 1995; Rösch and Kaltschmitt, 1999). Where the early 1980s focused on liquid fuel production, the 1990s are all about power production, both large scale (combustion and IGCC) as well as small-scale combined heat and power. In 2001 this becomes formalized by a directive on electricity from renewable energy sources. There is an increasing involvement of large power producers. Lignocellulosic biomass remains on top - with woody biomass already being applied at a large scale. The focus remains strongly on rural development (Auken, 1996; Chartier et al., 1996b; Millich, 2001; Palz, 2002). In 1998-2002 these trends are followed, with an increasing attention for climate change and the upcoming of biomass combustion and co-combustion and co-firing, the latter two involving both coal and biomass as feedstock. By that time, many countries have set ambitious targets for energy from biomass. However, there is growing concern that these targets will probably only be met in exceptional cases at best - or might be achieved later on or with a shift to other renewables. If the targets are to be achieved on the short term, more efforts are necessary to remove the (mainly non-technical) barriers for implementation (Beenackers, 2001; Rösch and Kaltschmitt, 1999; Spitzer, 2002). As the liberalization of the electricity markets progress, as a result of the EU policy on internal energy market that already started in the 1990s, the enthusiasm by power companies is tempered. Private power companies tend to have low RD&D budgets and be risk aversive - that is, if they invest at all they tend to go for proven technologies (Burdon, 2003). The foreseen expansion of the EU results in studies on the possible contributions or Poland and Ukraine, both large countries. In gasification, solving the persistent problems with tar becomes a focal point.

32

In the most recent period, 2004-2011, oil prices rise again, mainly from 2004 on with an all-time peak in 2008 (IEA, 2010; Wikipedia, 2011a). Although this was the result of multiple causes, some influencing factors are the increased use, especially by the upcoming economies like China, and a failure at the production side to keep up with demand. As these are considered structural changes, it is expected that on average oil prices will rise and become more volatile in the future (Hirst, 2007). However, this has not resulted in the strong sense of urgency that characterized the response to high oil prices in the late 1970s and early 1980s. More dramatically is the shift with respect to power application, where biomass has been surpassed by wind power, which became the new renewable energy technology of choice and the promise for the future. Biomass still holds a strong promise - that now focuses on application for heat, biofuels and in industrial applications (Beurskens and Hekkenberg, 2011; Eurostat, 2010, 2011; Jäger-Waldau and Ossenbrink, 2004). Over this period especially the international dimension comes up. Trade in biofuels takes off, both in pellets (solid biomass), but also in liquid fuels that start to dominate the agenda. In Europe the driver is set in the 2003 directive on liquid biofuels with specific targets for 2005 and 2010. Interest covers both first generation biofuels as well as the development and demonstration in second generation biofuels - although application remains restricted to first generation fuels (ethanol by fermentation and biodiesel) (Beurskens and Hekkenberg, 2011; Eurostat, 2011; Fjällström, 2004; Hamelinck and Bain, 2003; Kwant, 2004; Spitzer, 2002; Swaaij, Prins and Kersten, 2004b). The trade in biofuels relates mainly to countries like Canada, Brazil, Malaysia and Indonesia. But the scene is growing, with China and Russia also showing interest (Grassi and Utria, 2004; Maniatis, Grimm, Helm and Grassi, 2007b; Sjunneson, Carrasco, Helm and Grassi, 2005b). This results in new attention on strategies, policies, logistics and sustainability criteria, and interest in intermediate products (wood chips, pellets, pyrolysis oil) (Dallemand, 2009; Swaaij et al.,2004b). There is a growing interest in biorefineries and polygeneration. Bio-refineries may become the basis for cooperation amongst the various biomass sectors, to develop their synergies. There is also renewed attention for heat applications, the ‘Cinderella’ of bio-energy applications, and advanced applications like synthetic natural gas (SNG) and fuel cell applications (Carrasco, 2005; ETA/WIP, 2007; Hirst, 2007; Kwant, 2004; Maniatis, 2007; Pease, 2007). There seems to be a two path strategy to implement the utilization of biomass: the first one is integration in the existing infrastructure (power, blending biofuels, synthetic natural gas for existing gas networks); while the other one focuses on implementation at the best place on earth, including Clean Development Mechanisms and Joint Implementation and biomass trade (Kwant, 2004). There is some renewed sensing that biomass technologies are mature and that the interest for biomass commercial implementation has increased significantly, which could possibly result in

33

the initiation of the biomass deployment. However, there still is a need for strong policy, huge investments and strong support of other industries, especially the oil and car industries and the utilities (Carrasco, 2005; Hirst, 2007; Maniatis, 2007; Palz, 2007). Not included in this overview, but omnipresent over the years is the intensive and broad attention for biomass growing / productivity / harvesting. Also present from the start, although on a more limited scale, is the recognition of the importance of (traditional) biomass use for energy in developing countries and the possibilities that come along with it. This is very well articulated in the early 1980s and remains present over the years. 2.5 Quantitative analysis discourse As described in section 2.2.2, we used T-lab software to create a table of over- and under-used words in conference contributions over different periods, see table 1. To keep the results well-ordered, they are classified in the categories ‘regions’, ‘feedstock’, ‘conversion’, ‘application’ and ‘others. The results have been contextualized and will be discussed in thematic order, following the rows in table 2.1. Countries are selectively represented, focusing on most contributing countries. The early 1980s were dominated by the contributions of the UK and France. In the 1990s a series of smaller countries came up: Finland, Denmark, Austria and Portugal. For the period after 2000 the appearance of both Poland and Ukraine is striking. Both countries have been considered for their potential of biomass production, by both agriculture and forestry. Striking is also the limited reference to Germany in this period - a country known for its leading position in the field in recent years. The contribution of feedstock can only be interpreted when being aware of its subsequent possibilities of conversion and application. In the early 1980s there is mainly attention for forestry and wood, the traditional biomass sources for energy. Algae are considered as long term response to an increase of the pressure over land use. In the late 1980s attention shifts and start to include short rotation coppice, including eucalyptus and poplar, and crops suited for fermentation to produce ethanol, including Sweet Sorghum and Jerusalem Artichokes. The 1990s show a further increase in emphasis on short rotation coppice. Just after 2000 there is a remarkable attention for fossil fuels and especially coal, which are used in all sorts of hybrid processes using both coal and biomass. The most recent period shows extra attention for pellets, which are furthest in becoming a biomass commodity with

34

international trade, as well as some specific waste streams like tires and sewage sludge. The 1980s start off with mainly attention for digestion and fermentation that produce respectively methane and ethanol, although also methanol production by gasification was considered. As already mentioned, in the late 1980s there is a shift towards fermentation and ethanol production. In the 1990s there is a complete shift in focus. Fermentation and digestion and their respective products become under represented, although co-fermentation and co-digestion do receive some attention. The new kinds on the block are the use in gas turbines (IGCC) after gasification, combustion and combined heat and power - all to deliver mainly power but also heat. As such both the large and the small systems receive significant attention. After 2000 emphasis is on cogeneration (combined heat and power) and on co-firing / co-combustion, involving coal. After 2007 attention seems to broaden a bit. Gasification and co-firing are still in the spotlight, but now also bio-refineries are considered - which might actually accommodate a gasifier. There is a lot of attention for biodiesel and bioethanol. But also other, more advanced applications are considered, mainly based on gasification, for example synthetic natural gas (SNG) and fuel cells (SOFC). There is also attention for bio-hydrogen, both in general and more specific on production via a fermentation route.

35

Table 2.1 Over and under used words in different time periods, as present in titles of the European biomass conference Period 80-85 87-91 92-98 00-05 07-10

Ove

r-us

ed w

ords

Regions UK, France, Sweden France, EC, Brazil Finland, Denmark, Austria, Portugal, EC

Europe, Greece, Sweden, Denmark, Poland, Germany, Ukraine, United States

EU, Italy, UK, Thailand, Africa

Feedstock / crop

Alga, waste, forestry, fuel wood, wood, willow, crop, straw, sugar

Coppice, tree, pulp, timbre, eucalyptus, poplar, rice husk, agro forest, Sweet Sorghum, Jerusalem artichoke, straw, cotton, refuse, charcoal, fungus

Short rotation coppice, salix, eucalyptus, Miscanthus, coal, fuel wood, forestry, chip, bark, pulp, paper, Sweet Sorghum, reed, agricultural, bagasse, MSW

Biomass, coal, fossil, fuel, Kenaf, corn, stover, Switchgrass, prune, sunflower, olive

SRF, SRC, pellet, Jatropha curcas, palm, fruit, sewage, tyre

Conversion Digestion, fermentation

Digestion, fermentation, liquefaction, hydrolysis, pyrolysis, refinery

IGCC, combustion, Stirling, co-fermentation, co-digestion, combined heat and power

CHP, cogeneration, gasification, co-firing, co-combustion, Rankine

Gasification, torrefaction, bio-refinery, co-firing, ORC, extrusion

Application Alcohol, methanol, ethanol, gasoline, gas, methane, engine

Alcohol, ethanol, methane Electricity, heat, RME, compost Bioenergy, gas Biofuels, bioenergy, SNG, biogas, bio-hydrogen, syngas, SOFC, biodiesel, bioethanol

Others Animal, farm LEBEN, fast growing, yeast, fluidization

Plantation, non-food, ecological, carbon dioxide

CFB, tar, olivine, small scale, transportation, demonstration, socio-economic, optimization, prospect, implementation, Kyoto, NOx, life cycle, building, supercritical, CFD

GHG, CDM, mitigate, impact, sustainable, LCA, emission, chain, supply, trade, logistic, transport, pre-treatment, market, consumer, policy

Und

er-u

sed

wor

ds

Regions EU EU, France Denmark, Brazil, Austria Feedstock / crop

Sorghum Miscanthus, pellet Switchgrass, sunflower Sweet Sorghum, alga, Miscanthus, artichoke, fuel wood, forest, Jatropha curcas, rape, bagasse, Eucalyptus, charcoal

Waste, Sweet Sorghum, wood, coppice, Miscanthus, forestry, agriculture, eucalyptus, husk, straw, charcoal

Conversion Combustion Combustion, gasification Gasifier, digestion Fermentation, flash pyrolysis

(co)-Fermentation, digestion, hydrolysis, liquefaction

Application Biofuels, bioenergy, power, oil

Biofuels, bioenergy, gas, power, heat, biodiesel, bioethanol

Biofuels, bioenergy, syngas, hydrogen, methane, methanol biodiesel, bioethanol

Alcohol, ethanol, methane, biogas,

Alcohol, electricity

Others Tar Emission, ash, tar, potential market, policy, strategy

Sustainable, trade, chain, tar

Sustainable, harvest, supply, grid Animal, pulp

36

In the category other lemmas we find the word animal and farm in the early 1980s, that link to the emphasis on digestion and fermentation. Subsequently we see lemmas related to short rotation coppice (fast growing, LEBEN - Large European Biomass Energy Network, plantation, non-food). From the 1990s onwards one can see the increase in attention for reduction of greenhouse gasses (carbon dioxide, Kyoto, greenhouse gasses GHG, clean development mechanism CDM, mitigate). Typical is the emphasis on gasification, including CFB technology and the tar problem, for the period 2000-2005. After being present since the 1990s, sustainability finally breaks through in recent years. Along with that there is more attention of life cycle analysis, emissions and chain impacts. These are also the years of market formation, as reflected in lemmas like chain, supply, trade, logistics, transport, market and consumer. This involves biomass and bioenergy in general and more specific biofuels and pellets 2.6 Discussion and conclusions 2.6.1 Story lines We performed a discourse analysis by two different methods, resulting in the two storylines presented above. Both show similar trends - see table 2.2. The applied methods go well together: the literature based approach allows for more details and contextualization (arguments, drivers, etc.), while the quantitative approach shows that it is not just a discussion, an analysis of linguistics, but it also relates to a community effort that is represented - a discourse of language and practice. Results also show the relevance of the supranational approach: clearly there is a shared agenda, vision and effort. The broad, long term and international focus allows identifying shift in topics, technologies and contributing countries - shifts that in studies that focus on a specific technology or country would not show up. For example, the shift in the 1990s was not just favoring gasification based IGCC - that received most of the attention -, as it turns out to be part of a much broader shift away from biochemical towards thermochemical conversion routes. When checking the contributions of the conference, by words in titles, contributions per session theme and scores in key words indices, it shows that this shift from biochemical to thermochemical conversion has been rather strong. It resulted in the fourth framework Joule-Thermie program (1994-1998).

37

Table 2.2 Summary developments in energy from biomass RD&D and policy in Western Europe

1980-1985 Oil crisis shock, fuels for oil replacement like methanol by gasification Coming of age, growing interest Small impact near future, strategic importance long term Large government sponsored programs, need for long term support Short term waste, medium term energy crops Importance for Third World countries Biological routes

1985-1987 Low energy prices Dealing with waste, opportunity fuel Agricultural surpluses ethanol and short rotation forestry schemes Feedstock research and reduction of production costs

1989-1992 Idem Initially reduced interest, followed by positively reframing potential biomass Growing importance environmental problems, especially global warming International cooperation on biomass

1994-1996 Thermochemical conversion is maturing (combustion, pyrolysis, gasification) - and receives most interest

Electricity (incl. large power producers) and decentralized heat and power Unemployment disadavantaged regions Setting aside of farm land to limit excess agricultural production Bringing biomass energy to the market place, advantages and barriers

1998-2002 Worldwide strategies Bringing biomass energy to the market place, policy, climate protection (Advanced) combustion Coal (co-combustion, co-gasification)

2004-2011 International cooperation Biofuels trade, logistics, sustainability debate Bio-refineries, SNG and fuel cells

This shift from biochemical to thermochemical routes at the end of the 1980s / early 1990s coincides with a shift in contributing countries, away from the major contributions by France and the UK towards Germany, Italy and a series of smaller countries, including Finland and Sweden. Although Finland and Sweden had been active in this field already in the 1980s, these developments remained largely out of sight of the European biomass conference. To some extend the developments seem to be correlated. Bridgwater (1992) confirms that Europe in general, and especially France and the UK - together with Italy - took the lead in biochemical conversion in the late 1980s. However, Bridgwater also shows that the UK and France were very active in the field of thermochemical conversion of biomass. So apparently, other causes need to be considered as well. The case of the UK is not that clear, but for France we did find additional information. In the early 1980s the energy situation in France was characterized by high dependency on oil import and a high potential of biomass (Chartier, 1981; Durand,

38

1981). Small-scale air blown gasifiers found initial application, while France also actively participated in the EU gasification-to-methanol research (Beenackers and van Swaaij, 1985; Bridgwater and Beenackers, 1985). However, methanol production by this route could not compete with the low oil prices - that become a fact of life in the mid-1980s. In the EU in the 1990s, research was refocused to gasification for power generation (see the storyline and Bridgwater, 1995). However, for France this was not an attractive route to pursue, as it already shifted to nuclear power production in the 1980s on a large scale. Public funding of research on biomass gasification for power production was stopped in 1987 and France buy-back tariffs for power showed to be one of the worst in Europe (Boissonnet, Boudet, Seiler and Duplan, 2005; Edouard, 1986; Kaltschmitt et al., 1998). With the renewed interest in biofuels after 2000, France mainly focused on first generation fuels. 2.6.2 Duality The discourse on energy from biomass is surrounded by duality. For example, energy from biomass can be both culprit and savior for sustainable energy production. Surprisingly, already early on both the possible competition on land for food production and the energy balances received interest, as well as the impact of intensive land use (see e.g. Hall, 1982). In the 1990s attention intensified and focused on environmental problems and sustainability issues, including the discussion on sustainability criteria. Nevertheless, the biofuels directive of 2003 resulted in some malpractices and the public food versus fuel debate after the 2007-2008 world food prices crisis (Wikipedia, 2011b). This debate evolved around similar arguments that were already used by Hall in 1982. This raises the question how these arguments failed to make a contribution to the biofuel directive. Lehrer (2010, p440) describes a similar situation with respect to the US legislation over 2007-2008 with regard to biofuel development. “While concerns were raised about the environmental impacts of large-scale government investment in biofuels, these concerns gained political traction only after the 2007 (energy) and 2008 (farm bill) laws were passed.” She concludes that “policy is created at a particular moment in time and is influenced by situational, political and discursive forces that shape the type of policy change that becomes possible at that time”. She argues that discourse analysis can play an important role in identifying (and understanding) the opening and closing of these policy windows - the windows of opportunity for policy development in the given context. Another omnipresent duality is that of time horizons. On the one hand there is the focus on energy from biomass as a strategic option. This includes decreasing the dependency on oil and application to reduce greenhouse gasses. This focus also

39

relates to innovation: new, efficient and cheap technology that is a promise for the future; and highly productive crops and growing methods. As such it allows for long term RD&D trajectories that could easily encompass one or more decades. On the other hand there is a continuous pressure for short term application: to reduce the actual dependency on oil, to start reducing greenhouse gasses, to support rural areas and farmers, and to involve market parties and their sources and dynamics. The storylines clearly show the tension between the two approaches - long versus short term - that frequently require different policies, but also lead to different choices with respect to feedstock and technologies. 2.6.3 Discourses and discursive shifts Energy from biomass continues to be used as key word in potential studies, renewables and energy policy. However, as the story lines show, this is a non-homogeneous and ambiguous concept, a label that to some extent disguises the broad and complex options that it represents. It includes a wide variety of feedstock (wood, agricultural crops and residues, waste, algae), conversion processes (digestion, fermentation, combustion, pyrolysis, gasification, liquefaction, etc.) and applications (heat, power, fuels, chemicals, large vs. small scale), while also the drivers to work on these options change over time. Overall, one can say that energy from biomass referred to something completely different in the early 1980s, compared to the mid-1990s or to 2010. And it is this change in discourse that we are interested in. This variety in options provides on the one hand what innovation scientists call interpretive flexibility (Pinch and Bijker, 1984) - no matter what the new policy goal or large societal problems or promises are, energy from biomass can adapt to that like a chameleon and continues to receive some support. On the other hand, learning effects between options will be limited, each option will involve its own actors, and building up an image to the general public and media remains difficult, while other renewables like solar and wind seem to provide a much clearer profile. This is further eroded by the quarrels between different biomass conversion options (Maniatis, 2005; Palz, 2002; Pease, 2007; Rösch and Kaltschmitt, 1999). Several authors have structured these developments in one way or another. We will follow Chartier (1990) who distinguishes strategies based on knowledge intensity and scale, see figure 2.2. Each of the strategies also comes with different feedstock, conversion technologies and applications, and as such with different actors and supportive arguments. As these strategies fit well with our story lines, we will argue that they constitute different discourses that together have shaped the path of developments and as such can be used to analyze them.

40

Smal

l sc

ale

Knowledge intensive

Larg

e sc

ale

Biotechnology / bio-based economy • High tech / competitive • High added value • Medium to long term • Unconcerned farming land /

agricultural surpluses

• Limited attention lingo-cellulosic

Renewable energy (solid, liquid, power) • Ligno-cellulosic matter, wood,

short rotation coppice • Thermochemical conversion • Large volumes at low costs energy,

bulk chemistry • As such related to practices of

forestry, pulp & paper industry

• Conversion farm land to short rotation coppice

Small local systems • Farm based digestion • Chimney and stoves for space

heating

• Local / self sufficiency

Biofuels • Ethanol and vegetable oil,

wheat, beet, rapeseed • Short term • Maintain farm income,

stabilize agricultural population

• Compared to cost subsidized agriculture

Knowledge extensive

Figure 2.2 Four discourses in energy from biomass The first one relates to biofuels (bottom right) and is focused on large-scale ethanol and biodiesel production to use agricultural surpluses and support rural development. Research plays a secondary role in this strategy, as these were already more or less established technologies in the 1980s. The second strategy relates to renewable energy (top right) and is focused on the large-scale production of solid and liquid fuels and power, with an active interest in lingo-cellulosic matter. This includes options like advanced combustion, pyrolysis and gasification. This requires significant research and focuses on power and fuel production. The third strategy (top left) is focused on knowledge intensive biotechnology, with high added value products produced in small quantities. Examples include applications like fiber, specialty chemicals, biopolymers, lubricants, paints, etc. More importantly, this third strategy relates to what became known as the bio-based economy, a vision that received increasingly attention after 2000. Although in principal this refers to the more broad development away from fossil fuels to bio-materials as feedstock for energy and materials, its more specific promise is building strongly upon biotechnology (see also Asveld et al., 2011; EuropaBio, 2008-2011; OECD, 2009).

41

We added a fourth strategy (bottom left) - one that was not mentioned by Chartier - that relates to small scale and knowledge extensive application. This includes farm based digestion system and traditional combustion of biomass. Clearly, developments can be described as shift in attention between the four discourses. This is especially relevant as each involves a different policy arena and different actors. European efforts in the early 1980s mainly focused on developing and applying small local systems, including digesters; and on research and development on renewable energy that focused on methanol production by wood gasification for blending in petrol. Both technology and feedstock were considered suited (on the long run) for large-scale application. The main driver is finding domestic sources that can substitute oil. In the late 1980s the context shifts, as the oil prices no longer provide an incentive and the agricultural overproduction increasingly becomes a problem. Ethanol as biofuel is reconsidered. In the 1990s attention shifts away from biochemical routes to thermochemical applications and the renewable energy strategy. After 2000 attention diverges again. There is continued interest in renewable energy: gasification in research and development and (co-) combustion in applications. At the same time the biofuel market opens up, while there is also an increase in attention for the vision of a bio-based economy. Note that small local applications have to some extent been part of these developments, like the emphasis on digestion systems in the 1980s. However, it never received high priority since, although several sources indicate that small system, especially combustion for space heating, continue to make a significant contribution (EurObserv’ER, 2008; Eurostat, 2010). Also note that, although Chartier already in 1990 points at the increasing importance of agro-food industry focusing on new high-added value products, the promise of a bio-based economy is mainly articulated by other and more recent literature and received limited attention on the conference. Apparently the conference is not the best platform for life sciences, biochemistry, biotechnology, genetics, etc. Some of the developments cannot be fully captured by the identified discourses. Examples are the recent interest in bio-refineries that can be seen as an attempt to bridge the biotechnology and the renewable energy approach; or the fermentation of lignocellulosic biomass, which is a cross-over between the biofuel route and the renewable energy route.

42

Of course, one can also focus on the developments within each of these discourses. For example ethanol production that in the early 1980s was mainly considered for small applications with significant attention for the fermentation process. During the late 1980s and 1990s its role as potential biofuel is strengthened. When this materializes after 2000, the discussion is no longer about the fermentation process and technology, but about biofuels in their socio-economical context: policy, barriers, economics, trade, logistics, sustainability criteria, etc. 2.7 Future of energy from biomass Our results show for energy from biomass the 1980s as an area or exploration; the 1990s initial application for power production and glorious expectations of the future; and after 2000 the market on biofuels that opens up but also leads to a public debate and wind that takes over the power market. Where does it leave us? Has energy from biomass outlived its days of success of the 1990s? And what does this mean for the four discourses identified? Will they blend together, or will one become dominant, or will they continue to co-exist?

Figure 2.3 Trends in energy from biomass in EU-15:

(a) in application (thousand tons of oil equivalents) (source: Eurostat, 2011); (b) total of national RD&D budgets on energy from biomass (million euro, 2009 prices and exchange rates;

(3-year moving average) (source: OECD/IEA, 2011); (c) number of publications, in this case in Biomass (1982-1990) and in Biomass & Bioenergy (1991-2010)

(3-year moving average) (Thomson Reuters, 2014). In figure 2.3 some additional indicators are presented for the EU-15 (Western Europe), namely the total application of biomass for energy purposes, the total government based RD&D budgets and trends in scientific publications. Each to some extent shows the increased attention during the 1990s. But what stands out is the sharp increase after 2005. Apparently, the opening up of the combustion and the

0

20.000

40.000

60.000

80.000

1980 1990 2000 2010

a) application

0

50

100

150

200

1980 1990 2000 2010

b) RD&D budget

0

25

50

75

100

1980 1990 2000 2010

c) publications

43

biofuel market and the start up of the international market has brought new dynamics to the field. According to Beurskens and Hekkenberg (2011) - based on projections of the EU member states - the increase in application over the next decade will mainly take place in conservative applications, like heating, power production and first generation biofuels. However, the promise of a bio-based economy recently received increasing attention. At the same time, the long term promises that have driven the field over the past decades persist in relevance: that of reduced dependency on oil and reducing greenhouse gasses - both of which will require a large scale and advanced applications. In short: all signs show that the upcoming 30 years will be at least as interesting - as well as diversified! As a result, the multiple discourses identified are likely to continue to co-exist.

44

Chapter 3

Biomass Gasification: still promising? A 30 year global overview3

3 Kirkels, A.F. Biomass gasification: Still promising? A 30-year global overview. Renewable and Sustainable Energy Reviews 15 (2011) 471-481

45

3.1. Introduction A reliable, affordable and clean energy supply is of major importance for society, economy and the environment – and will prove to be crucial in the 21st century. In this context modern use of biomass (as opposed to traditional use) is considered very promising. The promise includes a widely available, renewable and CO2-neutral resource, suited for modern applications for power generation, fuels and chemicals. Biomass has a distinct advantage over the use of other renewables, like solar cells and wind power, which are restricted because of the intermittent power generation. Biomass is by far the most applied renewable at this moment and a further increase is believed to be possible (Faaij, 2007). Gasification is a clean and highly efficient conversion process that offers the possibility to convert various feedstock to a wide variety of applications, see figure 3.1. It has been considered both in advanced applications in developed countries, as well as for rural electrification in developing countries. As such it has been considered the enabling technology for modern biomass use. This raises the question whether it will be able to live up to these lofty expectations.

Figure 3.1 Gasification technology offers flexibility and enables biomass use in advanced

applications.

Textbox 3.1 Characterization of gasification technologies.

• Fixed bed technology: a fixed bed of feedstock is being gasified using a gasification medium, generally air at low velocity. Main subtypes are downdraft and updraft gasifiers, which are mainly applied at smaller scales.

• Fluid bed technology: a small fraction of feedstock is added to a much larger fraction of bed material, which is fluidized by a gasification medium (air, oxygen, steam) that flows through the bed at a high enough speed. Main subtypes are the bubbling and the circulating fluidized bed, which are mainly applied for biomass at medium scales. • Entrained flow gasification: small droplets or particles of feedstock are ‘entrained’ in a flow of

gasifying medium – in general oxygen or steam. Also referred to as suspension flow or dust cloud gasifiers. It has been mainly applied at larger scales for coal and petroleum based feedstock.

Gasifiers • Fixed bed

- downdraft - updraft

• Fluid bed • Entrained flow

Feedstock and pretreatment • Wood • Crop residues • Peat • Black liquor • Waste

Gas cleaning

Applications • Heat • Electricity • Chemicals • Transport fuels

46

Gasification can be understood as a thermo-chemical conversion with limited oxygen supply. This results in the production of a producer gas (also called syngas) with significant amounts of carbon monoxide (CO) and hydrogen (H2) and a low to medium energetic value. In textbox 1 the basic characteristics of the main gasifier technologies are presented – see also appendix A for a more detailed description. Gasification has a long history, with applications in town gas in the 19th and 20th century and a revival of small-scale gasification during World War II, due to an acute shortage in liquid fuels. More recently the oil crisis in the 1970s played a major role in the renewal of interest for biomass gasification. Since then a significant research, development and demonstration (RD&D) effort has been launched in Europe and North America, of which the more recent part has been extensively covered by literature, see e.g. Faaij (2006b), Knoef (2005a) and Maniatis (2001). This literature shows a true kaleidoscope of designs. This can be considered indicative for the large interest in the technology, but to what extent has this resulted in a distinct technological trajectory, one that justifies the high expectations? In this chapter we attempt to answer this question and we will assess the future potential of biomass gasification, based on a 30-year overview of the development of this technology. Drawing on evolutionary theories, we will study global developments, which will allow us to make geographically differentiated conclusions and include spill-over between regions. In addition, we will study both coal gasification and biomass gasification, because the literature suggests they share many characteristics: general gasifier designs, hot gas cleaning and (foreseen) applications (Babu, 1995; Babu and Whaley, 1992). After addressing methodological issues, we will provide an empirical overview based on literature and relevant science and technology indicators. The latter add detail, especially on dynamics, and ensure the coverage of global trends. In the final section we will come to conclusions and reflect on the chosen approach. 3.2 Theoretical framework and methodology 3.2.1 Evolutionary framework We will follow an evolutionary approach to analyse technological development. This approach addresses long-term processes that contain multiple product sequences; it is based on the mechanisms of variation, selection and retention. Variation refers to the creation of new designs by engineers and scientists in research and development laboratories or research institutes. This variation is not blind. It arises from firm specific differences in search processes and research and development, in attempts to generate alternatives and seek solution to problems (Dosi, 1982). These search

47

processes are directed by agency, strategies and expectations. History matters, as the process of variation builds on existing products and routines. The variations lead to somewhat different products, which compete in the market. Selection mainly takes place in these markets (sales, profits), although there is also selection on knowledge by communities of engineers and by firms internally and to the extent technologies are socially legitimate as reflected in governmental regulations and social norms. Retention, finally, refers to the mechanisms that retain the reproduction of successful variations, a process of institutionalization (Faber and Frenken, 2009; Geels and Schot, 2010). In this approach, different technologies can also co-evolve. A specific case are technologies that are rather ‘close’ and share a specific technology base, as this allows for spill-over and shared technological learning (Sandén and Jonasson, 2005). In our case there are two such cases to consider: the co-evolution of coal and biomass gasification, and of biomass combustion and gasification. Nelson and Winter (1982) considered the directed development of technology in the phase of variation as a sort of coordination within a population of firms (industry, sector). Because of shared routines, engineers in a technological field work in more or less the same direction. Hence, “sequences of minor variations … add up to global technological trajectories that proceed in particular directions” (Geels and Schot, 2010, p.37). These technological trajectories are not equivalent to focusing on a single technology. Contrary, evolutionary theory assumes that firms will be using somewhat different technologies, characterized by different fitness as expressed in profits. Only over time this will result in a dominant technology and dominant firm(s), as market selection singles out the best performing technology. Another source of diversity are the application of a technology in specific market segments or niches. “In markets with heterogeneous products, there are various user groups that differ in their valuations of a technology’s characteristics. In such environments, new technologies can be introduced in niche markets. … Once introduced, users and producers start learning and will introduce subsequent improvements. Such a gradual process allows the technology to diffuse niche-by-niche.” (Faber and Frenken, 2009, p.465) This is especially important, as new technologies often require protection from mainstream market selection to become more competitive and be able to escape lock in – for example from the fossil fuel based economy. Niches are not only present under heterogeneous market conditions, but can also be socially constructed by actors that are willing to invest time and resources in a new technology, because they believe in its potential (Schot and Geels, 2007). In that case (temporary) protection can come from policy makers through subsidies and regulatory adaptations. The recent strand of literature on Strategic Niche Management (SNM) is building on that idea (Ulmanen et al., 2009).

48

As such the evolutionary framework offers the possibility to deal with both diversity and directed development. We will reconstruct the technological trajectory of (biomass) gasification technology by identifying the main variations (designs, actors) and the major drivers and barriers – that is the (anticipated) socio-economical selection environment (niches) that directed the technological development. Special attention will be given to countries, as they provide protection and provide a specific selection environment for technology development, including factors like the presence of biomass and relevant industries, government regulations on energy and waste, and the attitude of the people towards environmental problems. In addition, governments also have more active involvement in innovations processes, by setting RD&D budgets, research agendas and making policy. 3.2.2 Methodology For the reconstruction we will apply two overlapping yet distinct methods. The first is a literature study, including overview articles, government reports, commercial status reports and books. A significant part of this literature has been written for IEA Bioenergy, an organization with the aim of improving cooperation and information exchange between countries. The literature offers a great detail on developments, including operational capacity, current status, socio-economical context (driving forces and barriers) and expectations and visions. The focus is mainly on the most successful and promising technologies. Detailed publications on technological developments have not been included. Second, we will present results of science and technology indicators related to biomass gasification. This offers the possibility of cross-examination of trends identified from the literature. The indicators present the dynamics over the full population, including the minor variations that add up to global trajectories. It also allows to identify relevant companies, research institutes and countries. Science and technology (ST) indicators are widely used. Indicators that we will use are publications to measure research activities or scientific productivity and patents to assess knowledge based innovation and commercialization activities. Both are long-term output indicators of the RD&D process. The content of patents and scientific articles are hardly overlapping, they are complementary to each other (Geisler, 1994; Oppenheim and Allen, 1979).4

4 Also input indicators of the innovation process can be used, like RD&D budgets. IEA holds a database on countries’ energy related RD&D spending (OECD/IEA, 2011). However, this indicator holds several disadvantages: the data are only available for IEA countries - which

49

The general patenting and publishing rate have increased significantly over the years. Our constructed time series have been corrected for that and represent trends as if overall patenting and publishing rate had been stable – thereby only indicating differences in interest in (biomass) gasification. With respect to geographical coverage, patent sets show the tendency to over-represent domestic patents and patenting intensity is not equal for all regions. We used patent sets from multiple offices, thereby reducing this geograhical bias. 3.3 Gasification An overview of large operational gasification plants (of capacity over 100 MWe electric equivalents, commercial industrial scale) has been presented by US DoE (2000, 2005, 2007a, 2007b). They include 144 plants and 427 gasifiers. The market is dominated by coal and petroleum based gasifiers. At least 15 different gasification technologies are in operation, of which three technologies are dominant: Sasol Lurgi, GE Energy (former Texaco) and Shell. The latter two are dominating recent developments and are examples of entrained flow gasification. Since 2001 new plants have mainly been built in China. Gasifiers in Europe are mainly located in Germany.

Figure 3.2 Accumulated capacity of main applications of gasification (data: US DoE, 2007b).

are mainly the developed countries; there are some gaps in data; only government based RD&D expenditures are covered, no private expenditures; and the data are only available at the aggregated level of ‘energy from biomass’, which is much more encompassing that just biomass gasification. For that reason we excluded it from our analysis.

02.0004.0006.0008.000

10.00012.00014.00016.00018.000MWth Fischer Tropsch Ammonia Methanol

Electricity hydrogen

50

The total capacity shows a growth until 1985, followed by a decade of market stabilization, to start growing again from 1993 on. Figure 3.2 shows the development of accumulated operational capacity for the five dominating applications. Leading applications are Fischer-Tropsch liquids (29% of 2006 capacity, 4 plants) and power production (24%, 22 plants). The production of chemicals like ammonia, methanol and hydrogen is steadily increasing. Other minor applications – not shown in the graph - are oxochemicals, carbon monoxide and others. All applications except integrated gasification combined cycles (IGCC) have already been applied before the 1970s. The few Fischer-Tropsch installations are typically very large plants, located mainly in South Africa (Sasol). They were built in the late 1970s and early 1980s under the Apartheid regime, when South Africa faced an international oil embargo. Gasification in combination with Fischer-Tropsch synthesis made it possible to convert the South African coal reserves into fuels and chemicals. Electricity production has been a late success, only gaining real significance in the late 1990s when IGCC started to find commercial application. IGCC made a coal-to-power conversion possible that was more efficient and less polluting. It has been mainly applied in Europe. The production of ammonia takes place in India and China. Ammonia is used for the production of fertilizers. The production of methanol is located in Germany and China. Its largest use by far is in the production of other chemicals and biodiesel. The two largest uses for hydrogen are in fossil fuel processing (e.g. hydrocracking) and ammonia production. The production of Substitute Natural Gas (SNG) has not been included in figure 3.2. SNG is methane that is produced by a gasification reaction, opposed to natural gas that is produced at oil and gas fields. Its production was especially considered in the USA in the 1970s and early 1980s. It included the development efforts by Exxon on a catalytic fluid bed technology. Another development effort resulted in the Great Plains plant in the USA in 1984 that proved to be uneconomical. Developments have stagnated since (Exxon, 1978; Harmsen, 2000; Longwell et al., 1995). Most of the research and development of coal gasification took place in Germany and USA, although the UK and Japan also played a role. Many developments in coal gasification are based on classical German coal gasification technologies: the Lurgi moving bed, the Winkler fluid bed and the Koppers Totzek entrained flow processes. Major developments in Germany were still going on in the 1970s and early 1980s. For the USA coal utilization for power production was and still is the core to its national

51

energy strategy. It was a motivation to start development programs for coal gasification in the 1970s after the first oil crisis (Harmsen, 2000; Longwell et al., 1995). According to Harmsen (2000) three different periods can be distinguished in the development of coal gasification between the 1970s and the 1990s. In the 1970s many companies started to develop their own processes and aimed at various applications. This period has been dominated by the two oil crises resulting in high prices for oil and strong concern by countries about self-sufficiency and fuel diversification. The second period covers the 1980s, in which a number of process developments were stopped that became economically unattractive due to decreasing oil prices. Remaining developments focused on IGCC that promised to be superior on environmental performance at similar costs compared to conventional coal-fired plants. This was considered especially important, as nuclear based electricity production was no longer considered a promising option. Also environmental awareness increased and climate change was first put on the international political agenda. In the third period, in the 1990s, the focus diverged again, considering more fuels and applications. However, IGCC had reached the stage of commercial demonstration. The period can be characterized by liberalization and deregulation of the electricity markets. This resulted in a strong preference for gas based power production, since it was a cheaper, cleaner and a reliable and proven technology. However, the liberalization also offered opportunities for gasification technology. An example is the poly-generation in refineries, in which oil residues are gasified to produce hydrogen and steam for the refinery and in addition electricity is produced and supplied to the grid (US DoE 2007a). More recent driving forces are the expanding economies, in particular China, the growing concerns regarding CO2 emissions and the volatile fuel prices (US DoE, 2005, 2007a). China has de facto become the world’s test-bed for large-scale coal utilization processes. According to Henley (2007), key economic factors are driving China to coal: very low costs of coal and labor and advantageous state financing. However, the largest contributing factor is the strong support from the Central Government and from provincial and local level. Coal gasification has not only found application, but has also been subject of Chinese research since the late 1990s (Liu et al., 2008). Ongoing concerns regarding CO2 emissions have resulted in both barriers and opportunities for further development. They are a fertile soil for anti coal sentiment. However, if carbon emissions become increasingly regulated, gasification based technologies will benefit from this: they offer increased efficiency and allow for carbon capture and storage (CCS). Decisions to move forward with gasification projects also depend highly on costs and prices of energy. Gasification applications are most directly in competition with

52

natural gas based technologies, both for power applications as for the production of ammonia, methanol and hydrogen. Between 1990 and 2000 natural gas was available at a relative low costs, after which costs increased with a factor three until 2007. This has improved the economic competitiveness of gasification technology (US DoE, 2005, 2007a). The US Department of Energy is continuing its gasification program, both research and development, to improve the overall economic and technological competitiveness, also in advanced applications. Each of the crucial components of the gasifier is subject of study, offering potential for improvement (Breault, 2008; US DoE, 2007a). 3.4 Biomass gasification 3.4.1 Overview In biomass gasification mostly wood is considered as feedstock. However, also peat, black liquor and rice hush gasification have been demonstrated. Black liquor is a byproduct of the paper industry, a lignin-rich mixture of cooking chemicals and dissolved wood material. Rice husk gasification has found application in Asia (Babu and Whaley, 1992; Morris et al., 2005). Canada, Finland, Sweden and the USA have been initially involved in the development of biomass gasification. Each of them has large woody biomass and/or peat resources. In the 1970s especially the USA fulfilled a leading role in response to the disruption of oil supply and high oil prices. This involved research and rapid development of gasification concepts. The potential to substitute natural gas or transportation fuels was viewed as being very important. However, initial applications were less advanced and focused on heat and power applications. Energy research in the 1980s shifted focus to long-term high risk research. Most financial incentives that were needed to stimulate the commercial use of biomass energy were eliminated – and so were many projects and plants (Klass, 1995; Miles and Miles, 1989; Stevens, 1994). Circulating fluid bed (CFB) gasifiers have been first applied in the early 1980s by Lurgi (Germany) and Ahlstrom (Finland, now Foster Wheeler). According to Basu (2006) both were based on their respective CFB combustion designs that were developed separate from the large government funded programs. Lurgi used its experience in ore roasting fluidized beds. Ahlstrom, a Finish engineering company and established producer of pulp and paper products, became interested in the technology as a method of burning a wide range of ‘difficult’ fuels for this sector – including biomass and bark (Watson, 1997).

53

The 1990s brought increased awareness of climate change, which resulted in a renewed interest in biomass gasification (Babu, 1995; Babu and Whaley, 1995; Klass, 1995; McGowin and Wiltsee, 1996). While some developments in the USA continued, European countries became increasingly involved. Germany and Austria have joint Sweden and Finland as leading countries, while many others became involved in development and implementation, including Netherlands, Italy, UK, Switzerland and Denmark (Dorca Duch and Huertas Bermejo, 2008; Kwant, 1998; Kwant and Knoef, 2004). Especially in countries with strong support for renewables and with availability of biomass, the development of biomass gasification has become an established practice (Faaij, 2006b; Morris et al., 2005). By 2005, the status of the technology was such, that there was significant interest for gasification but hardly any new commercially projects were implemented (Knoef, 2005a). A gasifier plant not only consists of a gasifier, but also includes feedstock pretreatment and feeding, gas cleaning and the end-use application. Over time each of these processes has been subject of continued research and development, with addition of the subjects of systems integration and scientific understanding of the gasification process. Since the late 1990s a significant amount of effort has been focused on gas cleaning (Babu, 1995; Kwant and Knoef, 2004; Maniatis, 2001; Kurkela and Kurkela, 2009). 3.4.2 Applications Modern use of biomass has been mainly based on combustion. By 2000 40 GWel of electricity production capacity was installed worldwide and 200 GWth of heat production capacity – over 90% of which was based on combustion. It has been successfully used in the lumber, paper and pulp industry and the cane-based sugar industry - sectors that also provide a huge potential for biomass gasification. Biomass gasification is applied on a much more modest scale, totaling about 1,4 GWth. Gasification has the advantage over combustion of more efficient and better controlled heating, higher efficiencies in power production and the possibility to be applied for chemicals and fuel production (Faaij, 2007). An overview of accumulated operational capacity of biomass gasifiers for different applications is provided in figure 3.3, based on Hellsmark (2007) who created a database for his research on biomass gasification in Europe. Details on the database have not yet been published. The overview is in line with developments described in the literature that also provides more detail on its current status (see Babu and Whaley, 1992; Faaij, 2006b, 2007; Kwant and Knoef, 2004; Maniatis, 2001; Morris et al., 2005). All applications of biomass gasification show an increase over time. This

54

increase is most significant for combined heat and power (CHP), which has become the main application.

Figure 3.3 Accumulated capacity of main applications of biomass gasification (data: Hellsmark, 2007).

Since the 1980s gasifiers for heat applications have been installed at lime kilns in the paper industry and cement kilns: a relative simple niche application in a sector that combined availability of feedstock and heat demand. Initial applications can be found in the USA, Sweden and Finland. The application has achieved commercial status, meaning that guarantees are supplied and the technology is competitive, but it shows limited diffusion. The heat applications were followed in 1990 by the first CHP applications, using diesel or gas engines. Deployment has been limited due to relatively high costs, critical operation demands, and fuel quality. It has been difficult to find areas where both heat and electricity feed-in tariffs are high. Deployment has often been related to national support, like the Swedish carbon tax and the Austrian CHP program. IGCC became the center of attention in the 1990s, mainly in Europe, in response to the promising results for coal IGCC. This is especially true for Europe. The technology promised high electrical efficiency at modest scales combined with modest capital costs. This resulted in a significant research and development effort and a few demonstration plants. But over time these plants have been canceled or shut down. Besides some technical issues this has mainly been due to the innovation gap: the unattractive phase between demonstration and market application characterized by an unproven technology available at high costs. Since 1998 biomass gasifiers have been implemented at coal-fired power stations for indirect co-combustion: the biomass is gasified, after which the resulting producer gas

0

100

200

300

400

500

600

700MWth

Heat or Kiln CHPCo-fire Fuels

55

is combusted together with the coal. This way biomass is introduced in the power industry with a minimum of potential risk to the boiler and to the quality of by-products, and this can be done at reasonably high efficiencies and limited costs. Interest in larger biomass co-firing shares and utilization of more advanced options is increasing. However, in practice often direct co-combustion is preferred, in which biomass is not gasified but directly combusted together with the coal. The latest development is a shift in interest to transport fuels (second generation fuels produced by gasification). These fuels offer a much better greenhouse gas performances and less competition with food compared to first generation biofuels. In addition, the transport fuels are high value energy carriers that might justify the use of (more expensive) cultivated biomass. Already in the 1980s, methanol, dimethyl ether (DME), Fischer-Tropsch liquids and hydrogen played a role, both in Europe as in the USA. For example, methanol production was tested and developed in France, Sweden and Canada. In Europe only recently, pushed by the EC biofuel directive (2003), attention for those routes is evident again in research programs of the EC (6th and 7th framework program) and countries like the Netherlands, Sweden and Germany. Current applications can best be qualified as research and demonstration and include the partnership involving Choren Fuel and the Schwarze Pümpe plant in Germany. The technological challenges are complex, since gas cleaning needs to be very effective in order to protect downstream catalytic gas processing equipment. There is a high confidence that once clean syngas is available, known process technology for producing the fuels can be applied (Faaij, 2006b; Maniatis, 2001; Morris et al., 2005). Also in the US biomass-derived fuels have received increasing attention in recent years, including a mandatory setting for renewable fuels. However, there is a focus on cellulosic ethanol, which does not require gasification. Nevertheless some research is going on for biomass gasification to produce clean syngas for the production of ethanol or other fuels or chemicals (Office of the Biomass Program, 2009). Advanced applications require significant upscaling: for IGCC a typical commercial scale is considered of 30-200 MWe; 50-200 MWth for the chemical sector and several 1000 MWth for transportation fuels. The required scales are the effect of market size and economies of scale (costs minimization). Besides the challenge of scaling up, this presents a mismatch with the dispersed availability of biomass of low energy density: larger plants result in higher feedstock costs because longer transport distances are involved. This can be overcome by converting biomass to an energy carrier with higher energy density, for example by local pyrolysis plants. Another option is to focus initially on applications that require less biomass capacity: co-gasification with coal or co-combustion of producer gas with natural gas in a combined cycle. A third option is

56

the cogeneration of multiple products in a pulp and paper mill. This would offer product flexibility and added value for a sector that provide the feedstock themselves (Boerrigter and Rauch, 2005; van der Drift and Boerrigter, 2006; Faaij, 2006b). Another problem of modern applications are the high initial investment costs, especially of the first plants, in combination with the risks involved. To achieve a reduction in capital costs for IGCC plants requires at least several successful demonstration plants, which themselves would be uneconomical. However, the liberalization of the energy markets has resulted in decreased direct support from national governments for technology development and of reducing investments of the energy sector in risky technology with long term benefits. This has stalled the application of the promising application (Bridgwater and Bolhàr-Nordenkampf, 2005; Faaij, 2006b; Morris et al., 2005). A similar effect can be expected in fuel applications, since these require even larger investments and are considered more risky (Bole and Londo, 2009; Hervouet, 2009). Three applications have not yet been discussed. The first one is high temperature fuels cells, which have only been considered in research (Babu, 1995); the second one is waste gasification; and the third one are applications in developing countries. Waste gasification found application in Japan since 1997. Main drivers have been the shortage of landfill and the policy to avoid incineration, emissions of dioxins and to increase plant size. The policy has been supported by technology development and demonstration programs. All leading Japanese thermal process companies now offer gasification solutions. The Nippon Steel (updraft) and Ebara (fluid bed) technology are fully commercial (Schwager and Whiting, 2003; Tanaka and Johnson, 2005). However, these technologies in general focus more on the production of manageable secondary waste products (solidified ash) than on energy recovery (Morris et al., 2005). Another country with significant experience in waste gasification is Germany. For developing countries the promise of rural electrification and local development has been a major driver for projects, amongst others by the World Bank and Western countries in the 1980s. These attempts were not very successful (Stassen, 1995).5 Main developments over the past decades can be found in India and China (Bhattacharya, 2005), that is for development, manufacturing, application and diffusion of the technology. Hundreds to thousands of small fixed bed gasifier systems

5 We have been very brief about these developments, as they are not the focal point of this article. However, rural application in developing countries really gained momentum in the early 1980s. Significant interest and early applications can be found in Brazil, Philippines, India and Indonesia, while a much broader set of countries showed interest - see for example FAO (1986), Foley and Barnard (1983), Gunawardhana/The Beijer Institute (1982) and Martins Coelho/The Beijer Institute (1985).

57

have been installed (Banerjee, 2006; Bridgwater et al., 1999). Applications remain troublesome, with problems regarding tar, operation, maintenance and economic feasibility. Ghosh, Sagar and Kishore (2004) and Verbong, Christiaens, Raven and Balkema (2010) conclude that small-scale rural electrification might not be the best way to introduce gasification technology in developing countries, especially India. Thermal applications are already applied and have a better track record. And medium scale power generation, either in industry or rural grid connected, is likely to face fewer difficulties (economical, technical and practical). 3.4.3 Technologies In contrast to coal gasification, which is dominated by entrained flow technology, in biomass gasification a range of technologies has been applied. At the end of the 1980s and the beginning of the 1990s, downdraft and updraft gasifiers with capacities of less than 100 kWth and up to a few MWth were developed and tested for small-scale power and heat generation (Faaij, 2006b). More recently the downdraft technology has become dominant, especially for power applications, due to its low tar content in the producer gas (Dorca Duch and Huertas Bermejo, 2008; EWAB, 2001). Major applications can be found in India and China – as was discussed above. Extensive lists of manufacturers can be found in the literature (see Biomass Energy Foundation, 2008; Bridgwater et al., 1999 (China); EWAB, 2001; Heerman et al., 2001; Kaltschmitt et al., 1998; Knoef, 1996, 2005b; Verbong et al., 2010 (India)), many of which are small companies with limited resources and some with a regional orientation on the market. Quite typical for this stage of market formation are takeovers of companies and technologies and new companies entering the market. For most equipment-suppliers gasification is not their core business, and plants are not mass produced. A selection of leading companies (most applied or advanced) for developed countries is presented in table 3.1.

58

Table 3.1 Leading manufacturers and technologies for small-scale biomass gasifiers (updraft and downdraft) in developed countries (Knoef, 2005b; Juniper, 2007).

Technology / company Country Gasifier

1 Bioneer (now Foster Wheeler) Finland Updraft, heat

2 PRM Energy Systems Inc. (PRME) USA Updraft, heat/power

3 Babcock Wilcox Volund Denmark Updraft, heat and power

4 REL Waterwide technology New Zealand Downdraft, heat

5 Chiptec Wood Energy Systems USA Downdraft, heat

6 Fluidyne Gasification New Zealand Downdraft, power

7 Xylowatt Belgium Downdraft, power

8 AHT Pyrogas Vertriebs Germany Double zone, heat and power

9 COWI/DTU 'Viking' gasifier Denmark Multi stage, electricity

10 Biomass Engineering UK Downdraft

11 ITI Energy UK Fixed bed, proprietary design

12 Puhdas Energia Oy Finland Downdraft

13 Host Netherlands Fixed bed

14 Condens Oy – Novel gasifier Finland Fixed bed, counter current bottum

In the case of nearly all medium-to-large scale gasification plants for power production, the preferred technology has been atmospheric circulating fluidized beds (CFB): it can handle a high throughput, is relatively easy to scale up and is capable of accepting a wide range in fuel quality – both in particle size and in ash properties (Dorca Duch and Huertas Bermejo, 2008; EWAB, 2001; Morris et al., 2005). However bubbling fluidized bed systems (BFB) are also still applied. Air blown gasifiers are preferred for heat and power applications, while the more advanced applications require oxygen blown gasification (and therefore an oxygen plant). Pressurized systems are considered for larger capacities and for IGCC, fuels and chemicals – in these cases the final conversion will take place under pressure anyway. Advanced applications require a strict gas cleaning. A specific application in fluid bed technology is the indirectly heated gasification, characterized by a separate gasification and combustion reactor/zone. In general, these gasification reactors are steam blown, which results in a higher heating value of the produced gas without the need of oxygen. This requires an air blown combustor to provide the required heat to keep the gasification reactions going. A disadvantage is the increase in complexity of the installation. Also oxygen blown entrained flow gasification is considered – the preferred technology for coal gasification – since it has the advantage of operating at very large capacities and producing clean syngas. However, this will require significant

59

pretreatment of the biomass (pyrolysis or CFB gasification), since entrained flow systems only gasifies very small particles. In the last decade a wider variety of technologies has found application, like pressurized and entrained flow gasification. Also plasma gasification for waste and gasification in supercritical water of wet biomass (like sewage sludge and pulp waste) has been further developed. Leading suppliers of large scale and advanced gasifiers are presented in table 3.2. Heerman, Schwager and Whiting (2001) emphasize that for different market segments and feedstock also different processes are leading. Especially Foster Wheeler (Europe) and Ebara and Nippon Steel (Japan) have realized several gasifiers (Morris et al., 2005; Schwager and Whiting, 2003).

Table 3.2 Leading gasification concepts for large scale or advanced cycles (Ciferno and Marano, 2002; Heermann et al., 2001; Juniper, 2007; Kavalov and Peteves, 2005; Morris et

al., 2005).

Company Countr

y Gasifier

1 Gas Technology Institute (GTI) - Renugas technology (Institute of Gas Technology (IGT))

USA BFB, air/oxygen blown, pressurized

2 Repotec Umwelttechnik/Austrian Energy and Environment (Güssing CHP plant)

Austria BFB, indirectly heated, steam blown (CFB air combustor)

3 Enerkem Technologies Inc. - BIOSYN technology Canada BFB, air/oxygen blown, pressurized

4 ThermoChem, (Manufacturing and Technology Conversion International (MTCI))

USA BFB, pulse enhanced, indirectly heated, steam blown, atmospheric, (also) black liquor gasification

5 Envirotherm GmbH, part of Allied Environmental Solutions Inc. (Lurgi technology, BGL at Schwarze Pümpe)

Germany/ USA

- BGL fixed bed, slagging bottom, pressurized; - CFB, atmospheric

6 Rentech Inc. - Rentech-Silvagas technology (Battelle Columbus Lab/Future Energy Resource Corporation (FERCO))

USA CFB, indirectly heated, steam/air blown, atmospheric/low pressure

7 TPS Termiska Processor AB (ex Studsvik Energiteknik AB) Sweden CFB, air blown, atmospheric

8 Foster Wheeler (ex Ahlström) USA/ Finland

CFB, air blown, atmospheric/pressurized

9 Ebara - Twin Rec UEP Gasification technology Japan CFB, gas to slagging combustor, air blown, waste

10 Choren Industries GmbH - Carbo V technology (Deutsche Brennstoff Institut)

Germany Entrained, involving pre-gasification or pyrolysis, air/oxygen blown, sewage sludge

11 Chemrec A.B. (ex Kvaerner Pulp & Paper) Sweden Entrained, air/oxygen blown, black liquor

12 Thermoselect S.A. Switser land

Pyrolyzer and entrained char gasifier, oxygen blown, waste

13 Siemens Fuel Gasification Technologies GmbH, (Future Energy, BBP, NOEL-KRC, Deutsche Brennstof Institut)

Germany Entrained, oxygen blown, pressurized

14 Energy Products of Idaho USA BFB

60

3.4.4 Conclusions The overview presented up till now had a focus on existing (demonstration) plants and current status. Developments in gasification for both biomass and coal are summarized in table 3.3. The technology has been successful in a few niche markets, but in general still is confined to RD&D niches. The literature research has been limited in its description of the RD&D stage. In addition it is lacking detail on demarcation of relevant time periods, the shift in development efforts over different nations and the role of India and China. In the next section we will try to fill these gaps by using science and technology indicators.

Table 3.3 Coal versus biomass gasification: a summary.

Coal gasification Biomass gasification Preferred technology

Entrained flow Updraft (small, mainly heat) Downdraft (small, mainly power) Circulating fluid bed (large) Entrained flow (large, fuels and chemicals)

Main applications (niches)

Fischer-Tropsch (South Africa) IGCC power Poly-generation in refineries China (ammonia, methanol)

Heat Combined heat and power Co-combustion IGCC (research) Fuels and chemicals (research) Rural electrification / developing countries Waste gasification

Scale 100-1000’s MWth 0.05 – 10’s MWth Dominant suppliers Lurgi, GE, Shell Multiple Dominant countries USA, Germany, China USA, Finland, Sweden, Germany,

Austria Japan (waste) China, India (small scale)

3.5 Science and technology indicators Patent indicators have been derived from datasets of the United States Patent and Trademark Office (USPTO) and the European Patent Office (EPO). In addition the datasets of patent offices of Japan, China and India have been analyzed in less detail. C10J3 is a patent class of the World Intellectual Property Organisation that covers patents on the production of combustible gases containing carbon monoxide from solid carbonaceous fuels. This class is most representative for gasification, as can be concluded from sampling of patents using a variety of key words. Within this class we

61

have made differentiations based on key words towards biomass and gasification technologies. Applications are checked upon in combination with the stem ‘gasif’ or a syngas synonym. Articles on biomass gasification are checked upon by using Web of Science by Thomson Reuters (2014). Web of Science is an international multidisciplinary index for journal articles in all sciences. It encompasses references to articles of over 8500 journals and thereby is one of the most complete databases available. Time series are constructed for several queries related to gasification (‘gasif*’), gasification technologies, feedstock (‘biomass or wood’) and applications. Applications are checked upon in combination with the stem ‘gasif’ or a syngas synonym. All were ‘topic’ searches, which means that the engine searches within the title, abstract and keywords. 3.5.1 Dynamics: intensity over time For both patents and publications, trends are shown over time. The results are presented in figure 3.4 for gasification and in figure 3.5 for biomass gasification. The maximum for each indicator is indexed at one. The gasification trend line shows one active period, starting in the 1970s and lasting until 1985 (patents) – 1990 (publications). For biomass gasification two relevant periods can be identified: the first one between 1981 and 1988, with a limited contribution from publications; and the second one from 1998 onwards including the 2008 peak.6 The USPTO dataset does not show this second peak – probably due the decreasing activity in the USA, as will be discussed in the next section. Concerns on climate change are presented in the literature as being the major driving force for the renewed interest in biomass gasification since the 1990s. These concerns have been articulated only recently (2002-2006) on a limited scale in the USPTO and Web of Science datasets.

6 The increase in publications on biomass gasification after 1990 might be the result of the expansion of coverage of journals by Web of Science (see the Thomson Reuter Community 2010 on this topic). However, also specific journals that are most relevant for the topic of biomass gasification and that had been included for longer periods show a strong upcoming publication rate. We checked ‘Fuel’ (1976-2010) and ‘Industrial Engineering Chemistry Research’ (1987-2010). So, the increase in publications is at least not only due to the expanding coverage of Web of Science.

62

Figure 3.4 Gasification, intensity of indicators over time (data: EPO, 2009; USPTO, 2009;

Thomson Reuters, 2014).

Figure 3.5 Biomass gasification, intensity of indicators over time (data: EPO, 2009; USPTO, 2009; Thomson Reuters, 2014).

3.5.2 Relevant countries In the USPTO and EPO patent sets three countries are dominant: USA, Germany and Japan. Apparently these countries hold the relevant industry – possibly with the recent addition of China that scores well on publications and in the Chinese patent set. Web of Science on publications provides a better worldwide coverage. It shows an increasing interest in biomass gasification: both the total number of countries

0,00

0,20

0,40

0,60

0,80

1,00USPTO patents

EPO patents

ISI Web of Sciencepublications

0,00

0,20

0,40

0,60

0,80

1,00

USPTOpatents

EPO patents

ISI Web ofSciencepublications

63

increases (the base) as well as the number of countries that have a significant amount of publications (the top). Figure 3.6 present the geographical distribution of publications on biomass gasification over time.

Figure 3.6 Geographical distribution of publications on biomass gasification (data: Thomson Reuters, 2014).

The figure shows that until 1988 the USA is dominating. Since 1997 Europe has become leading and the differences with other regions have increased ever since. Japan has only recently contributed to publications. These trends are in line with trends in patents. Chinese publications are also of recent date. China is hardly present in the USPTO and EPO datasets. However, Chinese patents confirm that attention for biomass gasification, although rather limited, has been coming up since 2000. China scores much better on gasification in general. Over half of the Chinese patents for both gasification and biomass gasification is held by Chinese companies – indicating a strong Chinese development effort. India has been publishing on biomass gasification already since 1989 and takes now a 6th position in Web of Science. However, Indian companies hardly patent on gasification, not in India nor abroad. The literature suggest that for specific feedstock distinctive development paths are in place with specific countries involved. Therefore these have been assessed in Web of Science – see table 3.4. They show domination by USA and Japan, with other

0

20

40

60

80

100

120

140

160# publications

EU

USA

Japan

China

64

countries involved for specific feedstock. Note that in this case the overall contribution of the EU has not been taken into account, only of its member states.

Table 3.4 Dominant countries involved in the gasification of different feedstock (data: Thomson Reuters, 2014)

Biomass Wood Peat Black

liquor Municipal

waste Agricul-

tural Sludge Rice

husk USA USA Finland USA USA USA USA India

Japan Japan USA Sweden Japan Greece Japan China China Finland Turkey Canada

Spain

3.5.3 Companies and institutes With respect to companies involved in gasification USPTO and EPO datasets are very consistent: Texaco and Shell score best with their respective technologies that currently dominate the market for coal gasification. Other companies in the top include: Metallgesellschaft (now marketed by Envirotherm) that developed the Lurgi fixed bed technology (applied in South Africa) and fluid bed technology; Exxon that was developing a catalytic fluid bed coal gasification process to produce SNG in the late 1970s and early 1980s; Krupp Koppers (now Uhde) that developed the PRENFLO entrained gasification process; and Foster Wheeler that worked on fluid bed gasification and is market leader in fluid bed combustion. In addition, in the USPTO patent set also the US Department of Energy is present as well as Combustion Engineering (now part of Alstom), whereas in EPO patent set the Japanese Ebara and Mitsubishi Heavy Industries take a prominent position. For biomass gasification the relevant companies are not that clear. About 80-85% of the patents are held by companies that only have 1 or 2 patents. Therefore we conducted a reversed search: for all companies in tables 3.1 and 3.2 we checked the datasets on their presence. Only a few companies involved in small-scale gasifiers hold patents. In contrast, most companies involved in large scale gasifiers hold patents, but they refer limitedly to biomass. Web of Science shows a much broader interest among research institutes. An overview of leading institutes is presented in table 3.5. On biomass gasification the Chinese Academy of Science is leading, followed by a broad range of institutes. The dominance of Europe since 1997, as was clearly shown in figure 3.6, is hardly present

65

in table 3.5. Apparently in Europe a large number of smaller institutes have been publishing, limiting the impact per institute.

Table 3.5 Rank of top institutes in biomass gasification. First mentioned is rank, followed by number of publications between brackets (data: Thomson Reuters, 2014).

Gasifi-

cation Gasifi-cation

biomass

Fluid bed

Fluid bed

biomass

Fixed bed

Fixed bed

biomass Tohoku University (JP) 1 (192) Chinese Academy of Science (CN) 2 (181) 1 (69) 1 (71) 1 (34) 2 (26) 4 (9) Pennsylvania State University (USA) 3 (147) Consejo Superior de Investigaciones Científicas (ES)

4 (127) 5 (12)

Hokkaido University (JP) 5 (92) 5 (29) 3 (20) 2 (14) Forschungszentrum Karlsruhe (DE) 2 (37) Nationale Renewable Energy Lab (USA) 3 (37) University of Tsukuba (JP) 4 (37) 4 (33) 3 (33) University Complutense Madrid (ES) 5 (35) 3 (34) 2 (34) Monash University (AU) 2 (34) 5 (19) 1 (29) 1 (18) Huazhong University of Science and Technolgy (CN)

4 (14) 3 (11)

3.5.4 Gasification technologies We have searched USPTO and Web of Science datasets on technologies; the EPO search engine Espacenet does not support such extensive queries. Technologies considered include updraft, downdraft, fixed bed, fluid bed (or Winkler) and entrained flow gasification. In general, all technologies follow similar trends over time, although minor variations do occur. Apparently, all are steered by similar driving forces leading to similar dynamics. Downdraft and updraft technologies receive rather limited hits. Most indicators show major intensities prior to 1986-1990, followed by a dramatic decreased. For patents, this decrease roughly continued ever since – possibly due to the lack of US developments. In contrast, publications show a steady increase. Most recent levels (2008) match the levels of the early 1980s. Levels of publication on biomass technologies, which were hardly present prior to 1990, have taken off. Only updraft technology seems to be lagging behind. In publications USA, China and Japan dominate on most technologies, both in general as for biomass. For updraft and downdraft technologies also India is relevant. In patents USA and Germany dominate – as is to be expected given their dominance in patents.

66

3.5.5 Applications USPTO and Web of Science data have also been searched for a variety of applications relevant for gasification. These included the production of methanol, ammonia and hydrogen and the application of IGCC, Fischer-Tropsch for transport fuels and electricity production involving engines. Total hits differ widely per application. USPTO patents show most hits prior to 1990, but with a very strong peak for the period 2003-2006 that disappeared as fast as it came. This peak is present for all applications and for both biomass and non-biomass. Many patents have mentioned multiple applications. Major independent trends in the peak have been those for fuel cells and for Fischer-Tropsch fuels. In general, the share of biomass increased. Web of Science publications mainly score after 1990, with a strong increase after 2004, including publications on biomass. USA is leading on most applications, both in general as biomass related. China does well on Fischer-Tropsch and hydrogen, Japan on all applications – but both show limited hits on biomass. In addition two specific trends are worth mentioning. SNG has been developed in the 1970s and show a decrease in patents and publications ever since. However, publications suggest a renewed interest since 2005 of biomass based SNG. IGCC is only scoring since 1990, with a peak around 1998 and continued scoring since. Both trends are consistent with the literature. 3.6. Conclusions and discussion 3.6.1 Conclusions We included in this study both biomass and coal gasification, as the literature suggests that their developments are closely linked. As we have shown, both are subjected to similar driving forces: availability of feedstock (either coal or biomass), prices of fossil fuels and concerns regarding disruption of supply and global warming. As such some synergies can be expected. However, coal and biomass show rather different characteristics: biomass is more fibrous and reactive and has a lower density and ash fusion temperature (Knoef, 2005a). This has resulted in the selection of different gasification technologies and, as a result, the involvement of different manufacturers. On a practical and industrial level the linkage has been limited. Another relevant dimension is the scale of application. The development of entrained flow gasification of coal and the fluid bed gasification of biomass involve global players that serve a global market. However, markets for biomass gasification have been highly depending on niche applications and government support; therefore, the application of this technology still is confined to the national level. For small-scale fixed-bed gasifiers, in which a wide variety of manufacturers are involved, there is not

67

(yet) a global market. This technology seems to be driven more by local developments that are not very well represented by patents and scientific articles. In biomass gasification, two equally relevant periods can be distinguished. The first one, from the 1970s until about 1987, is a response to the oil crisis, mainly led by developments in the USA. In this period biomass gasification is part of a more general interest in gasification. Remarkably, Lurgi and Foster Wheeler have developed successful concepts outside the large government supported programs of this period. A second period of activity takes off in the late 1990s and focuses mainly on biomass gasification. Concerns regarding climate change are a major driver. Europe has been dominating these developments. After 2000, Japan and China rapidly are emerging as important players in this field. We characterized the developments in (biomass) gasification as a process in which the technology has been successful in a few niche markets, but in general still is in an early stage of development, see table 3.3. Coal gasification has not yet become a mainstream technology and is open for improvement. However, it has found significant application in the market on a commercial scale: for power production by IGCC in Europe, poly-generation in refineries and in a variety of applications in China. The market has selected a dominant technology and a few dominant suppliers have emerged. Biomass gasification is less mature. Applications for heat, co-combustion in coal plants and combined heat and power show limited market penetration and some are depending on government regulation and support. High-end applications like IGCC and transport fuels (Fischer-Tropsch, fuel cells) are considered very promising and received a lot of attention recently in research and demonstration. The required end-use technologies are apparently not the problem, since most have been applied in combination with coal gasification. Technological hurdles for biomass gasification mainly include scaling up, tar reduction and gas cleaning. Also the technology needs to become more economically competitive, as the case of IGCC showed. Downdraft gasifiers and atmospheric air blown circulating fluid bed gasifies are the preferred technologies. However, a much wider range of technologies is considered. Market leaders for fluid bed technology, Foster Wheeler and Ebara, have realized only a limited number of plants. Fixed bed technology found wider application in India and China, but their market share remains low. Based on this, the stage of development of biomass gasification can best be characterized as one of limited niche development. Biomass gasification has demonstrated great flexibility in adapting to the requirements of different niches. However, this also holds the disadvantage of lack of focus and built up of momentum. This is strengthened by the very diffuse profile of

68

the biomass industry that groups together the forestry industry, boiler manufacturers, farming and agriculture, etc. And only relatively recently, international trade in biomass resources has become part of the portfolio of market dealers of fuels – a requirement for entering the major energy markets (Faaij, 2007). 3.6.2 Discussion The use of science and technology indicators requires a critical reflection, as the data selection and the procedure for correcting for the growth in patenting and publishing rate are somewhat ambiguous – they are open for slightly different outcomes. However, the long term trends presented here are robust, as these are that clear that they would not be affected by small variations. In addition, also the use of multiple indicators strengthens the conclusions. Also several subjects can benefit from more detailing: the downdraft and updraft systems that are described rather limitedly by patents and publications; and the developments in China and India, which remain somewhat underexposed due to the gap in the literature. Biomass gasification has been profiled as being CO2-neutral, having a high potential, improving security of supply, being able to provide power, chemicals and fuels. The promise of advanced applications has been important in both periods of development. However, the technology has mainly found application for heat and power and only on a limited scale. To live up to the high expectations the application of biomass gasification will have to expand. Although the advantage over conventional coal technology is clear, this is much less the case when compared to the use of natural gas or biomass combustion. In the market these latter have been the preferred options in recent years: natural gas for power, ammonia and methanol production; and biomass combustion to convert biomass to power or heat. This illustrates the fact that different technologies co-evolve, where (changes in) one technology affects the (commercial) fitness of other technologies. Proponents of biomass gasification often tend to ignore this when they articulate the high expectations for biomass gasification. A successful application will also be largely depending on the drivers: if the reduction of greenhouse gas emissions has the highest priority, biomass combustion might be the easier and preferred option; if it is about producing renewable and climate neutral biofuels or chemicals biomass gasification becomes relevant; if it is about dealing with disruption of supply and rising fuel prices, fossil fuel gasification might hold good cards. The recent increase in interest for advanced options has been mainly present in research and development. For both IGCC and transport fuels applications conquering

69

a market segment will be very difficult, given its current status of high-risk high-(initial)-cost technology. The technology seems to be lacking the strong commitment that coal gasification does receive – or it receives these from countries (Sweden, Finland) which have significant lower RD&D budgets, industry and market potential compared to those supporting coal gasification (USA, China). This is especially relevant if a fast implementation is to be achieved. Such a development should focus on involving the USA, Germany and Japan (and possibly China), as patents suggest they hold the relevant industrial companies. Given its currents status and support as well as the status of competing technologies, we seriously doubt that biomass gasification will meet the high expectations. It seems to be overly optimistic and probably mainly serves the advocacy coalition of the technology in embracing the technology and building up momentum for further development. We consider a process of gradual development in niches with limited market diffusion much more likely for the upcoming years. Acknowledgement We would like to Menno van Geloven and Fernanda Neira D’Angelo for their help with data collection.

70

Chapter 4

Punctuated continuity: the technological trajectory of advanced

biomass gasifiers7

7 Kirkels, A.F. Punctuated continuity: the technological trajectory of advanced biomass gasifiers. Energy Policy 68 (2014) 170–182.

71

4.1 Introduction Over the past few years, energy from biomass has received ample interest, with special attention for biofuels, bio-refineries and the concept of a bio-based economy. Crucial to these developments is the technology of biomass gasification. Biomass gasification is the thermochemical breakdown of biomass – at high temperature and frequently also at high pressure. Input can be a diversity of biomass feedstock, although each requires a somewhat specialized technology. In the gasifier, the feedstock is converted to syngas (also called producer gas) that mainly consists of carbon monoxide and hydrogen. Clean syngas can subsequently be converted in several products: heat, power, chemicals and fuels – like methanol and Fischer-Tropsch diesel. Biomass gasifiers come in a variety of designs. Typically applied at smaller scale are the updraft and downdraft gasifiers. Simple updraft gasifiers produce syngas full of contaminants and are mainly applied in heat applications. Downdraft gasifiers produce cleaner syngas that is mainly applied for power production by engines. At larger scales there is a diversity of fluidized bed and entrained flow designs that, combined with extensive gas cleaning, can produce clean syngas for the production of biofuels, chemicals and power. We focus on the latter category of advanced gasifiers. Advanced gasification fitted social concerns well over the past decades. As such, it received a lot of interest and support (Kirkels and Verbong, 2011), but only became applied in a few research, development and demonstration (RD&D) niches. This raises questions from an innovation perspective. What is limiting the success of this technology? And what does this mean for its future application? Only recently, the long-term development of biomass gasifiers has been studied. Hellsmark (2010) takes a Technological Innovation Systems (TIS) perspective on European countries that dominated developments in biomass gasification - Sweden, Finland, Germany and Austria. Kirkels and Verbong (2011) provide an overview of global long term developments in biomass gasification based on multiple indicators and literature, showing that interest came in three distinct waves: in the early 1980s with a focus on methanol production; in the 1990s with a focus on power production by Integrated Gasification Combined Cycles (IGCC); and after 2000 with a focus on biofuels. In this chapter we will follow a complementary approach. We will reconstruct the technological development of advanced gasifiers and try to answer the following questions: 1) what has influenced the initial momentum and focus of the technological path; and 2) what impact did the developments in the technological path have on the success and failure of the technology. For the latter, we will address four sub-questions: a) what have been the dominant technologies and companies; b)

72

what have been dominant research themes and lessons learned; c) to what extent did this result in patterns of variation, selection and (dis)continuous technological paths over time; and finally d) how did this influence the promise and failure of the technology? We will conduct extensive literature study and construct an overview of demo plants in order to answer these questions for each of the three periods identified by Kirkels and Verbong. In the next sections we will introduce the concepts that we will be building upon, followed by the methodology. Next we will describe for each period the empirical results. And finally we will come to conclusions and discussion. 4.2 Conceptual framework We use an evolutionary perspective on technological change, starting from the work by Dosi (1982) and Nelson and Winter (1982). It is evolutionary in the sense that it includes processes of variation, selection and retention. Variation comes from early engineering efforts in RD&D, in a phase characterized by high uncertainties, little alignment and no lock-in. Sources of variation are firm-specific differences and bounded rationality. Selection mainly takes place upon market introduction: picking technologies that perform best in a given socio-economical context. And finally retention, or continued existence, is driven by processes of success and institutionalization, e.g. setting standards, sharing knowledge, etc. As our interest is in both continuous technological change as well as discontinuities, we will be drawing upon the notions of technological paradigms and technological trajectories by Dosi (1982). Dosi starts from a broad notion of technology as:

a set of pieces of knowledge, both directly ‘practical’ (related to concrete problems and devices) and ‘theoretical’ (but practically applicable although not necessarily already applied), know-how, methods, procedures, experience of success and failures and also, of course, physical devices and equipment.8

Based on this, he defined technological paradigms (or research programs) in analogy of Kuhn’s notion of scientific paradigms as:

an ‘outlook’, a set of procedures, a definition of the ‘relevant’ problems and of the specific knowledge related to their solution.9

According to Dosi, the technological paradigm embodies strong prescriptions on the directions of technological change to pursue, and those to neglect. The identification

8 Dosi, 1982, p151/152 9 Dosi, 1982, p148

73

of a technological paradigm relates to the generic tasks to which it is applied, the material technology it selects, the physical or chemical properties it exploits and the technological and economic dimensions and trade-offs it focusses upon. These define an idea of progress as the improvement of the trade-offs related to those dimensions. As such Dosi sees continuity in technological development, or development that adds up to a technological trajectory, as a pattern of normal problem solving within the technological paradigm to achieve progress; while discontinuities are associated with the emergence of a new paradigm. Some of the characteristics of a technological trajectory are: it consists of a series of small innovations (local incremental variations) that built upon each other and as such are cumulative; once a path has been selected and established, it shows a momentum of its own and as such it might be difficult to switch from one trajectory to an alternative one; there are complementarities among trajectories; and it is doubtful whether it is possible a priori to compare and assess the superiority of one technological path over another. Geels (2002) and Rip and Kemp (1998) have argued against such a narrow perspective on technological change, as this put too much emphasis on the embedding of routines in the minds of engineers. The outcome of the innovation process is also determined by other social groups like policy makers, users and scientists. More recent innovation theories, like the field of Transition Studies that includes theories of Strategic Niche Management and the Technological Innovation Systems, take this criticism into account and approach technological change as a quasi-evolutionary process (Faber et al., 2005; Raven, 2006). The process is called quasi-evolutionary, as the variation of technologies is not random. Researchers and RD&D departments do take into account both what they consider most promising technologies based on performance in lab or merely by expectations, as also the perceived future socio-economic context in which the technology will have to perform. These approaches put more emphasis on cognitive rules like goals, problem agendas and expectations. According to Geels and Raven (2006) expectations, visions and beliefs have the dynamic of self-fulfilling prophecies, because they guide research and development activities that work towards realizing them. While shared cognitive rules and expectations create stable trajectories of technological change, change in the direction of the technological trajectories depends on a change in the content of cognitive rules and expectations. Geels and Raven argue (2006) that it is at the level of communities or emerging fields that the emerging technological trajectory can be found – see figure 4.1. This level is building upon (series of) local projects, characterized by actors directly involved in those projects and local variability (local networks, project definitions, skills). The global network consists of actors who have some distance to the project. It refers to

74

an emerging field or community. It is characterized by abstract, generic knowledge shared within the community (theories, technical models, agendas, expectations, etc.). The translation of local outcomes into generic lessons and cognitive rules requires aggregation activities (e.g. standardization, model building) and the circulation of knowledge and people to enable comparison between local practices and formulation of generic lessons (e.g. by conferences, workshops, proceedings, journals, etc.). According to Geels and Raven, the interplay between local projects and the global community is important, as feedback mechanisms between both can explain dynamics in developments.

Figure 4.1 Technical trajectory carried by local projects [Geels and Raven, 2006] This notion of technological trajectories by Geels and Raven as directed by shared frames leaves ample room for variation as: (1) local projects leave room for local interpretation and adjustments; and (2) especially early on, rules are not shared by everyone and efforts might have not yet aligned. Only over time, with more alignment of rules and possible institutionalization, this will result in a more guided and stable search. This is amplified as the selection environment also tends to prefer existing solutions owing to economies of scale and lock-in effects (Geels and Raven, 2006). However, Bakker et al. (2012) and Raven (2006) argue that guided variation and pre-selection can also result in early appearance of dominant designs in RD&D. Following the global-local distinction by Geels and Raven, we conclude that studying the global level would not be sufficient for reconstructing the technological trajectory. There is an additional contribution to be found by studying local projects and demo plants – see also appendix B on the role and definition of pilot and demo plants. First, according to Dosi technological knowledge is much less well articulated than scientific knowledge. Existing physical devices embody the achievements in the development of a technology and as such are of special interest.

75

Second, constructing demo plants involves heterogeneous actor groups. These projects provide a place for early interaction between the variation (scientists and engineers) and selection environment (project owners, financers, buyers, suppliers, regulators, etcetera). At this level also these actors become visible and can be mapped and analyzed (Geels and Raven, 2006). Third, demonstration plants can also be seen as an early indicator of success: demo plants define a certain stage of maturity of the technology; multiple actors share a belief in the technology that justifies investing in it; and the applied technology is preferred over alternative existing technologies. And finally, biomass gasification plants are a configurational technology: they contain several components – biomass pretreatment, feeding, gasifier, gas cleaning, final conversion – that are interdependent. Problems in one component as well as the interaction between components will affect the overall performance. Fleck (1994) stresses that for such a configurational technology local demonstration projects are crucial to facilitate learning by trying. 4.3 Methodology Following a quasi-evolutionary perspective, we will try to reconstruct the technological trajectory of advanced biomass gasifiers based on literature study. As such we will focus on processes of variation and selection, continuity and discontinuity. We will follow the analytical distinction made by Geels and Raven (2006) between the community level and local projects. For the community level we mainly studied overview articles and status reports, to identify preferred technology, the state of the technology, research themes and expectations. For the local level we identified demo plants, presenting an overview in tables including site, manufacturer, technology, feedstock, size and status. Over the late 1970s and early 1980s exposure in scientific journals has been limited. However, this was a period of high RD&D intensity, as Kirkels and Verbong (2011) show. This is documented in the proceedings of various conferences that are at the basis of our study. These include the Bioenergy ’80 and ’84 conferences; the (bi)-annual European Biomass Conferences (1980-2010); the IEA International conferences on thermochemical biomass conversion (1985-2001); four VTT conferences on Power production from biomass (1993-2002); an EC international workshop and conference (1984, 1989); and two expert meetings on pyrolysis and gasification organized by PyNe and GasNet networks (1997, 2003).

76

Gasifier technology can be classified as an open technology: its development is influenced by developments in related technologies like small-scale gasifiers, coal gasification and thermochemical technologies like combustion and pyrolysis. We take this into account whenever literature suggests it was relevant. 4.4 1970s and 1980s: methanol as transport fuel The 1970s can be characterized by high oil prices and concerns regarding depletion and dependency on oil. From about 1979 onwards biomass gasification took off - as shown in chemical abstracts on biomass gasification (Overend, in Stassen and van Swaaij, 1982) and biomass-gasification related patents (Kirkels and Verbong, 2011). A variety of manufacturers and technologies were involved, e.g. see Hodam, Williams and Lesser (1982), Klass (1985), Reed (1980a) and Shand and Bridgwater (1984). Developments in advanced biomass gasification concentrated on methanol production to replace oil-based fuels. This focus required a new, more advanced generation of gasifiers. This received attention from the USA, Canada, Sweden, and the European Community - see table 4.1 for an overview of RD&D plants. Emphasis was given to the development of gasifiers. The subsequent conversion to methanol was considered commercially available. Both the USA and Europe chose for an exploratory strategy - both developed multiple technologies in parallel to find out what would work best. Overall, these development programs were technologically successful to the extent that no further exploration of concepts was required at the end of the 1980s and up-scaling to demo plants was considered (Beenackers and van Swaaij, 1983, 1986; Hogan, 1992; Klass, 1985, 1987; Miles and Miles, 1989; Stevens, 1994; Ström et al., 1985; Strub, 1984).

77

Table 4.1 Pilot projects on advanced biomass gasifiers for methanol production in the early 1980s Consortium / contractors

(country) Reactor type Method Capacity

[kg wood/hr] Pressure

[bar] Site

1 John Brown Wellman (UK)

• Double fluidized bed, circulating carrier for oxygen and heat.

• Double fluidized bed, circulating heat carrier

• Gasification with chemically bound oxygen; air blown

• Gasification with steam, separate combustion of char for heat

440-900 1 Smethwick (UK)

2 Lurgi (D) Circulating fluidized bed (CFB) Oxygen or oxygen / steam blown 320-450 1 Frankfurt (D) 3 Creusot-Loire (F)

ASCAB / Framatome / Stein Industrie Bubbling fluidized bed (BFB) Oxygen or oxygen / steam blown 100-320

2.500 1

7-30 Le Creusot (F) Clamecy (F)

4 AGIP Nucleare / Italenergie (I) Indirectly heated fluidized bed Pyrolysis, combustion part of product gas separately for heat; steam and oxygen blown

500-800 1 Toscana (I)

5 Omnifuel / Biosyn (CA) Bubbling fluidized bed Pressurized, oxygen (or air) blown 10.000

~4.000*

16-20

????

St. Juste-de-Bretenières (Quebec, CA) Degrad des Cannes (French Guyana)

6 Royal Institute of Technology / Studsvik Energiteknik (SE)

Bubbling fluidized bed, high temperature filter and catalytic reformer; MINO process

Oxygen and steam (or air) blown; add. fixed bed oxygen blown catalytic reformer

10-15 300-500

30 10-30

Stockholm (SE) Studsvik (SE)

7 Batelle-Columbus Laboratories (USA) Indirectly heated fluidized bed, dual bed Air blown 154-1400 0.2-1 Columbus, Ohio (USA) 8 Institute of Gas Technology (USA) Bubbling fluidized bed; RENUGAS process Pressurized, oxygen blown 160-500 6-24 Chicago, Illinois (USA) 9 SERI (now NREL) (USA) Downdraft

Upscaled SynGas 1985 Pressurized, oxygen blown

38 560-900

11 1

Golden, Colorado (USA)

10 University of Missouri-Rolla (USA) Indirectly heated fluidized bed, fire tube with heat exchanger

Air blown 20-375 1 Rolla, Missouri (USA)

1 -4 Beenackers and van Swaaij (1983, 1986), Grassi and Pirrwitz (1983), Kaltschmitt et al. (1998), Strub (1984) 5 Hogan (1992, 1993), Corté, in Kaltschmitt et al. (1998) 6 Beenackers and van Swaaij (1983), Blomkvist et al. (1983), Ström et al. (1985) 7-10 Klass (1987, p50-52), Schiefelbein (1989), Stevens (1994, p23-43) * The Omnifuel plant at Degrad des Cannes was rated at 30.000 ton/year of dry wood, which at 8000-8600 operating hours is equivalent with 3500-4000 kg per hour.

78

In the late 1970s and early 1980s it was considered that the production of methanol required the production of clean medium-caloric-value syngas. Such a plant was likely to be sized at tens of mega Watt: economies of scale and the large market dictated a large size, while the dispersed availability of biomass was limiting it. To achieve both clean syngas production and reasonable scale, two different concepts were explored: the directly-heated and the indirectly-heated gasifier. The directly-heated gasifiers were blown by pure oxygen. Some extra oxygen was supplied to the gasifier to combust some of the biomass, which provided the energy to keep the gasification process running. Experience with the Purox process and the SERI gasifier indicated that this might be feasible. Recommendations were made to support research on energy efficient small scale oxygen plants. Alternatively, the oxygen could be bought in from nearby large-scale oxygen facilities, which limited the number of suitable locations and increased operational costs. The second option used indirectly-heated gasification or pyrolytic gasification. In this process, the gasifier receives oxygen from steam as medium or from a bed material that chemically binds oxygen (e.g. in the John Brown process). The process is driven by externally generated heat, most often a second vessel in which part of the biomass is combusted. Heat exchange between both vessels is based on exchange of bed material that takes the role of heat accumulator. The structural complexity of indirectly-heated gasifiers adds to the investment cost for the gasifier, but operational costs are lowered as no oxygen needs to be bought in. These gasifiers, although not as well developed as oxygen gasifiers, promised higher efficiencies and lower overall costs and received significant support, especially in the USA (Beenackers and Maniatis, 1984; Reed, 1980a) Another focal point was pressurized gasification. No pressurized biomass gasifiers existed yet (Reed, 1980a) and the technology would likely be more complex. However, pressurized gasification seemed attractive as it could have a positive effect on the overall economics, as the subsequent methanol synthesis required pressurized operation anyway. While the advanced concepts were explored in RD&D, simpler, small-scale, fixed-bed gasifiers already found application in the market for low-end applications. Given the policy interest in high-end applications, a handful manufacturers of these updraft and downdraft gasifiers did consider entering this market segment by the use of oxygen, pressurized operation and the use of catalyst (Beenackers and van Swaaij, 1986; Reed, 1980a). However, the updraft technology produced polluted gas, while the

79

downdraft technology was hard to scale up. Therefore both were considered less suitable - a realization that grew over time. Already by 1983 interest started to decline, and by 1985 efforts came to a standstill: oil prices were low and government based RD&D funding stopped - which in turn resulted in less projects realized and reduced RD&D output. We found that for the second half of the 1980s construction of plants and their performance were underexposed in literature. This is probably due to the lack of success of these projects as well as the drop of interest by the relevant community. We will provide a reconstruction. In the 1985-1986 US solicitation for funds, the proposals for demonstrating a near-commercial gasifier were withdrawn (Bain et al., 2003). In Europe, the Clamency plant was built over the period 1986-1989, significantly scaling up the Creusot Loire technology and applying it under pressurized conditions. Test runs were made in 1990 and subsequently the plant was stopped, as the French government stopped supporting gasification research (Corté, in Kaltschmitt et al., 1998; Marcellin, in Bridgwater and Evans, 1993). Omnifuel found application in a plant in French Guyana. The plant was built in 1987, tested for a few days, but never operated. Several technical problems have been reported (Corté, in Kaltschmitt et al., 1998). In 1985 it was decided to build the Kemira Oy plant in Finland. It started operation in 1988 using Rheinsbraun’s HTW technology. Although this technology has not been included in table 4.1, as its development focused on lignite gasification (Harmsen, 2000), this specific plant was of relevance as it produced ammonia from peat. Its operation was proof of the technical feasibility, although several technical problems were reported, among others due to the heterogeneous quality of the peat. The plant was shut down in 1990, at least partially based on economics (Kaltschmitt et al., 1998; Koljonen et al., 1993). Both this overview and literature suggest that advanced gasifiers that were installed in the late 1980s had been initiated when energy prices were still high, and were frequently abandoned based on their techno-economic performance (Kaltschmitt et al., 1998; Kirkels and Verbong, 2011; Miles and Miles, 1989; Stevens, 1994). In addition, attention shifted to (fast) pyrolysis, especially in Europe (Diebold and Stevens, 1989; Grassi, 1988). At the community level, we see community formation by the upcoming of biomass conferences from 1980 onwards. And from the second half of the 1980s cooperation started in IEA tasks on bio-energy. Bridgwater (1990a) concludes, based on a worldwide database of activities in thermochemical conversion over this period, that mainly academics, governments, industrial research and development and

80

manufacturers were involved - in line with the findings of our overview. Industries involved in large-scale gasification were mostly research institutes and process industries, like pulp and paper industry supplier, boiler manufacturers and industry dedicated to gasification. There seems to have been only limited interest by automotive industry and a lack of involvement of methanol and petrochemical industry (Kliman, 1983; Klass, 1985). But what were the lessons learned? Both Reed (1980a) and Stevens (1994) indicate that applications in the late 1970s and early 1980s were to a significant extent skill based. As this did not provide a solid basis to develop advanced and properly working concepts, they became accompanied by more fundamental studies. As such, technologies turned out costlier and took longer to develop than initially anticipated. Also, biomass was not easy-to-handle-coal, but rather different from coal. It had different characteristics that required different technology and handling. And by the mid-1980s the emerging insight was that gas cleaning and especially tar was a persistent problem and catalysis was given more attention. Other issues included system integration, reaction mechanisms and kinetics, and broadening the scope of fuels considered (Baker et al., 1986; Beenackers and Maniatis, 1984; Beenackers and van Swaaij, 1990; Bridgwater, 1984; Diebold and Stevens, 1989; Miles and Miles, 1989; Stevens, 1994). To conclude, in the early 1980s biomass gasification initially received interest as it was perceived as a relative easy technology that could reduce oil dependency by producing methanol. This anticipated application influenced the technological focus on the production of clean medium-caloric-value syngas. Strategies to develop the technology paid special attention to variation – the parallel exploration of concepts – along two technological paths: oxygen-blown, directly-heated gasifiers and indirectly-heated gasifiers. To improve progress along these trajectories, ongoing efforts became accompanied by more fundamental studies of fuels and gasification. After the exploration of the concepts, interest seems to have broadened to the configurational aspects of the technology, with more attention for system integration and gas cleaning – which proved to be a persistent problem. In the second half of the 1980s the concepts were evaluated. In both Europe and the USA this resulted in pre-selecting the best technologies at that time. However, this selection was hardly effectuated as demo-projects were discontinued, mainly due to the drop in oil prices which on its turn resulted in a drop of interest.

81

4.5 1990s to 2004: IGCC for high-efficiency power generation By the end of the 1980s energy policy had refocused: nuclear energy and coal were considered no longer attractive, and attention for global warming strengthened commitments to renewables. Power from biomass was considered one of the more promising renewables at the short term (Kaltschmitt et al., 1998; Williams and Larson, 1989). Low-efficiency steam turbines had already found application for converting biomass to power. Now the higher-efficiency Integrated-Gasification Combined Cycle (IGCC) started to draw attention: it redefined potential power output and improved cost efficiency (Bain, 1993; Grassi, 1993; Johansson, 1993; U.S. DoE, 1992; Williams and Larson, 1993). The biomass IGCC option was already recognized in the early 1980s, see for example Beenackers and van Swaaij (1984) and Reed (1980a), but at that time failed to receive support. This changed by the late 1980s. Coal-based IGCC, including hot gas cleaning and pressurized circulating-fluidized-bed (CFB) combustion, received significant attention. Natural gas became more applied for power generation using combined-cycle technology. And biomass combustion by CFB found application over the 1980s, especially in the USA. This provided many relevant learning experiences with respect to biomass as fuel, CFB technology and system integration. Two leading designs by the companies Lurgi and Ahlstrom even found application in CFB gasifiers (Bain, 1993; Koornneef et al., 2007; U.S. DoE, 1992; Watson, 1997; Williams and Larson, 1989). Given the progress in all these areas, a short development time of biomass IGCC was to be expected (Hall, 1997; Larson et al., 1989). However, initially a much wider range of technologies to realize high-efficiency biomass-to-power conversion was brought under the attention: steam injected turbines (STIG), pyrolysis oil - either in turbines or in co-firing, indirectly-heated turbines, ceramic turbines, etc. Dominant views in both the USA and Europe saw IGCC as a promising concept, but not the first one to blossom (Bain, 1993; Grassi and Bridgwater, 1990, 1993; Larson et al, 1989). Only by the early 1990s expectations started to align and IGCC became the focal point of attention. At that time mainly Finland, Sweden and the USA were involved. They all concentrated on large-scale pressurized gasification (Rensfelt, 1991; Stevens, 1992; U.S. DoE, 1992). To realize IGCC application, several types of gasifiers were considered. Fixed-bed updraft technology looked promising in combination with hot gas cleaning, keeping tars in vapor phase and combusting them in the turbine (Larson et al., 1989; Williams and Larson, 1993). Others considered medium-caloric-value gasifiers to be better

82

suited for gas turbines (Solantausta and Beenackers, 1989; Stevens, 1994). But neither of these ideas received a lot of support. In contrast to this, air-blown fluidized bed technology was now considered proven and commercially available and became widely applied - see table 4.2. This included both the bubbling (BFB) and circulating (CFB) type (Bridgwater, 1993; Grassi and Bridgwater, 1990; Johansson, 1993; Larson et al., 1989). Note that table 4.2 represents both plants as well as plans - as only the Värnamo plant was operated for longer periods of time. Other plants were constructed, but were less successful, like the ARBRE, Hawaii and Vermont plant. Others never materialized, but contributed by system analysis, feasibility studies and design - like the ones in Brazil, Minnesota and North Holland.

83

Table 4.2 Plants and plans for biomass IGCC application, late 1980s till 2003 CFB Circulating Fuidized Bed; BFB Bubbling Fluidized Bed; IGCC Integrated Gasification Combined Cycle;

MWe MWelectricity; MWth MWthermal

Site (country) Consortium /

contractors Manufacturer Technology

(type, MWe, pressure, medium) Värnamo (SE) Sydkraft /

Ahlstrom Foster Wheeler - Bioflow technology

Pressurized CFB; 6 MWe + 9 MWth; 20-24 bar; hot gas clean up

Vega (SE) Eskilstuna

Vattenfall Tampella Power / Enviropower

Indirectly fired IGCC, pressurized BFB; 20-25 bar, air blown; 65 MWth + 60 MWe

Brazil WBP-SIGAME Bahia state

World Bank, Electrobas

TPS Atmospheric CFB; 30-32 MWe

Mariestad (SE) Gullspång Kraft TPS / VBB Atmospheric CFB; 16 MWe 17 MWsteam 9 MWth

(FI) Imatran Voima Oy (IVO) IVOSDIG process

Air blown, steam injected gas turbine no steam turbine

Borås (SE) Borås Energi AB TPS Atmospheric CFB, 70 MW IGCC

Gaspi, Tampere (FI) Enviropower Air blown pressurized IGCC; 15 -20 MWth; dolomite as bed catalyst, hot gas clean up

Hawaii (USA) HC&S - Paia mill

PICHTR / WEC IGT Renugas Pressurized BFB, air-oxygen blown; 10-21 bar; 5MWe

ARBRE (UK) - EC proj Yorkshire Eggb. Power Station

Arbre Energy Limited (AEL)

TPS Atmospheric CFB; 8 MWe; tars cracked in second CFB

Energy farm (I) - EC proj Di Casina, close to Pisa

Biolettrica (ENEL, Lurgi a.o.)

Initially Lurgi, later on Carbona

• Atmospheric CFB; 12 MWe; 1.4 bar • Pressurized BFB

Biocycle (DK) - EC proj Assens, Maribo Kotka (FI)

Elsam/Elkraft a.o. Enviropower / Carbona

Pressurized air blown CFB; 7 MWe + 7-8 MWth; 22bar; hot gas cleanup

North Holland (NL)

ENW TPS or Lurgi Atmospheric CFB; 30 MWe

Burlington, Vermont (USA) McNeil power station

Batelle / FERCO Batelle Atmospheric, indirectly heated; air/steam blown; 8-15 MWe

Summa (FI) transferred to Äänekoski (Fi)

Tampella (Kvaerner) Enso / Metsäliitto

Foster Wheeler IGCC demo plants; 60-70 MWe

New Bern mill (USA) Weyerhaeuser Batelle / FERCO 39 MWe

Minnesota Alfalfa project (USA)

MVAPC Carbona Pressurized CFB; air and steam blown; 20 bar; 75 MWe

84

Table 4.2 Plants and plans for biomass IGCC application, late 1980s till 2003 (continued) Sources: Beenackers and Maniatis (1998), Bridgwater (1993), Hellsmark (2010), Kaltschmitt and Bridgwater

(1997), Kaltschmitt et al. (1998), Knoef (2005a), Sipilä and Korhonen (1993a, 1996, 1998), Sipilä and Rossi (2002)

Status / year Capacity

[ton/d] Feedstock

Studies late 80s; cooperation 1991; 1993 start gasifier, 1995 turbine; 2000 stop, uneconomical

90-100 Wood

Initiated 1990; 1994 terminated - due to large investment cost and power balance countries

720 biomass

Planning 1991-1998; design completed; not constructed; plans terminated by 2004

~435 Eucalyptus

1991-1992 feasibility study + pre-project study

1991 some tests; mid 90s on hold; required significant investment support

Wet fuels: peat, biomass

Planned early 90s; failed to acquire funding; terminated mid 90s

Oad chips

Test facility. At least operated 1993-1995

80 Wood chips

1989 request; 1994 start project; 1998 finalized; technical difficulties; gas turbine never installed

40-100 bagasse

1993 start; 1998-2001 construction; (non)technical problems; 2003 terminated

~138 SRF (willow, poplar), later sludge added

1993 start; 1998 design completed, construction start; non-technical delay; 2003 terminated

200-336 SRF plantations -poplar; wood chips / residues

• 1993/94 start; DK cancelled - competition natural gas; transferred to Kotka

• 1997 abandoned due to closure wood supplier

~84 • Willow, wood, later eucalypt. plantation. • Wood residue

1993 start; 1998 plans terminated lack of biomass and not competitive

Demolition wood, park wood, RDF

1994 design; 1997 operating; 2002 mothballed; not economic; gas turbine never installed

180-300 Tree chips, residue wood

~1994-1996; repowering paper mill; not realized, too limited investment subsidy

Feasibility study, 1994-1995 Waste biomass mill

Studies/planning 1994-1999; terminated, unable to meet deadline

1000 Alfalfa stem material

85

From the table we can identify several leading manufacturers. Foster Wheeler developed pressurized gasification and was involved in the successful Värnamo plant. Based on this it was pre-selected as possible technology for the Brazilian plant. It also was participating in plans for up scaled plants in Finland. In the late-1990s, it supplied two large-scale biomass gasifiers for co-firing in coal plants. These latter are not included in the table – as we did not consider them high-end applications. But at the time Foster Wheeler was the only manufacturer that had several relatively large-scale biomass gasifiers operational. TPS participated in several atmospheric plants. It build upon an atmospheric CFB gasifier for lime kilns (Bridgwater, 1993). In addition, it drew upon experience with catalysts and gas cleaning from its research and the MINO process it worked upon in the 1980s, as was shown in table 4.1. It became selected in the Brazil project based on its lower specific investment costs, lower risk and better technological readiness - in which it out-competed Foster Wheeler (CHESF, 1995). Only the ARBRE plant was realized, but it closed almost immediately upon completion. IGT technology, involving pressurized gasification, found application in the Hawaii demo plant. It licensed its technology to Enviropower / Carbona. As such it was considered for application in several European plants, in the USA for the Minnesota project and in India. Two interlinked parameters determined the two trajectories of technological development: size and pressure. Pressurized operation, including pressurized gas cleaning, was more efficient as the gas turbine needed pressurized operation anyway - but it also was more complex, less proven and added investment costs. Larger sizes of 50-80 MWe, or even up to 100 MWe, were considered realistic for Scandinavian and US conditions. For smaller scales of 20-50 MWe atmospheric operation seemed to be the logic choice, with more established, but less efficient gas cleaning. This lower range was considered by Europe, including several Scandinavian municipalities. Over time there was an ongoing debate about the relevant ranges of scales. (Bridgwater, 1995; Larson et al, 1989; Maniatis et al., 1997; Palonen et al., 1996; Rösch et al., 1998; Salo and Keränen, 1996; Salo and Patel, 1997; Solantausta et al., 1995; U.S. DoE, 1992; Wilén and Kurkela, 1998). Demo plants played an important role in demonstrating the technology and learning, especially for system integration and unproven parts: pressurized gasification, gas turbines, and gas cleaning. Extensive gas cleaning was required to prevent corrosion and tar condensation (Beenackers and van Swaaij, 1990; Bridgwater, 1993; Johansson, 1993). Note that due to differences in circumstances (scale, sort and quality of biomass, components applied) these first demo plants were often one of a

86

kind and specifically designed for their task, frequently requiring on site modifications. However, the demo projects showed it was a bumpy road to take. Each plant encountered technical problems. Gas cleaning and tar condensation proved to be especially persistent problems, but also biomass feeding frequently resulted in problems. However, many scholars consider the non-technical problems as more substantial. These included, but were not limited to, acquiring proper biomass (e.g. Biocycle and North Holland), permitting (e.g. Energy Farm), knowledge and capabilities, and last but not least the high investment costs and overall economics that did not show the anticipated reduction due to learning curves (e.g. North Holland, Värnamo) (Bain et al., 2003; De Lange and Barbucci, 1998; Kaltschmitt et al., 1998; Maniatis, 1998; Salo, 1998; Piterou et al., 2008). In the late 1990s, related gasification technologies were explored. Black liquor is a highly-corrosive byproduct of the paper industry for which specific gasifiers were developed by MTCI, Chemrec and Weyerhauser. And in biomass gasifiers there was frequent experimenting with waste feedstock, like RDF and tires. Also specific waste- and plasma-gasifiers were under development. Finally, biomass gasification was applied for co-firing in existing pulverized-coal boilers - offering limited adaptation of the power plant and limited operational risk (Kaltschmitt et al., 1998; Maniatis, 1998). At the community level knowledge exchange was stimulated by new and increased networking activities, involving both experts and more heterogeneous actors. It showed in expert meetings and in IEA conferences and tasks. But also in the European Biomass Conference, that over the 1990s showed an increase in participants and visitors, and that was linked to an industry exhibition (Kirkels, 2012). The demonstration efforts required the involvement of more heterogeneous actors: turbine or feedstock suppliers and actors involved in catalysis, feeding systems, construction, financing, etcetera. We are under the impression that especially feedstock industry and utilities started to show involvement at the community level. In parallel with the demonstration efforts a significant research effort was made, both by manufacturers and research institutes. Main topics included, but were not limited to pressurized gasification, gas cleaning and catalysis, agro-fuels and socio-economic issues. In addition, modeling was needed to acquire a more fundamental and detailed understanding of what was going on in gasifiers in order to be able to prevent problems and optimize designs (Beenackers and Maniatis, 1993; Bridgwater, 1995; Connor et al., 1997; Elliot and Maggi, 1997; Kurkela et al, 1993; Maniatis, 1998; Maniatis et al., 1998).

87

Just after 2000, developments came to a halt: natural gas technology was preferred in the power sector due to its very low specific investment costs; biomass combustion and co-combustion were closer-to-market; wind become the renewable of choice; and power companies were no longer interested in innovative but high-cost alternatives as the electricity markets became liberalized (Jäger-Waldau and Ossenbrink, 2004; Kirkels, 2012). To conclude, the socio-economic context initially drove the interest in biomass gasification in this period. Technological development benefitted from the insights of the 1980s as well as progress in a variety of related fields. Based on all these, a short development time was expected, as the attention for demonstration plants also illustrates. The initial policy interest in high-efficiency power resulted in diversity of technologies that were brought forward in conferences – an indication of strong variation. Only over time these aligned: first concentrating on IGCC; and finally on air-blown fluid-bed gasifiers. Two trajectories were explored in parallel: atmospheric small-scale gasifiers and pressurized large-scale gasifiers. Government based demo-programs seem to have institutionalized and centered ongoing efforts, effectively resulting in pre-selection of technologies. The shift in attention between the 1980s and the 1990s, as well as the discontinuity in interest in the late 1980s, resulted also in a discontinuity with respect to technologies and manufacturers. Some of the leading technologies of the 80s, like the Creusot-Loire and Omnifuel technologies, no longer played a role. Others, like TPS and IGT, worked on other technologies or adapted their original design. Another leading manufacturer came up, Foster Wheeler, based on its experience with gasification for heat applications, and in fluidized bed biomass combustion. Several factors seem to have affected the lack of progress over this period, including amongst others issues of gas cleaning and socio-economic issues like lack of learning curve and economic performance. As a result of these, and of the large changes in the external context, attention shifted to biofuels after 2000. 4.6 After 2000: biofuels Since the 1990s and especially shortly after 2000, the policy- and market-interest in liquid biofuels increased considerably in Europe and the USA (Costello and Finnell, 2002; Hall, 1997). In line with this trend, RD&D on biomass gasification refocused after 2000 on two topics. The first was the production of ultra clean syngas, a basic requirement for whatever fuel there was to be produced. This also benefitted from

88

the increased interest in bio-hydrogen and fuel cells at that time. The second topic was the concept of a bio-refinery, defined in analogue with a petrol refinery: an industrial site where multiple products are co-produced in an energetically and economically optimized way (Costello and Finnell, 2002; Maniatis, 2001; Maniatis et al., 2003; Kirkels, 2012). Biomass gasification is considered to be a second generation biofuel technology that can use woody biomass. As such it has a larger feedstock base, larger greenhouse gas reduction potential and less interference with the food supply compared to first generation biofuels that are based on agricultural crops. The other second generation technology is based on biochemical conversion and was also pursued after 2000. Apparently, there was a strong diversity of technologies that gasification had to compete with over the past decade. Fuel production imposed different requirements on the technology than IGCC did: the synthesis gas needed to be ultra clean; the relative amounts of main components (hydrogen and carbon monoxide) became important for the subsequent chemical conversion; and production was considered most economical at significant larger scale. Just after 2000, some argued that entrained flow was the only technology that could meet these requirements (Hamelinck and Bain, 2003; Kavalov and Peteves, 2005). Others preferred BFB, as it had been widely demonstrated and tested under high temperature and pressure (Ciferno and Marano, 2002). Table 4.3 shows an overview of pilot and demo plants, both operational and under construction. It shows that ultimately four different technologies were explored in parallel: fluidized bed gasifiers, indirectly-heated gasifiers, entrained flow gasifiers and hybrid technologies. In general, both fluidized bed and entrained flow gasifiers operate under pressurized conditions using steam or oxygen as medium. Hybrid technologies encompass both a thermochemical and a biochemical step, like the technologies developed by Coskata, ZeaChem and Iowa State University (Bacovsky et al., 2010; E4tech, 2009; Hellsmark and Jacobsson, 2012). In addition, different sorts of feedstock inputs and output fuels have been considered.

89

Table 4.3 Gasification plants for biofuel (pilot & demo, operat. or under constr.) after 2000 FICFB Fast Internally Circulating Fluidized Bed; CFB Circulating Fluidized Bed; BFB Bubbling Fluidized Bed; CHP Combined Heat

and Power; SNG Synthetic Natural Gas; FT Fischer-Tropsch; DME Dimethyl Ether; MWth MWthermal; t/a ton/year

Site (country) Consortium Manufacturer Technology Güssing (AU) TU-Vienna / Repotec / CTU /

Paul Scherrer Institute Repotec FICFB, atm. steam blown, CHP / SNG / FT

Rya (SE) GoBiGas: Gothenburg Energy / Repotec / Metso

Repotec technology FICFB, SNG retrofittinig CFB combustion plants

Piteå (SE) - Smurfit Kappa Kraftliner Mill

Chemrec / Volvo / Preem / Total / Delphi

Chemrec Entrained flow, oxygen blown, pressurized; methanol / DME / hydrogen

Värnamo (SE) Chrisgas / Växjo VVBGC / Linnaeus University (TPS, Volvo)

Sydkraft / Foster Wheeler Upgraded to steam / oxygen blown CFB; pressurized 22 bar; DME / methanol / hydrogen / FT

Chicago (USA) Andritz / Carbona / UPM / IGT

Carbona Pressurized, directly heated, oxygen and steam blown BFB, FT

Varkaus (FI) NSE Biofuels Oy / Foster Wheeler / Stora Enso / VTT

Foster Wheeler CFB, oxygen and steam blown, atmospheric or pressurized, FT

Freiberg (DE) Choren / Daimler / VW / Shell

Choren Industries GmbH Carbo-V process; oxygen blown, entrained flow, 6 bar; 3 step: low temp, high temp, entrained flow; FT diesel

Karlsruhe (DE) KIT / Lurgi / Südchemie / VW (Future Energy)

Karlsruhe Institute of Technology (incl former FZK)

Bioliq-process: decentralized pyrolysis + centralized entrained gasification; oxygen blown, 80 bar; methanol to gasoline and diesel

Denver (USA) - K2A Soperton (USA)

Range Fuels, Inc. Indirectly heated, pressurized, entrained flow; devolatilization low temp + steam reforming high temp; ethanol / mixed alcohols

Gridley + Aberdeen (USA) Hawaii (USA) Livingston (USA)

Pearson Technology / ClearFuels Technology (Rentech owned)

Multi stage, steam blown, indirectly heated; entrained flow gasifier; FT production of ethanol and methanol

Sherbrook (CA) Westbury (CA) Edmonton (CA))

Enerkem (Biosyn technology)

Bubbling fluidized bed, pressurized 2-10 atm, air or oxygen enriched air and steam blown; ethanol and methanol

Temiscaming (CA) Tembec Chemical Group Ethanol

Warrenville (USA) Madison (USA)

Coskata Westinghouse Plasma

Hybrid technology: plasma gasifier + bioreactor; ethanol

Boardman (USA) Zeachem / GreenWood Resources

ZeaChem Ethanol, mixed alcohols. Hybrid technology:

Durham (USA)

Themo Recovery International

Atmospheric BFB, steam blown, indirect heating pulse enhanced technology; FT diesel

Boone (USA) Iowa State University BECON techn.

Thermal ballasted latent heat BFB gasifier; ethanol, FT liquids, biodiesel, pyrolysis oils

Clausthal-Zellerfeld (DE) CUTEC CFB, FT liquids

Wilton (USA) Startech / Future Fuels Plasma; atmospheric; hydrogen / methanol

a.o. Puerto Rico S4 (USA) InEnTec Plasma; atmospheric; oxygen or steam; hydrogen/methanol/ethanol

Petten (NL) ECN SNG

90

Table 4.3 Gasification plants for biofuel (continued) Sources: Bacovsky et al. (2010), Bain (2011), E4tech / NNFCC (2009), Kolb (2011), Waldheim (2012).

Status / year Capacity Feedstock 2002 demo plant 2005-2008 adaptation for SNG and FT

8 MWth Wood chips Wood residue

2006 committed, 2010 decision taken

32 MWth Wood pellets

1997 initiated 2004-2012 DME pilot project

5 MWth Black liquor lignocellulosic

2004-2010 Chrisgas project, rebuilding for fuel 2011 mothballed: difficulty forming alliances, attracting industrial funding, IPR problems

18 MWth Wood chips?

2005 rebuilding test plant 3-7 MWth? Forest residues wood pellets / chips

2006 trials 2010 operational demo

0.5 MWth 12 MWth

Forest residues

1998 pilot operational; 2003-2008 construction demo plant (beta) 2011 filed insolvency

1 MWth 45 MWth

Wood chips, divers Dry wood chips

2008 construction demo 2010- gas purification and synthesis added

5 MWth Straw / agricultural residues

2008 pilot operational, mixed alcohols 2007 demo construction; 2011 bankruptcy

1 MWth 25 MWth

Wood & wood waste

2002-2004 pilot Gridley; ??? facility Aberdon 2006- construction validation plant Hawaii 2008 operational pilot Livingston

0.8 + 5 MWth 8.5 MWth 1 MWth

Divers

2003 pilot operational 2010 gasifier operational 2010 start construction Edmonton (comm.)

0.8 MWth 7.5 MWth

100.00 t/ain

Lignocellulosic Electricity poles Municipal waste

2003 demo operational 13000 t/aout Spent sulphite liquor

2003 pilot operational 2009 demo operational

?? 0.2 MWth

Various Wood chips/nat. gas

2010 pilot under construction 4500 t/aout Lignocellulosic, sugar, wood, chips

92-03 multiple MTCI black liquor plants 2003 pilot operational

3500 t/aout

black liquor Forestry residues

2002 construction 2009 pilot operational

1 MWtht Grains, oil seeds, vegetable oil

2008 operational 0.4 MWth Straw, wood, residues

Various plants 3.8-7.5 odt/d; 2006- syngas program ~1-1.5 MWth Waste

2001- several in operation

0.8-1.5 MWth Waste

2004 lab scale; 2008 pilot plant 0.8 MWth Multiple, wood pellets

91

Both the technologies for waste gasification by plasma gasifiers and black liquor gasification had matured and their developments also started to consider fuel production. Especially black liquor developments seemed relevant for three reasons: first, the Chemrec efforts contributed significantly to the European efforts and promise (Hellsmark, 2010); second, recent reports indicate that Thermochem is also looking at the gasification of forest residues (E4tech, 2009); and third, scholars like Kavalov and Peteves (2005) point at the many similarities between the Choren technology and Chemrec’s black liquor technology. Until 2005, mainly test and trial plants were constructed concentrating on clean syngas production. After 2005, pilot and demo plants were erected to produce biofuels. Several scholars indicate that there was a strong drive to upscale (Bacovsky et al., 2010; E4tech, 2009; EBTP, 2013). In the food-versus-fuel debate around 2007/2008 biofuels became criticized. It strengthened attention for second generation biofuels that also became embedded in legislation and policies (Bacovsky et al., 2010; Hellsmark, 2010; Sorda et al., 2010). However, this did not reduce their risk, as the bankruptcy of Range Fuels (2011) and Choren (2011) show - two of the leading demonstration efforts at that time. This seems not to have halted developments, with new demonstration plants under way: UPM Stracel (France), Ajos (Finland, licensed Choren’s Carbo-V technology), BioTfuel (France) (EBTP, 2013). Of the companies involved in advanced gasifiers, several could build on experience with IGCC efforts over the 1990s - like Chemrec and actors around Värnamo. Other companies already had extensive experience with biomass gasification in other applications, like Repotec (combined heat and power) and Choren. The latter started its development in the 1990s. Building on former East-Germany’s knowledge and experience in coal- and lignite-gasification, it designed entrained-flow gasifiers to produce biofuels - which at that time did not receive a lot of enthusiasm (E4tech, 2009; Hellsmark, 2010; Kavalov and Peteves, 2005). And finally there are also relatively newcomers in the field. The gasification companies that dominated the 1990s developments had a hard time competing, as technologies diversified. TPS got involved in the CHRISGAS efforts to revitalize the Värnamo plant - that failed. TPS filed for bankruptcy and is no longer active (Hellsmark, 2010). Lurgi sold its biomass CFB technology to Envirotherm - that has not planned any projects since. Lurgi remained involved by its decentralized pyrolysis and syngas technology, as applied in the KIT process (E4tech, 2009; Hellsmark, 2010).

92

The international technology group Andritz took over Carbona. Carbona constructed a CHP plant in Skive, Denmark. In an effort to become a large FT biodiesel producer, global forestry company UPM started cooperation with Carbona (Bacovsky et al., 2010; E4tech, 2009; Hellsmark, 2010). Foster Wheeler initially refocused its efforts on co-combustion and gasification of household waste, but was reluctant to enter this market. By 2010, a NSE Biofuels Oy / Foster Wheeler demo plant went online to demonstrate the production of ultra clean syngas. However, by 2012 the project was abandoned when it was not awarded funding (EBTP, 2013; Hellsmark, 2010). At the community level we see a shift. Until 2000 the community was based on biomass gasification. Although this continued to some extent, after 2000 the focus was much more on the broader overarching theme of biofuels. It brought new dynamics, also with respect to community formation, e.g. by the upcoming of conferences and journals on biofuels and bio-refineries and special interest groups like the European Biomass Technology Platform. These were not specifically focused on biomass gasification. In the field of biomass gasification, we observed that utilities showed less interest, as could be expected. From the start, just after 2000, both automotive and oil industries started to show interest in and join the biomass gasification field (Hamelinck and Bain, 2003; Hellsmark, 2010). Main research efforts over this period were less well articulated. They seem to include some of the struggles of the past: providing a clean synthesis gas, becoming more economically competitive, up scaling and system integration - although a variety of other (non)-technical issues also received attention (Babu, 2005; Lightner, 2009; Maniatis et al., 2003; OBP, 2005; Rensfelt and Gobel, 2003). To conclude, although the attention for biofuels and IGCC were partly overlapping in time, they constitute a discontinuity. A strong indication can be found in the focus of RD&D efforts: in the 1990s demo plants received most attention, while after 2000 efforts concentrated on RD&D and subsequently on test plants. Also traditional companies had a hard time surviving, while new leading companies were up coming. The technological focus was mainly on fuel production and on ultra clean syngas. Developments were characterized by strong variation. We identified four different technological paths that were developed in parallel. But there also was variation in feedstock considered and biofuels produced. And gasification technology has to compete with a broader range of both first and second generation technologies.

93

4.7 Conclusions and discussion We set out to reconstruct the technological trajectory of advanced biomass gasifiers by providing an overview of demo plants (local level) and developments in research (global level), the results of which are summarized in table 4.4. Our first research question was what has influenced the initial momentum and focus of biomass gasifiers’ technological path. Over each period the strongest influence seems to have been the socio-economic challenges of the time. This provided the interest and the momentum to commit to developing the technology. But it also offered guidance to the engineering community that translated these socio-economic challenges in technological requirements, which we showed to be different over the periods considered.

Table 4.4 Characterization of technological development of advanced biomass gasifiers

1980s methanol as fuel 1990s-2004 IGCC for power Since 2000 biofuels • Methanol fuel as future

replacement for oil fuel • Requires clean syngas • Exploration gasifier concepts • Oxygen blown &

indirectly heated gasifiers • Issues: fundamentals fuels

and gasifiers; later on gas cleaning

• Leading companies/technologies: Creusot-Loire; Batelle-Columbus; IGT; Omnifuel; MINO

• IGCC as high-efficiency biomass-to-power technology to reduce greenhouse gas emissions

• Demonstration plants • Air blown BFB and CFB,

atmospheric and pressurized • Technical issues: pressurized gasification,

gas cleaning, IGCC, biomass feeding • Non-technical issues: economics,

bringing down costs by learning curve • Leading companies/technologies:

Foster Wheeler, TPS

• Biofuels to address global warming, oil dependency and agricultural policy

• Initial RD&D: clean syngas, bio-refineries

• Variety gasification trajectories: fluid bed, indirectly heated, entrained flow, hybrid technologies

• Diversity of feedstock and biofuels • Issues: clean syngas, economics,

upscaling, system integration • Later: demo plants, failures, new

efforts

Other influences on the technological trajectory have been technological progress and company specific differences, including preference for feedstock and scale and - more recently - type of biofuels to be produced. The influence of technological progress in gasification and related technological fields is most apparent in the late 1980s and the early 1990s upcoming of interest in IGCC. The influence of preferred feedstock is apparent in two ways: first, typically technologies are developed or optimized for a specific feedstock; second, waste and black liquor set specific requirements to the technology and developed in separate, although interrelated technological trajectories. The influence of scale showed in the 1990s, when both small-scale atmospheric gasifiers and large-scale pressurized

94

gasifiers were developed. Over the 1980s and after 2000, the requirement of (very) large scales was considered as a side condition for the technology to develop. The company specific differences, like feedstock and scale, were strengthened by policies. Although the power plant suppliers and process industry typically serve an international market, the overview shows a strong national focus of governments supporting national product champions. It indicates that national industries, resources, policies and actors might have been of significant interest. Our second research question was how developments within the technological path have influenced the success and failure of the technology. By the end of each period we could identify preferred manufacturers. However, over longer time periods this did not lead to dominance or a strong competitive advantage. Companies showed to be locked in a specific design and had a hard time adapting to paradigm shifts. It did not only result in a regular shake out of technologies and manufacturers – getting rid of losers that for some reason at that time are considered less fit – as could be expected by the evolutionary processes of variation and selection. We showed that also premium technologies had a hard time to survive. Each of these paradigm shifts had a distinctive different impact on the promise of the technology. The shift between the 1980s and 1990s seems not to have hold back the promise of the technology, as all countries involved at the time considered the technology ready for demonstration. We explain this by three reasons: the progress made in the field in the 1980s; the complexity of the technology, as air-blown gasifiers were considered less complicated than the medium-caloric-value gasifiers of the 1980s; and the progress made in related technological fields. Although Kirkels and Verbong (2011) portray the developments since the 1990s as one on continuous growing interest in gasification, our study comes with a more differentiated conclusion. The paradigm shift between the 1990s and 2000s came with a refocusing on the more complicated clean-syngas production of medium caloric value. This resulted in a shift in manufacturers involved and a large variation in technologies. The attention changed from demo plants in the 1990s towards research, developments and pilot plants after 2000. This indicates a setback with respect to the perceived maturity of the technology and the time-to-market. Our overview indicates how the community evolved. In the 1980s it was an upcoming community within the energy from biomass field. It mainly involved researchers, governments and institutions. Over the 1990s the community broadened and starting to involve more heterogeneous actors, including some utilities and feedstock companies. After 2000 we saw the upcoming interest by automotive and

95

petrochemical industry. Over this period, the biomass gasification community became embedded in a much broader community working on biofuels. With respect to dominant research themes and lessons learned, our study shows that initially biomass gasification was perceived to be a new technology that would not be hard to master. When actors became aware that both the feedstock and the technology were more complex, more fundamental studies started to be conducted. Also gas cleaning turned out to be most relevant and not easy to handle. With the attention shifting to demo plants, initially in the 1990s and later after 2000, other aspects started to receive attention: biomass feeding, system integration, up scaling and improving economics – all of which we perceive as quite typical for this stage of plant development, as compared to, for example, the development of fluid bed combustion (Koornneef et al., 2007; Watson, 1997). Variation can be important for technological development, as it is a way of overcoming uncertainty, diversify learning experiences and as such contribute to a more robust technological development. Our study shows that especially early on in each period, characterized by the upcoming social and political interest in the technology, a wide variety of ideas were plugged. Only over time these aligned when government applied pre-selection of technologies to support and pursue within their RD&D programs. But also within these programs there was large technological variation. One source of this variation seems to be firm specific differences facilitated by the national focus of innovation policies, which resulted in significant variation at the global field level. Another source is to be found in the policies pursued by the USA and the European Commission. Both are characterized by relatively large budgets and explicitly supported multiple technologies in parallel. Especially after 2000 we see large variation that can be explained by multiple factors: the market that does not have a clear preference on what biofuels to produce; the technology of the 1990s (air-blown fluid bed) that probably was most mature at the time, but there was no consensus whether it would be the appropriate technology to produce clean syngas on large scale; and the maturing over time of other gasification technologies, like entrained flow and hybrid technologies. As such, we conclude that lack of progress in the field of gasification is unlikely to be caused by lack of variation. On the contrary, for the period after 2000 the level of variation raises concerns of too much variation, which would result in lack of focus and scattered funding. However, a normative measure of what level of variation is suitable is lacking. So far, from literature we did not pick up any major signals that the level of variation was problematic.

96

A final aspect that draws our attention is the long development time of several decades. In addition, according to IEA (2011) it still requires decades to prove the gasification-to-biofuels technology and come to a reasonable diffusion in the market. Long development times are quite typical for energy and process plants and more specific for gasification-related technologies, as these are complex and systemic technologies that are to a significant extent knowledge driven (Martin, 1996; Watson, 1997; Harmsen, 2000). Our study suggests that these long lead times are hard to manage or even survive by companies depending on an immature technology and government support – especially given the dynamic socio-economic context. Innovation literature suggests reducing uncertainties by variation (robust technological development) and by institutionalizing efforts e.g. by providing stable frameworks and consistent policies. In our case of biomass gasification both were applied: variation we already discussed in detail; while institutionalization was provided by RD&D programs and more recently biofuel legislation. However, our case also shows the limitations of both approaches to overcome the impact of socio-economic dynamics over a long time frame of decades. This case touches upon the inherent uncertainties of these long term innovation processes. It is a warning, especially for policy makers and innovation scientists, against belief in easy solutions and too optimistic views of steering long term innovation processes. Acknowledgement I would like to thank Geert Verbong and two anonymous reviewers for providing me with constructive feedback and suggestions for improvement.

97

98

Chapter 5

Biomass boom or bubble? Expectations and dynamics

in biomass gasification10

10 Based on Kirkels, A. (2016). Biomass boom or bubble? A longitudinal study on expectation dynamics. Technological Forecasting & Social Change, 83-96.

99

5.1 Introduction Biomass is the most applied renewable at this moment and a further increase is believed to be possible (Faaij, 2007; Hoogwijk, 2004; OECD/IEA, 2014; World Energy Council, 2013). Biomass can be converted to energy using advanced gasification technology. This is considered an efficient and clean conversion technology, one that facilitates large-scale high-end applications. The technology started to receive attention in the late-1970s, driven by expectations on the high potential of the technology – although working concepts did not yet exist. Over time it continued to receive attention over different waves of interest. Kirkels (2014) showed that these waves of interest relate to three periods. Over each period the focus was on a different end-use application: during the first period (late 1970s till mid 1980s) the main focus was on methanol production; during the second period (1990-2003) the focus was on Integrated Gasification Combined Cycles (IGCC) for high-efficiency power production; and from 2003 onwards the focus changed to the production of a variety of biofuels. High expectations regarding these applications were crucial for the momentum and direction in this field. A body of literature studies the role of expectations in innovation processes and innovation dynamics. Early research, development and demonstration (RD&D) of technologies tends to be strongly driven by expectations of the future performance of the technology, as in this stage the actual technological and economic performance of technologies is often poor. Based on these characteristics Mokyr (1990) characterized these technologies as ‘hopeful monstrosities’. The expectations can take the form of visions, agenda’s, strategies and beliefs (Van Lente, 1993). In the sociology of expectations it is not about whether these expectations are right or wrong – something that can only be assessed in retrospective. It is rather about the performative role of expectations, the impact that expectations have on the innovation process by providing direction, coordination and legitimacy (Dignum, 2013; Van Lente, 2012). The dynamics of these expectations caught the interest of scholars like Fenn and Raskino (2008), Ruef and Markard (2010) and Van Lente, Spitters and Peine (2013). In general they are building upon the concept of a hype-disappointment cycle: the fast upcoming and downfall over time of attention for and expectations of new technologies in a stage before market diffusion, also referred to as waves of interest - short periods of large momentum. However, scholarly interest has broadened to alternative patterns in the dynamics of expectations as well as to a better understanding of this dynamics. Cases in literature cover, amongst others, the field of

100

emerging sustainable technologies in which governments play a large role, like biofuels, hydrogen and fuel cells. Our research question is what were expectations of advanced biomass gasifiers and how did these influence developments of the technology? The content of expectations is best constructed by a qualitative approach. It will allow us to answer questions like who are the actors involved; why do they support this technology; and are their expectations positive or negative. The dynamics of expectations can benefit from a quantitative approach. We opt for the best of both worlds and will apply a mixed-methods approach. We argue that a mixed method approach will offer additional benefits beyond this complementarity: triangulation of trends and the possibility to overcome specific bias related to each method. This is especially important for the challenging case of advanced biomass gasifiers: the technology developed as the result of international efforts (broad geographic focus), over four decades (long term, longitudinal study), refocusing on different end-use applications, while its development was interrelated with that of other technologies (Kirkels, 2014). As such it is substantially different from previous cases studied in this field that have a much more narrow focus in one or more of these dimensions. It raises a methodological question: how do both qualitative and quantitative approaches handle these issues, and what are the benefits of a mixed-method approach? In the next section we will discuss the conceptual framework drawing upon the literature on dynamics in expectations. In section 3 we introduce the methodology: the concept of mixed-methods as well as the methodological choices we made. In section 4 we discuss quantitative trends in science and technology indicators. In sections 5-7 we discuss expectation dynamics as reconstructed by the mixed method approach, each covering a promise of a specific end-use application and related to a specific period. And finally in section 8 we will come to conclusions. 5.2 Theoretical background 5.2.1 Gartner’s cycle The dynamics of expectations of emerging technologies has attracted a lot of attention, often drawing upon the hype-disappointment cycle as introduced by Gartner Consultancy Group in the late 1990s, see figure 5.1. The hype-disappointment cycle is building upon the observation that many technologies temporarily succeed in capturing attention in a pre-mature stage during the Peak of Inflated Expectations, before the technology shows diffusion in the market at the

101

Plateau of Productivity. Therefore the curve is formed by conjunction of first the expectations of (the progress in) an innovation and second a maturity and diffusion related curve (Fenn and Raskino, 2008). According to Fenn and Raskino, ‘expectation’ in the curve should reflect market’s assessment of a technologies future expected value, the sentiment of potential and actual adopters of the technology and the shifting pressures surrounding investment decisions. For sexy products in the consumer world, like some ICT gadgets, hype cycles might be well visible in mass media. For business-to-business technologies, hypes might be better traceable in industry magazines or at conferences.

Figure 5.1 Gartner’s hype-disappointment cycle (Based on Fenn and Raskino, 2008)

Gartner’s concept of hype-disappointment cycles has been criticized on both methodological and empirical grounds. Steinert and Leifer (2010) draw attention to the curious merging in one curve of expectations on the one hand and technological maturity and diffusion on the other. Also the use and interpretation of the model has been criticized. Van Lente argues that the working of the Gartner group is based on an implicit perspective that “others may be victims of the game of expectations, but the analyst, who studies them, understands the game and is smarter than others” (2012, p776). Steinert and Leifer (2010) show that predictions by the Gartner group based on the cycle turned out to be imprecise, inconsistent and subjective. It shows that the implicit perspective as applied by Gartner does not hold. Borup, Brown, Konrad and van Lente (2006) add the principal point that the model is too general and does not provide room for variation and unpredictability - over different technologies, socio-economic contexts and actors involved. Strictly following the cycle does re-introduce an unwanted understanding of a technology’s path dependency, one that is deterministic.

Time

Expe

ctat

ions

Peak of inflated expectations

Technology trigger

Slope of enlightenment

Trough of disillusionment

Plateau of productivity

102

5.2.2 Hype-disappointment pattern Among scholars these criticisms have been embraced and the full Gartner cycle received little further attention. However, the presence of hype-disappointment patterns is widely accepted, see for example Alkemade and Suurs (2012), Bakker and Budde (2012), Borup et al. (2006), Brown (2003), Jun (2012), Konrad (2006), Ruef and Markard (2010), and Van Lente et al. (2013). Some scholars start from exposure in media, for which patterns over time can easily and objectively be assessed. Several scholars have questioned whether these hypes (in practice or when studied by scholars due to applied methodology) actually reflect expectations or exposure, especially as media exposure itself tends to follow hype dynamics (issue-attention cycles as Downs calls it in his 1972 paper); and what sources (media, actor groups) are valid for reconstructing expectations (Järvenpää and Mäkinene, 2008; Jun, 2012; Konrad, 2006; Ruef and Markard, 2010; Steinert and Leifer, 2010). Järvenpää and Mäkinen (2008) argue in this context for the value of conferences and conference attendees: these can be related to hype cycles as it measures both excitement and visibility among important interest groups. Others include qualitative sources (e.g. literature, interviews) or science and technology indicators. Often a combination of different sources is applied, depending on the goal of the study and in order to secure coverage of different aspects of technological development (e.g. related to different stages of maturity or different discourse spheres) – see for example Budde (2015), Jun (2012) and Ruef and Markard (2010). Both hype and disappointment are driven by several forces: the novelty preference, the bandwagon effect, competitive threat and access to first hand learning (Fenn and Raskino, 2008; Van Lente et al., 2013). The hype-disappointment pattern is further strengthened by the ambivalent nature of expectations, as Van Lente and Rip (1998) point out: stating that a technology or field is strong, that it will break through or provide a good rate of return after a certain amount of work and investment, implies its weakness – without the work, investment and time the technology or field will wither away. In periods of high expectations the focus is on the positive and high hopes; but in a period of disappointment the new technology might become criticized on these grounds (Konrad, 2006). The hype constitutes both an opportunity as well as a pitfall, as it is likely to be followed by disappointment that can have a detrimental effect on the credibility of specific actors or an innovation field (Konrad, 2006). Several scholars stress in this context that it is what happens after the hype what is important: have the efforts during the hype provided enough learning, involved enough heterogeneous actors and institutionalized developments, so that innovation activities can continue in periods of limited interest (Borup et al., 2006; Geels, Pieters and Snelders, 2007; Ruef and Markard, 2010; Van Lente et al., 2013).

103

Although the pattern of hype-disappointment by now serves as a reference in the field, it has been recognized that alternative patterns exist. Van Lente et al. (2013) look at micro differences in patterns: the shape of the peak, the depth of the through and the length of the hype. Others looked at macro differences: technologies that fail; technologies that are merged with other technologies; several subsequent hype-disappointment cycles; contested embedding, in which both resistance and enthusiasm co-exist; and alternating waves of enthusiasm and resistance (Fenn and Raskino, 2008; Konrad, 2006; Steinert and Leifer, 2010; Van Lente et al., 2013) And finally, several scholars indicate the theoretical option of a more gradual technological development, in order to avoid the negative effects of hype-disappointment. However, it is not clear to what extent hype patterns can be ‘prevented’. And if so, whether this will not result in the loss of the beneficial aspects of the hype, like constructing legitimacy and acquiring support (Bakker and Budde, 2012; Geels and Raven, 2006). 5.2.3 Expectations of technology Van Lente (1993, p182/183) differentiated between micro, meso and macro expectations of technologies. Other scholars have been building upon this distinction, e.g. Ruef and Markard (2010) adapted it to the analysis of hypes; and Budde, Alkemade and Weber (2012) adapted it to align with the multiple level perspective (MLP) of transition theory. Crucial in each of these perspectives is that the interaction between levels explains part of the dynamics (Budde, 2015; Geels and Raven, 2006; Konrad, 2006; Ruef and Markard, 2010; van Lente, 1993; Van Lente et al., 2013). We will follow the distinction made by van Lente, as it is well established in literature and suits our analytical needs. The micro level is defined at the level of the artifact and applications of research, often involving specialists. Micro level expectations are rather specific and might shape the local agenda. Achievements in micro level projects feedback to expectations on the meso level, shape the field agenda with issues that need attention or demand action and might provide the trigger for hype or disappointment. Meso level expectations are shared among a community or field. They tend to express functions that the technology presumably will fulfill: the general direction of the field and opportunities the field offers, in our case the field of biomass gasification. This is the level at which hype-disappointment cycles are located, as they relate to the general expectations, visibility and maturity of technologies at the field level. Macro expectations are general, broad and diffuse. These might encompass statements about technology as a whole in relation to societal trends, e.g. with respect to large scale application of energy-from-biomass for advanced applications, or in relation to oil prices or concerns regarding global

104

warming. Typically they are ideographs that are not clearly defined and cannot be falsified. Positive expectations at macro level can provide legitimacy, direction and resources for the field or for specific projects at the micro level. 5.2.4 Actor involvement Actors take a crucial position in understanding expectation dynamics: they are the ones that hold and voice expectations; they act upon these expectations; they take decisions on RD&D of technologies; and as such cause dynamics in expectations and developments. From an analytical perspective it is important to realize that expectations of actors can be diffuse, as they can hold multiple expectations regarding a technology: e.g. expectations with respect to short term and long term performance, development costs, profitability, market size, social support, etcetera. Different scholars have argued that it is important to differentiate between actors involved: e.g. policy makers, research groups, investors, end-users, and etcetera. There can be differences between actor groups in support for and opposition against a technology (Brown, 2003; Budde et al., 2012; Jun, 2012). In addition, different actors respond differently to hypes and disappointment and to expectations at different levels. Konrad, Markard, Ruef and Truffer (2012) conclude that research institutes react more slowly to disappointment than industry actors. According to Bakker, van Lente and Meeus (2011 p154) “Scientists tend to maintain positive expectations as long as these provide them with a mandate and with funding to continue their research activities. For industry it seems that meeting expectations is more vital and hence they tend to compare and test expectations.” And industry tends to have a strong focus at expectations at field level and non-technical expectations including on regulations and customer demand, as Budde et al. (2012) showed for the German hydrogen hype. Contrasting this, for policy actors expectations on the macro level are most crucial (Budde et al., 2012; Dignum, 2013). In this domain short lived attention and hype-disappointment patterns are most important. Verbong, Geels and Raven (2008) and Dignum (2013) explain this by the mechanism that policy actors face credibility pressures with regard to societal problems, such as sustainability. Therefore policy actors need to make promises about how they will solve these problems; they desire to put forward a policy connected to their names; while budgets constrain the support for multiple technologies. It frequently results in shifts of preference for renewable energy technologies (Verbong et al., 2008), as well as for energy-from-biomass technologies (Kirkels, 2012).

105

5.3 Methodology and sources From the overview of literature we conclude that dynamics in expectations are to be understood as interactions between expectations and developments at micro, meso and macro levels. It is the result of interaction between multiple actors, for which it is important to differentiate between types of actors involved. The dynamic might result in hype-disappointment patterns, however cases show a much wider variety of patterns. The pattern offers limited guidance. However, it can be considered an indication of mechanisms and phenomena that are relevant (e.g. disappointment, competitive threat, increasing returns to scale). The case of advanced biomass gasifiers requires a longitudinal study of three different promises that relate to both advanced gasifiers as well as specific application domains. These shifts in foreseen applications are not uncommon for emerging technologies that are the generic basis for applications (Borup et al. 2006; Ruef and Markard, 2010; Verbong et al., 2008). However, such a reorientation can involve an adaptation or discontinuation of specific activities, technologies considered and actors involved – as also Kirkels (2014) showed for advanced biomass gasifiers. We assume that actors might position a new promise in relation to past developments: referring to technological progress made, or emphasizing differences between earlier and current promises, emphasizing why it would work this time. This would provide credibility and direction to the field. In order to understand dynamics in the field of biomass gasification, we reconstruct the expectations as well as the dynamics in innovation activities in the field. We focus on capturing developments in the USA and EU, as these were dominating the field (Kirkels, 2014; Kirkels and Verbong, 2011). We do so by applying an inside-out perspective, following the actors involved in the field. It will allow us to capture the inner dynamics and is in line with earlier literature on this case. We consider competing technologies as far as they show up as being relevant. We will apply a mixed-methods approach, reconstructing dynamics from science and technology indicators in combination with a reconstruction of expectations and developments by literature study and by drawing upon earlier studies on this case. For a more extensive discussion of the merits of mixed-methods research see Johnson and Onwuegbuzie (2004), Johnson, Onwuegbuzie and Turner (2007) and Sandelowski (2013). A variety of RD&D indicators exist. Strengths are that there are specific indicators for different stages of technological maturity and for different discourse spheres; they represent a large number of activities which allows for a bottom up reconstruction of developments; they can reveal trends as far as data are available over longer periods;

106

and that they are relatively easy to reproduce (Järvenpää and Mäkinen, 2008; Jun, 2012; Watts and Porter, 1997). The validity of the indicators is dependent on their specificity (do they capture the trends of the technology that we are interested in and only these trends); and as far as the practice of patenting, publishing or doing research did not change over time (e.g. changes in scope, intensity and what these indicators represent). We reconstructed four indicators. The first one is on governmental RD&D budgets, as developments showed large governmental involvement. The second one is the European Biomass Conference – both contributions as well as attendees, given the relevance of conferences for business-to-business innovations (Fenn and Raskino, 2008; Järvenpää and Mäkinen, 2008) and for this particular field (Kirkels, 2014). Both RD&D budgets and conference contributions relate to developments over different stages of technological maturity. Our third indicator is Web of Science publications as measure of academic interest. And finally our fourth indicator is Derwent patents, as indicator for the development stage of technologies depicting interest by industry. We will reflect on the strengths and weaknesses of these indicators in the results section. We did not include media exposure, as a quick scan of a few newspapers showed limited exposure – except for the recent interest in biofuels and the public food versus fuel debate around 2007. In addition, a valid reconstruction of global coverage by newspapers would be hard to do. The use of indicators comes with several limitations: it will be hard to capture all dimensions of expectations as identified in the conceptual framework; indicators don’t show interactions; they do not make expectations explicit; nor do they differentiate between (or sometimes do not represent) proponents and opponents, or between success and failure, or more and less important achievements/quality. These limitations can be overcome by qualitative research, in our case literature study, as it allows for a more detailed, contextualized and appreciative reconstruction of developments. Typical disadvantages of such a method are subjectivity and improper generalizations. Two main sources of subjectivity are the scholar that is conducting the research and the material he is drawing upon. The scholar makes choices, provides direction, and comes to an interpretation and representation of results, which seem incompatible with the concepts of objectivity and reproducibility. However, also in quantitative approaches often decisions need to be made (e.g. about proper key words for indicator data) and therefore these involve some subjectivity as well; and approach and interpretation of both quantitative and qualitative methods are bound by codes of scientific conduct (transparency, good referencing, nuanced conclusions, be critical and reflexive). The second source of subjectivity originates from the material studied: claims made in literature for which it is not always clear what they are based upon or under what circumstances they will hold. In general

107

these statements will not hold for the whole field or cover global developments, but will be restricted to limited time periods, to specific countries, research programs, manufacturers or conferences. Improper generalization follows from ignoring these limitations. This can be reduced by making the relevant boundaries explicit – see the earlier work on this case by Kirkels (2012, 2015) and Kirkels and Verbong (2011); by focusing on overview studies and literature covering key conferences, dominant countries and dominant companies as these relate best to dominant developments; and by cross-examination of the results with the outcome of the indicators and by the results as published in this prior literature. For our case especially conference proceedings are relevant: they extensively covered developments since the 1980s; they brought together policy makers, scientists and industry; while exposure in academic journals was limited over the earlier period. Conferences that are at the basis of our study include the Bioenergy ’80 and ’84 conferences; the (bi)-annual European Biomass Conferences (1980-2010); the IEA International conferences on thermochemical biomass conversion (1985-2001); four VTT conferences on Power production from biomass (1993-2002); an EC international workshop and conference (1984, 1989); and two expert meetings on pyrolysis and gasification organized by PyNe and GasNet networks (1997, 2003). In addition we also studied broader scientific literature and reports. In the next section we present the results of the indicators (trends 1976-2014) and their methodological choices and limitations. In the subsequent sections, sections 5 till 7, we reconstruct the expectations and field level dynamics for the three different promises (respectively methanol, IGCC and biofuels). Each of these sections is structured in the same way: 1) a reconstruction of the underlying rationale in order to better understand the legitimacy of and support for the promise; 2) the actors involved in building the promise; 3) a reconstruction of dynamics in expectations and developments; and 4) intermediate conclusions. Micro and macro expectations are only included whenever they started to effect (positively or negatively) the expectations on meso level. The reconstruction is the result of the applied mixed-method approach, so the findings on indicators feed into these reconstructions.

108

5.4 Indicator trends 5.4.1 Governmental funding RD&D Governmental funding of RD&D of IEA countries was of vital importance for the developments in the field. Trends in funding are reconstructed in figure 5.2. Included are data for USA, Canada, Japan and EU-15 – which according to Kirkels and Verbong (2011) took a leading role in the developments in biomass gasification. The figure shows that in 1976 hardly any budgets were available for energy from biomass. Funding increased fast, resulting in a peak in1983 at 280 million US$. By 1986 this had already dropped to 180 million US$, coming to a low in 1993 at 108 million US$. Especially after 2003 funding starts to increase again, with a sharp increase from 2007 onwards. This resulted in the high peak at 1849 million US$ in 2009 that dominates the figure, amongst others due to the American Recovery and Reinvestment Act of 2009 (OECD/IEA, 2015).

Figure 5.2 Total governmental Research, Development and Demonstration (RD&D) budgets on energy from biomass in 2013 million US$. Included are data for USA, Canada, Japan and

EU-15. (Data: OECD/IEA, 2015 - indicator 34 biofuels incl. liquids, solids and biogases)

Main limitation of this dataset is the general scope: data represent different stages of technological maturity and thereby different discourse spheres; and data are only available for the general category of energy from biomass. For recent years more detailed data are available, although this is fragmented. Trends suggest that since 2009 liquid biofuel received most funding. For the production of biogases, which is a different category and is including thermochemical conversion technologies like gasification, budgets showed a significant rise since 2007 and a strong increase over 2011-2013. Data on early years, just after 1976, do not cover all countries; however, countries that at that time dominated the field (USA, Canada, Sweden, UK, and France – according to Kirkels (2012) and Kirkels and Verbong (2011)) are largely covered.

0

400

800

1200

1600

2000

millionUS$ RD&D budgets energy from biomass

109

5.4.2 European biomass conference The European Biomass Conference (EBC) is a large international conference supported by the European Commission. To some extent it also covers broader developments, e.g. leading demo projects or US policy. It has been held annually or bi-annually. It encompasses the whole field of energy from biomass, from biomass production and harvesting, to processing and conversion, to final application, market formation, environmental impacts and policy. Gasification was the technology most published upon. Initially it was one of only few conferences on energy from biomass. From the start it covered over one hundred contributions and was visited by over 500 participants. As such it was of major relevance for the field. It is also the only conference that was held for more than three decades that we are aware of.

Figure 5.3 Trends in European Biomass Conference. Participants as reported in proceedings.

Articles identified by reading title and abstract, looking for gasification and syngas synonyms. Maxima have been indexed on 1.

(Data: European Biomass Conference, 1980-2010)

Table 5.1 Contributions of European Biomass Conference on biomass gasification and end-use applications. End-use application by key words in title: methanol; IGCC, BIG CC, combined cycle, gas turbine; biofuel, fuel, synfuel, methanol, ethanol, diesel, Fischer-Tropsch, BTL, DME. Only unique titles are counted (no double counts if multiple key words are used in one title). Only relevant titles are counted, e.g. not titles in which biofuel refers to input/feedstock. Not included are titles on SNG/natural gas and on fuel cells/hydrogen as we considered these out

of scope.

1980

1982

1985

1987

1989

1991

1992

1994

1996

1998

2000

2002

2004

2005

2007

2008

2009

2010

Methanol 2 1 1 2 1 1 1 1 3 1 1 Biofuel 2 2 1 3 3 2 1 1 3 2 6 14 8 14 4 4 1 IGCC 1 1 5 4 5 1 3 2 3 1 1

0%

5%

10%

15%

20%

25%

0,0

0,2

0,4

0,6

0,8

1,0European Biomass Conference

participantsarticles on gasificationarticles on gasification %

110

For the conference trends are reconstructed for participants and articles on biomass gasification, see figure 5.3. The first conference of 1980 can be considered an indication of an upcoming in interest at that time. While the number of publications on biomass gasification increased over the 1980s, the percentage of publications related to gasification remained more or less constant. Conferences of 1991 and 1992 show a significant drop in contributions overall and especially in those related to gasification. Between 1992 and 2004 there is a strong increase in participants and contributions that was even stronger for contributions on biomass gasification. This is not only the result of an increase in interest, as the setup of the conference changed, enlarging the circle of participants (Scheer, 1995). After 2004 the number of articles on biomass gasification show a decrease. This should not be interpreted as a decrease in interest, as over this period a large variety of conferences on biofuels and related technologies came up. In table 5.1 we specified contributions on the three applications of our interest. Only a small number of articles on gasification refer in the title to end-use applications. Over the 1980s articles mostly refer to methanol as biofuel. Articles on IGCC are mostly published from 1996 till 2000, although the technology continued to receive attention in later years. There is a strong increase in articles on biofuels in the period 2002-2007. The broad scope of the conference makes it hard to draw detailed conclusions, e.g. to use it for cross-examination with more specific indicators or trends in literature. In addition, contributions on biomass gasification in figure 5.3 include papers on small-scale fixed-bed gasifiers, which are not of our interest. 5.4.3 Web of science publications Academic involvement at international level is reconstructed from the Web of Science database by Thomson Reuters. Web of Science is an international multidisciplinary index for journal articles in all sciences. It encompasses references to articles of over 12.000 journals (Thomson Reuters, September 2015d) and thereby is one of the most complete databases available. Trends are reconstructed, see figure 5.4. It shows the three different periods of interest: first for methanol (1980-1987); than for IGCC (1993-2003); and finally for biofuels (2006-2014). Note than methanol is a specific type of biofuel: it explains for the publications on biofuels over the 1980s that mainly focused on methanol production; is also explains the more recent publications on methanol as part of the interest in biofuels. The figure reveals a residual interest after the initial wave of interest is gone: during the 1990s there are still some publications on methanol; and

111

after 2008 IGCC seems to have attracted a new wave of attention. Biofuels received by far the most attention (2014 publications: methanol 45, IGCC 35 and biofuels 241). The main limitation of this dataset is the limited coverage of developments prior to 1990: trends over this first period are based on only a handful of publications, while conference contributions suggest a much larger research effort at that time.

Figure 5.4 Publications related to biomass gasification (topic search on ‘gasif* and (biomass

or wood)’. Applications have been identified by topic searches on respectively ‘methanol’, ‘IGCC or (combined cycle and gas turbine)’, and ‘(Fischer-Tropsch or *diesel or biofuel or *gasoline or *ethanol or *methanol)’. Values are corrected for increases in publication

intensity in literature related to ‘energy’. Maxima have been indexed on 1. (Data: Thomson Reuters Web of Science (WoS), 2015c)

5.4.4 Derwent patent families Derwent patents are an indicator of technology development and industry involvement, with a broad international coverage - covering 40 worldwide patent-issuing authorities (Thomson Reuters, September 2015b). Double counts of identical patents in different countries is prevented as it counts patent families. Trend analysis (see figure 5.5) reveals the early interest in methanol and the recent upcoming of patents on biofuels, methanol and IGCC. The 2009/2010 peak illustrate the dominance of patents related to biofuels (38 patents), although a significant part of these are also related to methanol (22 patents). IGCC was much less patented (4 patents). Surprising is the lack of patents on IGCC over the 1990s. Developments prior to 2000 are based on only two hands full of patents, which makes it hard to come up with a reliable representation of detailed trends.

0

0,2

0,4

0,6

0,8

1Web of Science publications

methanol IGCC fuel

112

Figure 5.5 Patents related to gasification (IPC class C10J3*) and biomass (topic search ‘biomass or wood’). Applications have been identified by topic searches on respectively

‘methanol’, ‘IGCC or (combined cycle and gas turbine)’, and ‘(Fischer-Tropsch or diesel or biofuel or gasoline or ethanol or methanol)’. Values are corrected for average increase in patenting intensity in the IPC classes C10* ((fossil) fuel related) and F23* (combustion). Maxima have been indexed on 1. (Data: Thomson Reuters Derwent Innovation Index,

2015a)11 The findings of chapter 6 nuance this approach. In chapter 6 we study the development of fluidized bed biomass gasification more extensively by drawing upon the USPTO patent set. Results show that there are limited patents on wood or biomass; that there are many more on carbonaceous fuels – also by leading manufacturers in biomass gasification; and that this development is also strongly overlapping and linked to the development of fluid bed coal gasifiers. Another conclusion is that the strong interest in methanol and biofuel production did not show in the patent sets: apparently the development of end-use technologies are not closely related to the development of biomass gasifiers. In addition, the research reveals intensive and broad patenting on fluidized bed gasifiers in the late-1970s and early-1980s, which can be considered indicative for strong development efforts and progress made at that time. Developments in the late-1980s and early-1990s, the period of the increased interest in IGCC, strongly relate to fluidized bed combustion and steam technology that shows links with gas turbine technology.

11 The drop in patents after 2009 is unlikely to be the result of a backlog of unexamined patents, as we depicted indexed trends and not absolute amounts. To be sure we did check absolute values: while number of yearly patents in the broader field continue to rise until 2013, the number of yearly patents related to fuels and methanol show a drop after 2009 and continue to decrease over the period 2009-2014.

0,0

0,2

0,4

0,6

0,8

1,0Derwent patent families

IGCC

methanol

fuel

113

5.5 Late 1970s - early 1980s: methanol production 5.5.1 Underlying rationale – understanding the promise Energy policy in the 1970s can be characterized by great concerns regarding high oil prices and issues of depletion and dependency, mainly with respect to oil. Explorative studies emphasized the potential of domestic biomass (5-20% of the energy supply) - see for the USA Colitti and Baronti (1981), Klass (1985) and Stelson (1980); and for Europe Chartier and Hall (1982) and Palz (1980). Lignocellulosic biomass was the preferred feedstock, based on its annual world production and its limited competition with food (Chartier, 1981; Chartier and Hall, 1981; Klass, 1985). Advanced gasification technology for biomass was considered as an enabling and flexible technology that would make it possible for the first time to apply biomass for high-end applications, over a wide variety of applications and scales. The main focus was on application for methanol production, although some scholars actually favored other applications for initial market introduction (Brandon, King and Kinsey, 1984; Brandon and Kinsey, 1984; Bridgwater and Beenackers, 1985; Reed and others, 1979). Methanol as transport fuel in order to reduce oil dependency received significant attention by scholars and policy makers in the late 1970s and early 1980s. The application as octane booster in gasoline, replacing lead additives that were under scrutiny, was foreseen as initial niche market. This niche could be initiated using methanol from fossil fuels (coal, natural gas), as those technologies were considered more mature (Brandon et al., 1984; Klass, 1985, 1987; Sperling and DeLuchi, 1989; USDA, 1983). Biomass gasification hardly found application during the 1970s. It was framed as a promising but challenging technology. There was some speculation that biomass would be better suited for clean syngas production or for application in smaller plants. In addition, scholars mention the widespread application of small biomass-gasifiers in Europe during World War II and the availability of (near) commercial gasifiers for small-scale and less advanced applications. Other scholars tried to nuance these arguments. However, the arguments attracted attention and as such did contribute to the general promise of biomass gasification (Chartier and Palz, 1980; Overend, 1980; Reed, 1980b; Reed and others, 1979). 5.5.2 Actors involved building the promise Developments were driven by interests of academics, research institutes, gasification industry and governments. Industries involved in large-scale gasification were process industries, like pulp and paper industry supplier, boiler manufacturers and industry dedicated to gasification (Kirkels, 2014). There was only limited interest by the

114

automotive industry and a lack of involvement of methanol industry and petrochemical industry. Oil companies did not share the concern of short-term oil depletion (Bridgwater, 1990a; Klass, 1985; Kliman, 1983). Also the power industry was not yet involved (Klass, 1985, 1987; Synthetic Fuels Associates, 1982). Several new networking platforms were formed for scholars, industries and countries working on the topic. 5.5.3 Expectation dynamics From 1976 onwards we see an increase in budgets on energy from biomass (OECD/IEA, 2015) and broad patenting of fluidized bed technology, including coal gasification and combustion (see chapter 6). From 1979 onwards the interest in biomass gasification increased, as showed in chemical abstracts, scientific publications, patents and a further increase in budgets (OECD/IEA, 2015; Overend, in Stassen and van Swaaij, 1982; Thomson Reuters Derwent, 2015a; Thomson Reuters WoS, 2015c). The RD&D efforts were mainly focused on developing a new generation of advanced gasifiers. Both in the USA and in Europe multiple technologies were developed in parallel, while also in Canada and Sweden development efforts were ongoing. In each case, the subsequent conversion to methanol hardly received attention, as it was considered commercially available (Kirkels, 2014). Only a few scientific publications and conference contributions focus on the link with methanol production (EBC, 1980-2010; Thomson Reuters WoS, 2015c). The interest did not last. Already by 1983 the promise was being eroded; by 1985 the development came to a standstill. Several reasons caused this downfall. Scholars that looked back recognized that technologies turned out to be costlier and less competitive than originally hoped for; and that they took longer to develop, especially when compared to the sense of urgency that characterized the initial interest (Beenackers and Maniatis, 1984; Beenackers and van Swaaij, 1990; Klass, 1985; Stevens, 1994). In addition, gas cleaning proved to be important and difficult (Baker, Brown, Moore, Mudge and Elliot, 1986; Beenackers and Maniatis, 1984; Beenackers and van Swaaij, 1990; Rensfelt, 1984). And at the end of the 1980s attention started to shift towards the competing technology of biomass pyrolysis, at least in EU (Kirkels, 2012). Possibly more important was the failed market introduction of immature small-scale gasifiers. The poor reputation also affected interest in RD&D and industry participation on more advanced systems (Beenackers and van Swaaij, 1990; Hayes, 1989; Miles and Miles, 1989; Reed and Miles, 1988). An even stronger explanation can be found in the economic and political context. In 1983 the price of oil started to drop. The USA had been leading in funding and

115

development of energy from biomass and more specifically biomass gasification, but under the Reagan administration started to reduce RD&D budgets and refocused on longer-term projects. In 1985 oil prices dropped further and would remain at a relatively low level until 2000, which affected the prospect of biomass-to-methanol (Bridgwater and Beenackers, 1985; Klass, 1985, 1987; Rensfelt, 1984). Both funding and innovation activities came to a low by 1988 (see all indicators; Beenackers and van Swaaij, 1990; Diebold and Stevens, 1989; Fabry and Ferrero, 1989). Methanol production did continue to receive some attention over the 1990s, mainly in scientific publications that mostly considered a variety of fuels (Thomson Reuters WoS, 2015c). It also showed in the initial plans in the early 1990s for the Hawaii demo plant that focused on both IGCC as well as methanol production. Interest mainly revitalized with the broad upcoming of interest in second generation biofuels after 2006 – as is visible in trends of patents, scientific publications and conference publications. 5.5.4 Intermediate conclusions Dynamics of expectations at field level can be characterized as a hype-disappointment: overpromising early on, fast upcoming of interest, as well as a fast downfall of interest and disappointment. It mainly reflects expectations by researchers and policy makers and not so much expectations of possible adopters of the technology. Contextual factors show to be very important in both the upcoming and downfall of expectations. During the 1970s they helped in raising interest for the technology: the large concerns regarding high oil prices, depletion and oil dependency in the transport sector resulted in legitimacy, urgency and support for biomass gasificatin – as well as many other energy technologies. The macro expectation can be defined as the large scale deployment of biomass as modern energy carrier - locally available, clean, and efficient and for high-end applications. Expectations focused on the conversion of biomass-to-methanol to be applied as automotive fuel to reduce the risks of depletion and dependency. Clearly, this was a long-term strategic technological promise that provided legitimacy for the RD&D developments at the meso and micro level. It was technology driven; societal embedding was not given much consideration. Dynamics also worked the other way around: expectations at the meso level did contribute to these macro level expectations. The expectations of biomass gasification as a field made it possible to consider large scale deployment of biomass as modern energy carrier. This was a major shift compared to ongoing practices at

116

that time in biomass-to-energy, which were largely focused on relatively simple and small-scale applications in combustion, fermentation and digestion (Kirkels, 2012). The positive general expectations of methanol were soon translated in more specific expectations that fitted better developments in the market: methanol from biomass as possible future feed-in for a still to establish niche of methanol as octane booster. It would not require competition purely on fuel prices. Some did suggest other applications, but with the mandate for enactment intact by RD&D programs this did not result in much debate. This soon changed, when oil prices dropped and a change in policy was effectuated. Expectations at macro level turned very negative - to the extent that most actors halfway the 1980s considered biomass-gasification-to-methanol not competitive and no longer worthwhile pursuing. This also strongly effected expectations at meso level that initially were ambiguous: scholars were happy with the proof-of-concepts that had been achieved, but had to recognize that it took more time and effort than expected. Disappointment with respect to small-scale gasifiers effected the reputation of the whole field and the search for other niche-applications was unsuccessful, thereby reducing expectations for further development. Attention shifted to biomass pyrolysis. The micro expectations of early research programs and research plants fed back into the field (proof on concepts!), those of larger demo plants did not: by the time the larger demo plants had been completed in the second half of the 1980s, all interest had already vanished and therefore the demo plants were stopped (Miles and Miles, 1989; Kaltschmitt, Rösch and Dinkelbach, 1998). 5.6 1990s: IGCC power production 5.6.1 Underlying rationale – understanding the promise By the end of the 1980s energy policy had refocused. Coal and nuclear energy were considered no longer attractive, although electricity demand kept growing. The increasing attention for global warming strengthened commitments to renewables. Biomass-to-power was considered one of the more promising applications on the short term (Kaltschmitt et al., 1998; Williams and Larson, 1989). Traditionally, in biomass-to-power plants condensing steam turbines had been applied, mainly in biomass based industries: a well-proven process with limited efficiency (15-25%). Application of the more advanced Integrated Gasification Combined Cycle (IGCC) looked promising, as it would result in higher efficiencies (35-40%). Given the progress in the field of biomass gasification as well as related technological fields like fluidized bed combustion, a relatively short development time was expected, with demonstration before the turn of the century and fast commercialization (Grassi and

117

Bridgwater, 1990; Hall, 1997; Larson, Svenningsson and Bjerle, 1989; Overend, 1995; U.S. DoE, 1992). This technology was expected to be simpler and cheaper compared to the technologies developed in the 1980s (Kirkels, 2014). Crucial to the promise was the expected increase in efficiency. Based on it, the USA and the European Commission started to reframe the potential of indigenous biomass: with the same amount of biomass more energy could be produced. In addition, both were considering the cultivation of energy crops on agricultural lands to strengthen rural development, create jobs and further increase the potential of biomass-to-energy. However, the use of energy crops was relatively expensive. This could be counteracted by the increase in final conversion efficiency by applying IGCC, which made the whole concept more feasible (Bain, 1993; Bridgwater, 1993; Grassi, 1993; Grassi and Maniatis and Ferrero, 1995; Johansson, 1993; U.S. DoE, 1992; Walter, 1988). The increase of efficiency also ensured that the pulp, paper and wood industries could continue to cover their internal power demand (Wilén and Kurkela, 1998). Also for producing power for the grid it was important to improve competitiveness by increasing conversion efficiency and economies of scale (Overend and Chum, 1993; U.S. DoE, 1992). And finally and not least important, several scholars voiced the expectation that over time at larger scales advanced biomass-to-power, including biomass IGCC, would be able to compete with many alternatives, including clean coal and biomass combustion – see e.g. Johansson (1993), U.S. DoE (1992), and Williams and Larson (1993). 5.6.2 Actors involved building the promise Initial interest came from gasification industries and governments of a few countries: they voice the expectation, explore the potential and relevance of the technology and start showing commitment, for example by constructing demonstration plants. These efforts on demonstration plants involved utilities and gasification industry, but also World Bank and governments for the financial support, and a more heterogeneous set of actors: turbine supplier, wood supply system, constructor, financers, and etcetera. By new and increased networking activities knowledge exchange was stimulated, involving both experts and more heterogeneous actors. Several of the leading designs of the 1980s had been discontinued and manufacturers were no longer involved. Some other designs were adapted and new manufacturers came up (Kirkels, 2014; Stevens, 1992). Feedstock-based industries recognized the potential and showed interest, but mainly at arm’s length. In specific projects they were actively involved (Hawaii, Minnesota, Westinghouse), but they did not take a leading role in overall developments (Kirkels,

118

2014). EU projects were required to use crops or wood from agro-forestry plantations on set aside land - that had to be started up in parallel to the gasification plants. But within three years the EU policy completely changed: ‘the set-aside policy has been greatly reduced in scope, while incentives for traditional food production have been re-introduced’ (de Lange and Barbucci, 1998, p223). 5.6.3 Expectation dynamics The biomass IGCC option was already recognized in the early 1980s, see for example Beenackers and van Swaaij (1984), but at that time failed to receive support. By 1988 growing interest resulted in dedicated efforts: in Finland the 9 million ECU research program JALO started (Sipilä, 1993). The enthusiasm started to spread (Reed and Miles, 1988; Rensfelt, 1991; Solantausta and Beenackers, 1989; Stevens, 1992). In Sweden the government shifted focus towards renewables and domestic resources, forcing industry and utilities to start looking for alternatives (Hellsmark, 2010). In the USA in 1989, the Department of Energy started looking for proof-of-concept of an up-scaled gasifier (Diebold and Stevens, 1989). In addition, in an ambitious long-term plan the department together with EPRI - representing the electric utility companies in the USA - evaluated the potential use of biomass-to-power (U.S. DoE, 1992). From 1991 onwards the commitments in the USA resulted in a strong increase in funding12 (Overend and Chum, 1993). Broader and international interest was given an enormous boost from 1991 onwards when several parties publicly committed themselves to the realization of three demo plants (Bain, 1993; Elliot, 1993; Lundqvist, 1993; see also Kirkels, 2014). A first requirement in launching a biomass-to-power industry, the demonstration of IGCC, seemed to be fulfilled. The increase in interest also showed by an increase in conference participation and contributions (EBC, 1990-2010; Sipilä and Korhonen, 1993b), as well as in academic publications (Thomson Reuters WoS, 2015c). It did not result in a strong increase in patenting (Thomson Reuters Derwent, 2015a), which can be explained by the focus on demonstration of technologies that already were largely developed. After feasibility studies in 1992/1993, the European Commission joined efforts by presenting a plan for supporting three demo plants using wood from energy plantations. The plan strengthened expectations and ongoing efforts, as it covered most technologically significant areas; it strengthened the role of atmospheric gasifiers; it provided a further up-scaling; and it contributed to the volume (funding

12 Despite growing commitments this did not result in a strong increase in governmental RD&D funding (OECD/IEA, 2015). However, according to Kirkels (2012) - studying European Biomass Conferences – it was accompanied by a shift in focus: from biochemical to thermochemical technologies and from pyrolysis to gasification.

119

and number of plants) of ongoing efforts (Bridgwater, 1993; Johansson, 1993; Maniatis, 1998; Maniatis and Ferrero, 1995). Positive expectations were building on the foreseen effect of experience curves. Building several plants was expected to result in up-scaled and better integrated technology and an optimized operation of plants. This would reduce specific investment costs up to an acceptable level (Elliot, 1993; Sterzinger, 1994; U.S. DoE, 1992). By taking into account the total efforts and the planned demonstration plants, the general (and largely implicit) expectation was that significant cost reductions could be achieved. Actors were not very critical towards these expectations. Relevant questions would be whether these cost reductions would be realistic, or who would have to pay for the first several uncompetitive plants, or what to do with them – but these were hardly raised. By 1995-1996 expectations started to shift. Actors in the Scandinavian countries terminated plans due to the lack of funding for further demo plants. By 1998-2000 also the international efforts reached a turning point, which showed in a decrease of publications in the following years (EBC, 1990-2010; Thomson Reuters WoS, 2015c). Also, demo plants turned out to be significantly more expensive as expected, sometimes exceeding cost estimates by a factor two. Other demo plants had a problematic operational record, failed to compete with natural gas prices, failed to secure biomass supply, or suffered delays (Babu, 1999; Faaij, Meuleman and Ree, 1998; Hughes, 1998; Maniatis, 1998; Morris, Waldheim, Faaij and Stahl, 2005; Salo, 1998). This affected expectations, as was reflected in the working title of a Dutch report called ‘Identification of the real long term perspective for BIG/CC’ [emphasis in the original] (Faaij et al., 1998, preface); and was voiced by Rösch and Kaltschmitt (1998, p.211): ‘…It is not likely that in the near future an electricity generation from biomass via the gasification process will become competitive…. It seems to be rather unlikely that there will be many additional or new biomass gasifiers build in the near future on a commercial basis.’ However, several demo projects and research programs were still under way. Some – mainly in policy circles - still held high hopes or engaged in reconstructive efforts, see Maniatis and Kotronaros (2002) and Maniatis and Millich (1998). Around 2000 interest in biomass IGCC dropped. All remaining demo plants and efforts were stopped for both economic and technical reasons and many ongoing plans were cancelled (Bain, Amos, Downing and Perlack, 2003; Morris et al., 2005; Piterou, Shackley and Upham, 2008; Rensfelt, Morris and Waldheim, 2004; Stahl, Waldheim, Morris, Johnsson and Gardmark, 2004). Attempts to restart both the ARBRE and Värnamo plant with the help of the European Commission failed (Maniatis, Guiu,

120

Riesgo, 2003). Closing plants and discontinuing efforts limited the effect of experience curves, thereby further limiting the prospect for cost reductions of the already too expensive technology. Despite the stopping of demonstration projects, there was some continued interest in IGCC. It initially showed in conference contributions (EBC, 1990-2010) and after 2007 by scientific publications and patents, which suggests a new wave of interest. Part of this interest refocused on application in bio-refineries. (Bain et al., 2003; Kurkela and Kurkela, 2009; Thomson Reuters Derwent, 2015a; Thomson Reuters WoS, 2015c;). 5.6.4 Intermediate conclusions Expectations drew upon the prior developments of the 1980s in two distinctive ways: 1) the concept of biomass as modern energy carrier was already part of the repertoire; and 2) the technological progress, especially as the new technology of the 1990s was expected to be simpler and more in line with biomass combustion. The expectation dynamics at meso level can be characterized as a hype-disappointment cycle: overpromising early on with respect to potential, costs and development time and a fast upcoming of interest, and more realism and disappointment later on accompanied by a drop of interest. In the first phase of the hype we see positive expectations on all three levels of technological expectations that reinforced each other. The increased concerns regarding climate change resulted in commitments to reduce greenhouse gas emissions that provided direction and legitimacy in support of renewables. This provided the window of opportunity for the macro level expectations of large scale biomass-to-power based on domestic biomass supply. They were reinforced by expectations at the meso level: the prospect of achieving higher efficiencies by applying IGCC and a technology that could be applied on the short term fed back into the macro level expectations and allowed to redefine the total potential of biomass and provided policy makers a large-scale short-term option for addressing the issue of global warming. In addition, especially early on the micro level expectations of demo plants seem to have pushed the hype at meso level: the realization of demo plants suggests a certain maturity of the technology close to market as well as possibilities for learning, including the required cost reductions by going through the experience curve. The disappointment at the end of the period was also apparent in macro, meso and micro expectations. Emphasis on liberalization resulted in privatized power companies that looked for less expensive and less risky technology, which affected macro level expectations. Incumbent power producers favored co-combustion in coal-fired power plants and the application of wind energy (Kirkels, 2012). At the micro level several gasification plants were delayed or abandoned and suffered a variety of

121

social-economic and technological problems - including problems with gas cleaning and cost effectiveness. At the meso level the disappointment has not been well articulated. With the benefit of hindsight, we suggest this can be explained by different actors holding different expectations. In the market, among utilities and industries, interest dropped from about 1998 onwards. The technology was considered to be too expensive, too problematic, and not competitive. Policy makers, both in EU and the USA, remained much more positive as reflected in positive rhetoric as well as new plans. A second explanation is that even after the disappointment phase there was some continued interest, resulting in a new wave of interest from 2007 onwards. We consider the expectations over the 1990s a good example of ambivalence of promises. For the IGCC case issues like efficiency, market segments, diversity of scales, potential to scale up, and cost reductions by learning curve showed to be of major importance to raise interest. Values for these in literature were rather vague: not specifying what they were based upon, in contradiction with other sources by other actors, and depending on assumptions made. This resulted in claims and counterclaims made and a lot of uncertainty, which can be considered quite typical for this level of maturity. From the perspective of expectations, we conclude that this initial uncertainty did contribute to the promise - as it was not excluding any of the actors or applications! It was only over time, when the uncertainty persisted or in other cases the outcome was worse as expected (e.g. the promised effects of the experience curve which largely did not deliver) that it backfired at the expectations of the technology and contributed to the disappointment. 5.7 After 2000: fuel production 5.7.1 Underlying rationale – understanding the promise The application of biomass gasification for the production of biofuels was actively explored in the 1970s and 1980s, as has been described in section 5. The ammonia producing plant Kemira Oy (Finland, 1988-1990) and the methanol producing plant Schwarze Pumpe (Germany, 1995-2007) proved the technical feasibility of the concept (E4tech, 2009; Kirkels, 2014). In the meantime first generation biofuels based on the conversion of agricultural crops started to find application, especially ethanol in the USA and in Brazil. Over the 1990s developments in Europe were limited. By 1997 the EU published its White Paper on renewables, stressing the need for biofuels

122

(IEA, 2011; Van Thuijl, Roos and Beurskens, 2003). This ultimately resulted in the 2003 biofuel directive. Main drivers for support of biofuels were the projected growth in use of transport fuels, the wish to decrease oil dependency and CO2-emissions in this sector, coupled with the desire to sustain the agricultural sector and revitalize the rural economy (IEA, 2011; Kavalov and Peteves, 2005; Sims, Mabee, Saddler and Taylor, 2010; Sorda, Banse and Kemfert, 2010). There were strong regional differences in focus due to trends in the market and the imbalance of refinery output: predominantly on gasoline in the USA and on diesel in EU (Kavalov and Peteves, 2005; Shore and Hackworth, 2007). Short term application favored ethanol and bio-diesel production, as they could be applied using existing infrastructure and cars. In the market this mainly resulted in the application of first generation technologies that are based on the conversion of agricultural crops - mature technologies that were of support for rural areas. The USA and Brazil continued to dominate ethanol production and use, while EU started to dominate the biodiesel production and use (IEA, 2010a in IEA, 2011; Lamers, Hamelinck, Junginger and Faaij, 2011; Sorda et al., 2010). For long term applications, scholars considered other conversion technologies and fuels more appropriate, putting more emphasis on issues of high efficiency and low CO2-emissions. One of these technologies was biomass gasification: a second generation technology that could use lignocellulosic or woody biomass, and as such was considered to have a larger feedstock base, larger CO2-reduction potential and less interference with food supply compared to first generation technology. It also is a generic technology, as by gasification a variety of fuels can be produced. This fitted well the uncertainty with respect to preferred future fuels. 5.7.2 Actors involved building the promise The increase in interest, the size of the market and strong policy support by EU and USA created large momentum over different discourse spheres, as all the indicators show. It resulted in more and a broader variety of actors getting involved. At a community level, we saw an increase of dedicated journals, conferences, network organizations and lobby organizations. All can be considered indicative of a more active and dynamic role of the technological field. The gasification industry of the 1990s had a hard time to refocus, especially as their technology did not provide a perfect fit for biofuels production. Several new biomass gasification industries came up that frequently were building upon experience in other applications (Kirkels, 2014). And the momentum together with the start of biofuel trade also resulted in more countries getting involved (Kirkels, 2012).

123

From the start, just after 2000, both automotive and oil industries showed interest in the gasification route (Hamelinck and Bain, 2003). As the production of fuels would require specific knowledge and technology (e.g. competencies related to gas cleaning, conditioning and catalysis of the product gas, fuel distribution, etcetera) several alternative routes and demonstration projects were explored by consortia of industrial actors, including to some extent automotive and petroleum industry (Hellsmark, 2010). However, as Oberling, Obermaier, Skzlo and Lèbre La Rovere (2012) and Sims et al. (2010) show, the petroleum industry was more extensively involved in first generation biofuels as well as second generation technologies based on biochemical conversion. 5.7.3 Expectation dynamics The shift in policy interest and the relevance of biofuels was recognized by the biomass gasification community in the late-1990s, as showed in conference contributions, scientific publications and patents. Biomass gasification became increasingly framed as an enabler of the biofuels option, see e.g. Hamelinck and Bain (2003), Kwant and Knoef (2004), Maniatis (2001) and Maniatis et al. (2003). Kwant (2004, p.li), summarizing the 2nd World Conference on Energy from biomass, indicates about biofuels that ‘there was overwhelming attention for this subject’, which also showed in interest in second generation technologies. After 2006 it results in a strong increase in momentum, which shows by all indicators – except for the European Biomass Conference that had to face strong competition of other upcoming conferences. It coincides with renewed increase in patents and publications on IGCC, but absolute indicators on biofuels are much higher, thereby confirming the strong focus and large momentum. Just after 2000, biomass gasification could not compete in the market with first generation biofuels, nor was it considered mature enough for full-scale demonstration. Developments were focused on research and development on ultra-clean syngas and bio-refineries (Costello and Finnell, 2002; Kwant, 2004; Maniatis et al., 2003; Raldow, 2004). Production of ultra-clean syngas was considered a basic requirement for the production of biofuels and hydrogen (for fuel cells) and would also be beneficial for IGCC. Until 2005 mainly research plants were erected (Bacovsky, Dallos and Wörgetter, 2010; E4tech, 2009). The bio-refinery concept is defined in analogue with a petrol refinery: an industrial plant where multiple products are co-produced, which allows for energetic and economic optimization. Gasification can contribute in two ways: as the central converter in the case of Fischer-Tropsch diesel production; or when the focus is on biochemical conversion to ethanol (as in the USA is often the case) a gasifier can be added to convert residue into heat and power.

124

Over time, first generation technologies increasingly started to face opposition: assessments showed their limited contribution to reducing greenhouse gas emissions; several scholars point at the high production costs and high processing costs; and the competition with food production for available land and water formed the basis for the public food-versus-fuel debate around 2007/2008 (Sims et al., 2010). This contributed to an increased policy and market attention for second generation biofuels, which showed in an increase of demonstration plants - both biochemical and by gasification – as well as in plans for further up scaling. Government legislation, policies and funding played a major role in these developments. It also resulted in a strong increase in patents and scientific literature. After 2009 governmental funding and patenting intensity started to drop, although more money was allocated to RD&D on biogas (Bacovsky et al., 2010; E4tech, 2009; IEA, 2011; OECD/IEA, 2015; Sims et al., 2010; Thomson Reuters Derwent, 2015a; Thomson Reuters WoS, 2015c). The increased RD&D effort and government support did alleviate some of the uncertainties and risks from market parties, but could not take them away. In 2011 Choren Industries in Germany and Range Fuels in the USA closed down. Both had been considered leading initiatives: by the size of their demo plants, advanced concepts, and by large amounts of money and subsidies invested (Chapman, 2012; Hellsmark, 2010; Wikipedia, 2014). This has not halted developments, with new demonstration projects under way (EBTP, 2013). Typical, and deviant from prior developments, was the constant presence of a critical reflective line of reasoning that focused on what realistically could be achieved and the commitments required - see e.g. Bole and Londo (2009); Hamelinck and Bain (2003); Hellsmark and Jacobsson (2012); IEA (2011) and Kavalov and Peteves (2005);. 5.7.4 Intermediate conclusions Our reconstruction suggests initially a gradual rise in expectations at field level. Concerns regarding global warming and the focus on increased use of renewables were still present. Combined with a strong interest in agricultural policy and in reducing oil dependency this resulted in high macro level expectations with respect to the introduction of biofuels, with the first generation of biofuels being ready for market introduction. This also reflected on second generation technologies and biomass gasification. Both indicators and our reconstruction show a strong increase in interest and activities around 2003, in parallel with the UE biofuel directive. Around 2007/2008 expectations were strengthened by the public food-versus-fuels debate, and further institutionalization of biofuel efforts by policy makers that specifically targeted second generation technologies – resulting in a strong wave of interest. Over the most recent period 2010-2014 trends are ambivalent: patenting intensity and

125

governmental funding of RD&D on biofuels was dropping; while intensity of Web of Science publications remained high and governmental funding of RD&D on biogases increased; and this period is underexposed in the qualitative reconstruction. At the micro level expectations were initially low, as showed in the shift from demo plants to research plants. It took years before expectations with respect to specific plants started to contribute to the meso level - with a few exceptions like Choren. There were some efforts to build upon ongoing initiatives and plants, like the Värnamo plant, but using them for biofuel applications turned out to be difficult for a variety of reasons. At the end of the period, governments provided stronger and longer term support, thereby preventing that the disappointment in Choren and Range Fuels affected too much the expectations at meso level. At the meso level we see a redefinition of the field level: the new focus on biofuels opened up a much larger field, with many more actors and countries getting involved. The narrow field of biomass gasification industry and researchers initially did not have high hopes. They tried to refocus on the new application field, although with mixed success as other technologies were required and leading manufacturers of the 1990s did not succeed in leading developments on biofuels. The focus on clean syngas and bio-refineries were not only most relevant for biofuel production, but were also ideal bridging themes - themes that could contribute to both the ‘old’ focus on IGCC and the ‘new’ ones on biofuels and hydrogen production around 2003. Despite the obvious parallel with the developments in the 1980s of biomass-to-methanol, we did not find evidence that this strongly influenced macro or meso level expectations. 5.8 Conclusions and discussion 5.8.1 Conclusions We started from our interest in expectations of advanced biomass gasifiers and how these influence developments of the technology. We reconstructed and explained these dynamics by differentiating to levels of expectations and actors involved. We took into consideration expectations regarding biomass, advanced gasifiers and end-use applications. Meso level expectations on methanol production over the late 1970s and early 1980s showed a hype-disappointment pattern. This was mainly driven by contextual factors like high oil prices and the perceived risk of depletion: it showed to be of major importance in both the upcoming and downfall of expectations. Biomass-to-methanol-fuel served as a collective expectation that provided legitimacy and direction for policy makers, scientists and the gasification industry.

126

The expectations over the 1990s also showed a hype-disappointment pattern. In this case expectations were mostly well aligned over the three levels, both during the rise and fall of interest. However, the disappointment was hardly visible at meso level. We explain this by the different expectations of industry and utilities compared to policy makers, while renewed interest in the technology also came up relatively fast, only a few years after disappointment. We argue that this hype-disappointment cycle resembles more a Gartner type hype cycle, as diffusion of the technology was already foreseen. From the late 1990s onwards we saw a more gradual growth in interest in biofuels from gasification that from the start was accompanied by a critical reflexive line of reasoning. This interest was mainly driven by macro expectations and policy with respect to the broader and diverse field of biofuels – which is also encompassing first generation biofuels as well as second generation biofuels based on biochemical conversion. Actors within the gasification community tried to refocus and align to it. From 2006-2008 onwards this resulted in specific support for second generation biofuel technologies, including biomass gasification, resulting in a wave of interest – one with larger momentum and actor involvement compared to the earlier developments. After 2010 trends in the field are a bit ambivalent, with a drop in patenting intensity and governmental funding of RD&D on biofuels, while WoS publications remained high and governmental funding of RD&D on biogases increased. Both automotive and oil industries showed interest, although the latter were mainly involved in other biofuel technologies. We studied how the technology developed over three subsequent promises. We assumed that actors would position a new promise in relation to past developments. However, new expectations were only partially based on (progress in) the technology and references to or comparison with the previous promise were hardly made. Our analysis suggests three factors that played a crucial role: external circumstances, broader technological expectations, and alignment of policies. Contextual factors played a crucial and dominant role: the oil crisis, concerns regarding global warming, and agricultural policy. They influenced timing, momentum and direction of developments. The promises of methanol, IGCC and biofuels linked to these contextual factors, but also hooked onto the large technological promise of methanol-as-fuel during the 1970s and 1980s and of biofuels after 2000 – both of which were essentially building upon other technologies. Finally we saw that especially the EU and the US policies contributed by large RD&D budgets, involving many actors, upholding the promise of a breakthrough and extensive learning effects. Policies of both EU and USA showed to be relatively well aligned with respect to momentum and direction.

127

We applied a mixed method approach in order to accommodate the reconstruction of both expectations as well as dynamics in expectations and developments. In addition, a mixed methods approach can bring additional benefits – the triangulation of trends and the possibility to overcome specific bias of methods. We considered this especially important for the challenging case of advanced biomass gasifiers – a long term longitudinal study of global efforts of an open technology. Our findings show that quantitative indicators hold specific strengths over qualitative sources for reconstructing longitudinal patterns. They allowed us to better trace developments long after the initial hype had gone and dominant interests had refocused on other themes – and thereby our own analytical lens. In our case it revealed the continued low-level interest in methanol over the 1990s and a renewed upcoming of interest as part of the recent biofuel hype; as well as renewed interest in IGCC after 2006 – possibly in relation to the concept of bio-refineries. Quantitative indicators can also provide an indication of the extent of the hype. In our case, budgets and the total numbers of patents and publications showed that the biofuel hype was much broader and more encompassing compared to earlier hypes. However, it also showed the shortcoming of the indicators in practice, often due to limitations in the used datasets: a too broad scope (RD&D funding, conference contributions, conference participants); not covering the full time series (scientific publications that did not cover RD&D before 1990); changes of practice (European Biomass Conference in mid-1990s); or having difficulty to grasp both technological development as well as promise (patents). The literature sources counter-balanced this: often they contain some quantitative data as well as statements on trends – although often only relevant for a specific time period, country, or policy program. This is not enough to allow for quantitative assessment, but it can be very useful in support of trend analysis. We conclude that the mixed-method approach allows for cross-examination over sources and methodologies as validation of developments and trends. But we argue that it goes a step further: the outcome of a mixed-method approach is not always confirmation, it can also be contradiction, contestation, challenging findings. It offers a reality check for quantitative research by showing how specific datasets and algorithms relate to practice; as well as a reality check for qualitative research that corrects for subjectivity and cherry picking due to selective literature selection or interpretation by the scholar. It requires a reflexive approach, as methodologies and datasets do not always go well together. It offers the opportunity to reveal and overcome their inherent restrictions, but also the pitfall not to fully acknowledge them.

128

5.8.2 Discussion In the conclusions, we argued that quantitative indicators are proxies of trends and qualitative sources also contain detailed, valuable and sometimes additional information on trends. Based on this, we suggest an alternative visualization of field level dynamics: indicating the relative change in reconstructed expectations in year t2 compared to earlier year t1 and later year t3 (increase, equal, or decrease – which is a representation at ordinal scale). Hype-disappointment graphs based on this approach do not reveal how high expectations are, but whether expectations were rising or not compared to prior or later years – see figure 5.6. As such, the main focus should be on identifying starting points (take off, first upcoming of interest), turning points and end points (figure 5.6 respectively 1, 3 and 5). Points in between (2 and 4) can add to the reconstruction by revealing the extent of increase or decrease of expectations. As such, the figure is not the starting point for evaluating field level dynamics, but rather a representation of the outcome of a more detailed study on these dynamics. By adding information one can provide some insight of the underlying complexity and interrelatedness. The figures are a powerful and meaningful way to communicate trends. However, they are proxies and as such not very accurate. Although more complex dynamics can be visualized in this way, it is not recommended as it does not allow for a meaningful comparison between multiple hypes or between multiple periods of low interest.

Figure 5.6 (left) Reconstruction of hype-disappointment based on trend analysis on ordinal scale. (Right) Reconstructed dynamics of meso or field-level expectations of biomass

gasification for combined cycles (IGCC). Main influences have been made explicit, including relevant contextual factors and the macro and micro expectations that fed into the

expectations of the field. In this article we reconstructed developments of biomass gasification from an inside-out perspective, mainly following contributions of actors working in or contributing to

expe

ctat

ions

(or

dina

l sca

le)

time

1

2

3

4

5

129

the field. We identified the strong role of policy makers and the weak role of related industries and of end-used sectors. However, we do not reveal the internal dynamics of policy making or the detailed position of related industries – like petrochemical industries, pulp and paper industries and utilities. We argue that detailing this out might add understanding to the developments. We showed that developments in biomass gasification received long term governmental support and that there was strong international alignment of efforts. Both strong alignment as well as long term support seem to conflict with literature: policy interest is characterized as opportunistic, influenced by image pressure of governments and politicians and as such affected by election cycles. It can partially be explained by the nature of the technology, with long time spans for RD&D and the construction of demo plants. For this study we choose an international perspective. It allows us to trace expectations in the field of biomass gasification and to take into account the interplay of local, national and international expectations. Based on this article we see two interesting areas for further inquiry: differences in the roles of countries involved; and international alignment. We showed that different countries can have different roles. This is influenced, amongst others, by their resources, policies and size. We showed the importance of momentum that provided direction and scale to developments. Especially efforts by the EU and USA played a large role in creating the scale. On the other hand we recognize specific roles for smaller countries. Sweden provided expertise and continuity to developments. Finland contributed with the Jalo program to the upcoming of the IGCC hype. Smaller countries might even hold specific beneficial characteristics due to their smaller size: more responsive, better locally embedded, higher social proximity of actors involved in the innovation process. Most important for the momentum is the alignment between countries. Partially this can be explained by (expectations of) technological maturity, competitive threat amongst manufacturers and countries, and windows of opportunity related to external forces and circumstances operating at international scale. However, the mechanisms and dynamics of alignment between nations remain underexposed and are not fully understood. To shed more light on these alignment processes might require looking at other sources, like political and policy debates to reveal the country perspective – although also politics is not always (or by definition not?) transparent about strategic motives and choices. It also might require drawing upon other conceptual frameworks that focus more specifically on politics and policy.

130

Acknowledgements The author likes to thank Marloes Dignum, Geert Verbong, Floortje Alkemade and two anonymous reviewers for their helpful comments.

131

132

Chapter 6

Embedding emerging technological trajectories:

a patent analysis of biomass gasification

133

6.1 Introduction Global sustainability requires the development and deployment of new technological options based on renewables. The energy domain is particularly under pressure as the use of fossil fuels contributes to depletion, global warming and air pollution (UNDP, 2004; WCED, 1987). Worldwide, initiatives are developed to take up the challenge. Many technologies have been developed and start to find application (REN21, 2014). However, often market shares lag behind targets and the general expectation is that we will have to rely on fossil fuels for at least some more decades (Verbong and Loorbach, 2012; World Energy Council, 2013). A faster development and uptake of renewables is required to avoid the worst effects of climate change (IEA, 2015). This requires not only knowledge on individual technological options but also on the general process through which these new technological options emerge and develop. We study the development process of biomass gasification technologies. Biomass is currently the most applied renewable energy source and a further increase in application is expected (Faaij, 2007; Hoogwijk, 2004; OECD/IEA, 2014; World Energy Council, 2013). The development of advanced biomass gasification plants received a lot of attention over the past four decades. Nowadays application is foreseen for the production of biofuels and in bio-refineries that contribute to the bio-based economy. One of the most often applied technologies over time is fluidized bed gasification (Kirkels, 2014; Knoef, 2005a). Biomass gasification is thus an example of a new technological option that succeeded in attracting attention, with many different options being proposed and explored. As such it will allow us to study early stages of development and interaction between technologies. Innovation scholars study the development of technologies in order to better understand the innovation process and determine factors that influence successful development and diffusion. These scholars argue that technological developments are highly selective: of all possible technological options only a few get explored. Often technologies develop over technological trajectories - improvement sequences of a technology characterized by incremental innovation resulting in technological continuity. Because of shared routines, engineers in a technological field work in more or less the same direction. Discontinuities are associated with radical innovations that happen less frequently (Dosi, 1982; Nelson and Winter, 1982). These technological trajectories show path dependency which indicates that history matters, e.g. by building upon prior knowledge, experience and routines (Garud and Karnøe, 2001; Garud, Kumaraswamy and Karnøe, 2010). This implies that initial starting conditions are of relevance. A second determinant of path dependency are the self-reinforcing processes that can explain the stickiness of trajectories, like sunk costs, learning effects, increasing returns and bandwagon effects. They result in quasi-

134

irreversibility or lock-in, where “effort and imagination … remain ‘blind’ to other technological options” (Biondi and Galli, 1993 p580). And finally the paths are the result of contingencies: somewhat unpredictable and exogenous events, also described as chance or historical events (Arthur, 1988; Garud et al., 2010). Other scholars embraced the concept of technological trajectories, but have argued that the original concept has a too narrow perspective on technological change. These scholars have argued for a stronger role of agency to highlight how different stakeholders embrace and influence technological trajectories - see Garud and Karnøe (2001), Garud et al. (2010), Geels (2002), Kemp, Rip and Schot (2001), and Rip and Kemp (1998). They therefore coined the term path creation. This perspective affects all the determinants of path dependency (Garud et al., 2010): initial conditions are constructed, with actors playing an active role in determining what portions of the past they will draw upon; actors know that unexpected events will arise and have to be prepared to improvise in order to deal with these contingencies; actors can anticipate and even strategically manipulate self-reinforcing mechanisms; while lock-in provides provisional stabilization within a broader structuration process. In this paper we argue that an understanding of the technological trajectories of biomass gasification requires an analysis of both the factors that led to path creation and the further development of the technological trajectory. This requires taking a ‘real-time perspective’, i.e. place oneself at the time that events occurred. “Otherwise, it would be tempting to think of any sequence of events (retrospectively labelled as a path) as having been inevitable.” (Garud et al., 2010, p770). To study path dependency often quantitative methods are used that draw upon large datasets. They assess in retrospect the dominance of technological paths. An example is the approach by Verspagen (2007) analyzing patent-citation networks: patents relate to inventive activities, while citations link between patents and can be considered a representation of knowledge flows between patents. Verspagen uses the patent-citation network to reconstruct the dominant development path based on the structure of the network, by identifying a sequence of patents with highest connectivity. The method has been criticized for being over selective: not representing well broader technological developments (Bekkers and Martinelli, 2012; Bhupatiraju, Nomaler, Triulzi and Verspagen, 2012). However, the underlying patent-citation network does allow to uncover these broader developments in real-time perspective and to assess how these developments are related. Kirkels (2014) studied the development of advanced biomass gasifiers. He suggests that it was essentially an interrelated technological development: on the one hand influenced by closely related technologies that share a knowledge and technology base (e.g. biomass combustion, other types of gasifiers); and on the other hand influenced by complementarity of technologies that are applied together in a biomass

135

gasification plant (e.g. gas cleaning unit, end-use conversion, but also combustion or pyrolysis units). Our aim is to uncover these interrelated developments as they might have facilitated the emergence of paths and the further path formation. Understanding their role might help to structure the search for high-potential technologies; and it might provide clues on how to foster initial development. We pose two research questions. How did biomass gasifiers develop over time and did this result in technological trajectories? And how did interrelated technological developments influence the emergence and formation of paths? In the next section we will discuss possibilities to uncover technological developments by studying patent-citation networks. In section 3 we will describe the methodology, including the identification of relevant datasets. To answer our research questions we will draw upon analysis of patent-citation networks: patent landscaping in order to reconstruct ‘real time’ technological developments by studying patents (section 4); studying cross-citation in order to better understand how developments are linked (section 5); and linking this with the Verspagen approach to see how this influenced technological trajectories (section 6). Finally, we will draw conclusions on technological developments, interrelated developments and how these influenced the reconstructed technological trajectories. 6.2 Uncovering patent citation networks A patent is a document, issued by an authorized government agency, granting the exclusive right to manufacture, use, or sell an invention for a certain number of years. Therefore, by definition patents are related to inventive activities. Patents contain a richness on information regarding inventor, company, country, priority year and year of publication and technology (e.g. by classification in pre-defined classes). In addition, patents are required to make citations to prior art. Several mechanisms ensure an extensive coverage of this prior art: in the US there is the ‘duty of candor’ by which applicants should disclose all the prior knowledge they are aware of; the threat of law suits based on patent infringement; and finally patent examiners of the patent office that add citations to the patent (Alcácer, Gittelman and Sampat, 2009; Hall, Jaffe and Trajtenberg, 2001). Citations provide a ‘paper trail’ of spillover: the fact that patent B cites patent A is indicative of knowledge flowing from A to B (Hall et al., 2001; Verspagen, 2007). Even in cases where citations are added by the patent examiner, and therefore do not necessarily represent the prior knowledge that the inventor actively drew upon, citations signal technological similarity. The large amount of available patents, the requirement to cite prior art combined with the systematic registration of patents and citations, allows for a bottom-up

136

reconstruction of technological developments (Albert, Avery, Narin and McAllister, 1991; Basberg, 1987; Griliches, 1990). Studying broad patent sets involves patent landscaping: by straightforward patent counts reconstruct technological developments over time and identify companies involved. However, this implies that all patents are considered to be of equal importance – which is highly contested (Albert et al., 1991; Griliches, 1990; Hall et al., 2001; Harhoff, Scherer and Vopel, 2003; Karki and Krishnan, 1997; Narin and Olivastro, 1988; Trajtenberg, 1990). The number of citations received by a patent can be considered a measure of the technological and economic importance of the patent, as this indicates the significance for later technological developments. The distribution of citations shows to be highly skewed: a significant part of the patents is hardly cited and therefore has little significance, while a few others are highly cited. The focus should be on the latter, the highly cited patents. For complementary technologies such an approach may have drawbacks, as innovation is distributed over multiple interacting components (Fontana, Nuvolari and Verspagen, 2009). Also, a strict application of such an approach results in fragmented findings: by only looking at highly cited patents it is very difficult to grasp the overall relationship among all patents (Yoon and Park, 2004) and thereby understand how technologies evolve. This criticism can partially be overcome by looking at paths. The focus is no longer on single patents, but on series of subsequent patents connected by citations (a path) that characterize technological development over time. One can see the analogy with historical studies that describe dominant developments over time. In general, such a path covers some of the highly-cited patents, but also patents that connect highly-cited patents, or patents that represent more incremental innovation (Verspagen, 2007) - “improvement sequences in which devices or procedures are progressively refined and extended in their scope of application” (Mina, Ramlogan, Tampubolon and Metcalfe, 2007, p791). Of special relevance for our purpose is the work by Verspagen (2007), drawing upon earlier work by Hummon and Doreian (1989). He identifies a top path in a patent citation network as representation of dominant knowledge flows through the network. He does so by taking into account both direct as well as indirect citations (e.g. if document A cites document B and B cites C, an indirect knowledge flow goes from C to A); and by considering two-sided connectivity (citations received as they indicate influence on later developments, as well as citations made as this accounts for the cumulative nature and path dependency of technological development). As such, the path represents a local and cumulative chain of innovations consistent with the definition of technological trajectories (Bekkers and Martinelli, 2012; Bhupatiraju

137

et al., 2012; Verspagen, 2007). The approach has been successfully applied on several domains: fuel cells (Verspagen, 2007), medical knowledge (Min et al., 2007), Ethernet LAN communication technologies (Fontana et al., 2009) and mobile telecommunications (Bekkers and Martinelli, 2012). Identifying a dominant top path leaves many questions open. What prior art is it drawing upon? Does the network also contain other technological developments? To answer these questions requires a study of broader technological developments – note the strong parallel with the earlier criticism by Yoon and Park regarding a too strong focus on highly cited patents. To do so, Verspagen (2007) focused on mapping top paths over different periods in order to identify shifts in focus; Mina et al. (2007) and Bhupatiraju et al. (2012) apply different sorts of clustering analyses to identify different developments; Bekkers and Martinelli (2012) suggest to study not only the patents on the path, but also (a selection of) the patents that they draw upon; while Fontana et al. (2009) focus on the relatedness technologies by comparing groups of patents. 6.3 Methodology We focus only on US patents. This holds the advantage that patent and citation tables are homogenous for metric purposes. We realize that this focus might introduce a geographic bias. However, the US patent system is an open system – a relative large share of patents is held by non-US assignees (Albert et al., 1991; Griliches, 1990; Karki and Krishnan, 1997; Hall et al., 2001). Hung and Wang (2010) argue that one of the reasons behind this is that US patents offer strong protection in a large market. We assume that dominant international developments will be covered by the patent set, as RD&D on advanced biomass gasifiers are typically undertaken by large firms that serve an international market (Kirkels, 2014), with a significant contribution from US companies and research institutes, especially from the late 1970s to the early 1990s (Kirkels and Verbong, 2011). In addition, the alternative strategy for protecting innovations - keeping things secret - does not work well in the sector, with very long times to commercialization and public exposure of demo plants. For selecting datasets the scope is of crucial importance, especially as interrelated developments lack clear boundaries: a broad scope ensures coverage of all relevant patents, but is also likely to include a significant share of non-relevant patents; while a narrow scope is more specific, but not necessarily all inclusive. Prior literature indicates that advanced biomass gasifiers are mainly using wood as feedstock; are mainly of the fluidized bed type, although other types of gasifiers have

138

also been applied; and that over time multiple manufacturers were involved (Kirkels, 2014; Knoef, 2005a). Therefore we did two key word searches (‘gasif$ and (biomass or wood)’ and ‘gasif$ and (fluidized or fluidised)’), using the search engine of the United States Patent and Trademark Office (USPTO, 2009-2015). For both resulting patent sets we specified the technological classification, revealing the dominance of classification C10J3 of the International Patent Classification13. It covers patents on the production of combustible gases containing carbon monoxide from solid carbonaceous fuels (WIPO, 2015). Fluidized bed or Winkler technology is classified in subgroups 54 & 56. By the USPTO search engine we retrieved all patents in these subgroups over the period 1976-2008 (Fl1 dataset) to represent the developments of fluid bed gasification. The start year precedes the upcoming of interest in biomass gasification (see Kirkels, 2014; Kirkels and Verbong, 2011), while the end year is given in by the start of this study. For citations and patent attributes (e.g. year of publication14, company, etc.) we have been drawing upon the Hall, Jaffe and Trajtenberg patent citation data file covering 1976-2003. We updated it to include patents and their citations made for the 2003-2008 period. The dataset shows relatively low connectivity of 1.5 citations per patent, see table 6.1 and figure 6.1. The patents in the dataset cite many patents outside the dataset (53%), so are strongly embedded in a broader technological context. For each citation made to earlier patents on fluidized bed gasifiers also 2-3 citations are made to patents on other technologies. It confirms the importance of one of our starting points – that of interrelated technological developments. It also shows the need to expand the dataset in order to take these developments into account. We defined an additional broader dataset ‘FL2’ by snowballing: including all patents that are citing to or are cited by patents in the Fl1 dataset. This way patents on a variety of related technologies are included, which allowed us to study interrelated developments. In the resulting FL2 network the connectivity has increased to 3 citations/patent (see table 6.1 and figure 6.2). However, it does not resolve the demarcation issue: 48% of citations go to patents outside this dataset, so also this dataset is strongly embedded in a broader technological context. For patents in the Fl2 dataset additional patent information (company names, title, abstract) was retrieved by using Derwent (Thomson Reuters, 2015a). The international patent

13 The classification is by code, like C10J 3/54. Each digit adds extra detail: C is referring to Chemistry; C10 to petroleum, gas and coke industries; etcetera. Both USPTO and IPC classification were checked upon, with IPC class C10J3 showing the best fit 14 For our datasets the date of publication (issue date) shows on average a time lag of three years with the priority date, which is the date that the patent or a patent family member was first filed.

139

classification was retrieved by using Patstat (European Patent Office, 2015) for both the Fl1 and Fl2 datasets.

Table 6.1 (left) Characterization of patent-citation datasets used Figure 6.1 (right) Visual representation of the relationship between the Fl1 dataset and the

Fl2 dataset. Fl1 Fl2 indicates Fl1-patent citing Fl2-patent-not-also-being-Fl1-patent. Fluid bed Fl1

C10J 3/54&56

1976-2008

Fluid bed Fl2 incl. citing &

cited 1976-2008

Patents Internal citations Internal citations / patent % internal to dataset % citations outside dataset % citations <1976

389 565 1.5

19% 53% 27%

2433 7273 3.0

24% 48% 28%

Over the next sections we will come to a patent analysis by an informed search, drawing upon the rich body of literature on biomass gasification: it provides structure that helps to understand the diffuse technological developments; and it links our findings to the literature. We will also specify per section methodological details (procedure followed). In the next section, section 4, we will start with patent landscaping: by patent counts and citations made identifying dominant technological developments and companies involved over time. In section 5 we explicitly study the interrelatedness of technological developments. We do so by identifying relevant subsets for different gasification technologies, thermochemical conversion technologies, feedstock and end-use applications. For the subsets we assess cross-citation and co-classification. The level of cross-citation indicates the extent to which one technology draws upon another technology. Co-classification indicates a level of similarity of technologies or the presence of hybrid technologies. As such it also might explain to some extent the level of cross-citation. In section 6 we come to a reconstruction of technological trajectories by the algorithm as used by Verspagen and their embedding in a broader technological context.

Fl1Fl2 1560 cit. Fl2Fl2 3480 cit. Fl2Fl1 1668 cit.

Fl2 2433 patents 7273 citations Fl1

389 patents 565 citations

140

6.4 Developments in fluidized bed gasification First we started with patent landscaping in order to identify dominant developments over time. For both patent sets we identified companies involved, frequency over time and an outline of technological development. By identifying dominant companies we can see who was involved in the technological development, as well as check whether the datasets represent the companies that are known from literature to be involved in biomass gasification. In response to the argument that not all patents are of equal importance, we checked on dominant companies by three different measures – see table 6.2: number of patents per company; number of citations received by company (which is the result of both amount of patents and their respective citations); and companies that hold the most cited patents. Upfront we would expect a relative good fit between companies involved in biomass gasification and the Fl1 dataset that is committed to gasification; and a somewhat less fit with the Fl2 dataset as it also extends to other technologies. Exxon is dominant according to all three measures, especially in the Fl1 dataset. The development effort by Exxon is strongly concentrated in the period 1976-1982. Other companies in the datasets are stronger associated with biomass gasification (e.g. see Kirkels, 2014): IGT, Metallgesellschaft, Foster Wheeler, Ahlstrom and Ebara. They are especially strongly present in the Fl2 dataset, which suggest that the links with broader technological fields have especially been important for biomass gasification. Other companies that Kirkels (2014) identifies as being important include TPS/Studsvik and Framatome – which are not present in the top-10. They do show up in the broader patent sets, but only hold a few patents. This might be explained as these companies are not broad engineering firms or boiler manufacturers and as such did not provide broad contributions to the field. Their focus seems to have been much more on the implementation and diffusion of specific technologies. As a consequence of the snowballing, the dataset does include a large number of Texaco and Shell patents – which are known to dominate entrained flow coal gasification (Harmsen, 2000; Kirkels and Verbong, 2011). However, their patents don’t receive a lot of citations in the dataset, from which we conclude that they had little influence on overall developments. As the role of Exxon in biomass gasification is not widely recognized in prior literature, we made their contribution explicit by detailed literature study. Exxon’s Oxygen Donor Process has been applied by John Brown and Wellman in a pilot project of advanced biomass gasifiers for methanol production that received support from the European Commission in the early 1980s. (Dallas, McLellan, Scanlon and Smith, 1986). Also, Exxon is known for the catalytic gasification of coal to produce methane (van

141

Swaaij, 1981), and for the Exxon Donor Solvent process for direct coal liquefaction (Neavel, Knights and Schulz, 1981). Table 6.2 Dominant companies in Fl1 and Fl2 datasets as identified by # patents, # citations

received, and by patents that received most citations

Fl1 companies # pat. Fl1 companies # cit. rec.

FL1 patents # cit. rec. company

EXXON 52 EXXON 119 3957459 12 EXXON IGT 18 IGT 32 4676177 12 A. AHLSTROM USA DoE 17 USA DoE 31 4157245 11 CHEVRON METALLGESELLSCHAFT 10 METALLGESELLSCHAFT 24 4017272 11 BAMAG FOSTER WHEELER 9 EBARA, NIKKISO 24 3966634 10 COGAS BERGWERKSVERBAND 9 A. AHLSTROM 21 3985519 10 EXXON CHEVRON 8 CHEVRON 20 4032305 10

Fl2 companies # pat. Fl2 companies # cit. rec.

FL2 patents # cit. rec. company

EXXON 94 EXXON 505 4165717 47 METALLGESELLSCHAFT FOSTER WHEELER 74 FOSTER WHEELER 460 4469050 41 YORK-SHIPLEY TEXACO 62 METALLGESELLSCHAFT 241 4111158 37 METALLGESELLSCHAFT USA DoE 49 IGT 213 4157245 37 CHEVRON METALLGESELLSCHAFT 38 A. AHLSTROM 203 4594967 36 FOSTER WHEELER IGT 37 USA DoE 196 4338283 35 BABCOCK-HITACHI

SHELL OIL COMPANY 32 EBARA, NIKKISO 163 4039272 34 STONE-PLATT FLUIDFIRE

A. AHLSTROM 32 TEXACO 159 4896717 29

The patenting intensity over time can be considered an indication of intensity of development effort over time. Patenting intensity for both datasets is displayed in figure 6.2, both in absolute numbers as well as corrected for the decrease in patenting intensity over the years.15 The broader dataset FL2 is after 2004 clearly affected by the incomplete citation tables used. The significant amount of patents already published in 1976 is not in line with literature that places the upcoming of interest in biomass gasification at the late-1970s based on literature study as well as a broad variety of indicators – see Kirkels (2014, 2016) and Kirkels and Verbong (2011). It suggests that around 1976 developments were not limited to biomass gasification; that there was a broader engineering effort on which the subsequent work on biomass gasification could build. After 1984 patenting intensity drops. Kirkels and Verbong (2011) explain this by a shift

15 The general tendency is a strong increase in patenting intensity over the years, however this is not true for the relevant sectors for our case – see subscript figure 6.2.

142

in geographic focus by which Europe starts to dominate the field. However, another explanation might be that the technology shifted from a phase of exploration to a phase of diffusion: several of the leading technologies of the 1990s (like TPS/Studsvik and Lurgi) were already developed and patented over the 1980s – as our patent sets reveal; while developments over the 1990s focused on demonstration (Kirkels, 2014). However, this does not explain the rise that the Fl1 trend line shows over the 1994-1998 period.

Figure 6.2 Patent intensity over time in absolute numbers (bars, left axis);

and indexed (trend lines, right axis). Indexed values are corrected for average decrease in patenting intensity in the IPC classes C10* ((fossil) fuel related)

and F23* (combustion). Maxima have been indexed on 1. In addition we reconstructed technological developments according to IPC classification. We did so for both the Fl1 and Fl2 patents (see table 6.3), as well as by citations from Fl1 patents to FL2 patents which are representing the knowledge flows of related technologies that the development of fluid bed gasification has been building upon. All show similar results. To represent developments we used four digit IPC classes for reasons of clarity and to ensure substantial amounts of hits per category. For the interpretation of trends we also looked at more detailed classification. Over all periods subclass B01J is relevant. It refers to (fluidized bed) apparatus for chemical processes and catalysis. Many of the patents on gasification or combustion are also classified in this subclass. Over the period 1976-1985 the subclasses C10G and C10B are relevant. C10G refers to hydrocarbon cracking and includes the production of liquid hydrocarbon mixtures from coal and wood. C10B refers to the destructive distillation of carbonaceous materials. These are mostly patented by Exxon, while also Chevron and Occidental Petroleum Corporation hold several patents. Over the period 1981-1996 (fluidized bed) combustion is often patented

0

1

0

20

40

60

80

100

120

inde

xed

# pa

tent

s

Fl1 Fl2

143

(F23C). Foster Wheeler and Ahlstrom hold most of these patents. This is followed by a shift in focus to the incineration of waste and low grade fuels (F23G) over the period 1997-2002, mainly patented by Ebara. Over both periods there is attention for the separation of gasses and the purification of flue gasses (B01D). During the last period there is significant attention for the conversion of chemical energy into electrical energy (H01M) which largely evolves around fuel cells (H01M8); as well as for hydrogen production (C01B).

Table 6.3 Dominant classifications of Fl2 patents over time, as percentage of total classifications.

1976-1980 1981-1985 1986-1990 1991-1996 1997-2002 2003-2008

C10J 20% C10J 20% B01J 14% B01J 11% F23G 10% B01J 17% B01J 10% B01J 13% F23C 14% F23C 11% C10J 9% C10J 15% C10G 9% C10G 10% C10J 12% C10J 9% B01J 8% C01B 12% C10B 7% F23C 7% B01D 6% B01D 8% B01D 7% H01M 9% C01B 6% C10B 5% F23G 5% C01B 5% F01N 5% F23G 5%

As a preliminary conclusion we can state that both data sets seem to be a good representation of biomass gasification. Companies that are known by literature for their contribution to the development of advanced biomass gasifiers are covered, but both data sets cover broader developments. There is not a clear preference for one specific dataset: the patenting intensity of the Fl1 dataset - which is restricted to fluidized bed gasifiers – reflects the three waves of interest that characterize the development of biomass gasification according to Kirkels (2014); while the companies involved in biomass gasification hold more patents in the broader Fl2 dataset. In addition we identified trends over time. High patenting intensity in 1976 and subsequent years suggest a broader engineering effort on which the development of advanced biomass gasifiers could built – as these started to receive interest in the late-1970s. This engineering effort includes the work by Exxon, which is dominant in that period, amongst others on cracking and destructive distillation. Over the period 1986-1996 combustion technology was of large relevance, followed by waste incineration (1997-2003). Most recent developments link to fuel cell technology and hydrogen production. In the next section we study in detail how different developments are related.

144

6.5 Interrelated technological developments Both the first analysis of datasets as well as the analysis in the previous section indicate that both datasets encompass multiple developments. We would like to know how these developments are related, as this might provide an understanding on how they facilitated the emergence of paths and the further path formation. Therefore we need to study developments in more detail and assess to what extent these developments are related. We will do so based on co-classification and cross-citation for different subsets of patents that relate to a variety of gasification technologies; several thermochemical conversion processes (combustion and pyrolysis); gasification of different types of feedstock; and for different end-use applications. 6.5.1 Gasification technologies Several different gasification technologies have been developed, see textbox 6.1 and appendix A for a more detailed description. They share social and technological proximity: some of the literature and conference sessions cover both; and some manufacturers have been involved in multiple technologies (e.g. Ahlstrom and Lurgi that have both been involved with fixed bed and fluid bed technology). Kirkels (2014) identified two periods during which developments might have been affected by spillover. During the late-1970s and early-1980s small-scale, fixed-bed gasifiers were considered more mature and became applied for low-end applications. Some of their manufacturers did consider entering the market for more advanced applications. During the second period, from the late-1990s onwards, there was an upcoming interest in entrained flow gasifiers for the gasification of black liquor and for biofuel production.

Textbox 6.1 Characterization of gasification technologies.

Fixed bed technology: a fixed bed of feedstock is being gasified using a gasification medium, generally air at low velocity. Main subtypes are downdraft and updraft gasifiers, which are mainly applied at smaller scales. Fluid bed technology: a small fraction of feedstock is added to a much larger fraction of bed material, which is fluidized by a gasification medium (air, oxygen, steam) that flows through the bed at a high enough speed. Main subtypes are the bubbling and the circulating fluidized bed, which are mainly applied for biomass at medium scales. Entrained flow gasification: small droplets or particles of feedstock are ‘entrained’ in a flow of gasifying medium – in general oxygen or steam. Also referred to as suspension flow or dust cloud gasifiers. It has been mainly applied at larger scales for coal and petroleum based feedstock.

145

We analyzed the broader FL2 dataset that encompasses possible spillover between gasification technologies. We identified per technology the related classification, as well as companies involved and most relevant periods. We assessed interrelatedness by checking on co-classification and cross-citation. The patent set includes a limited amount of patents on fixed bed technology, mainly published over the period 1979-1984, that mainly influenced developments in fluid bed gasifiers over the period 1984-1988 and 2004-2005. They are held by a variety of companies, including Metallgesellschaft. There are significantly more patents on entrained flow technology. They are mainly held by Texaco, Shell, Ebara and Combustion Engineering. Results on co-classification (table 6.4) and cross-citation (table 6.5) indicate the relevance of entrained flow technology and especially subgroup 46. Ebara holds many of these patents that are co-classified; and Texaco, Metallgesellschaft, Combustion Engineering and Ebara hold many of cited patents. The entrained flow technology was mainly patented over the late-1970s and early-1980s, but these patents were mainly cited after 1996 and thereby influenced the development of Winkler technology over that period.

Table 6.4 Co-classification of patents on different gasifying technologies within the Fl2 dataset, as percentage of total patents in respectively subgroups 54 and 56.

IPC subgroup & description

2 Fixed bed

20 Fixed bed, apparatus or plant

46 Granular or pulverulent fuels in suspension (=entrained flow

)

48 Granular or pulverulent fuels in suspension (=entrained flow

), apparatus or plant

total

patents 51 33 159 47

54 By Winkler technique / fluidization

317 3% 1% 13% 4%

56 By Winkler technique / fluidization, apparatus or plants

151 1% 1% 13% 13%

Table 6.5 Citations made by patents on fluid bed gasification to patents on other gasifier types within the Fl2 dataset, as percentage of total citations made.

IPC subgrou

p Description

Total citation

s

Cited subgroups

2 20 46 48

54-56 By Winkler technique / fluidization

2125 2% 1% 8% 2%

146

6.5.2 Combustion and pyrolysis As a next step we will study the relatedness of similar technologies. Several thermochemical processes exist that use fluidized beds to break down biomass under high temperature, see textbox 6.2: combustion, pyrolysis and gasification. According to literature these technologies showed some level of proximity (shared technology and knowledge base) that might have facilitated synergies. Hundreds of combustion plants have been erected, largely by Foster Wheeler / Ahlstrom and Combustion Engineering that licensed Lurgi technology (Koornneef, Junginger and Faaij, 2007; Watson, 1997). Both designs were subsequently adapted for gasifiers that were first applied in the early 1980s (Basu, 2006). Also Foster Wheeler’s pressurized circulating fluidized bed (CFB) technology as applied in the successful Värnamo plant was based on its Pyroflow coal combustion process (Kaltschmitt, Rösch and Dinkelbach, 1998). The combustion technology provided many potentially relevant learning experiences for gasification technology, with respect to biomass as fuel, CFB technology and system integration. Pyrolysis attracted less attention than combustion or gasification. Fast pyrolysis especially received attention in the late-1980s and early-1990s. Later on, after 2000, pyrolysis has been applied by several manufacturers as initial step before gasification (Kirkels, 2014). The spillover from pyrolysis to gasification or vice versa has not been made explicit in literature, as far as we know of.

Textbox 6.2 Themochemical processes that can convert biomass under high temperature.

Combustion: takes place in the presence of oxygen (air) by which the fuel is converted to carbon dioxide and water and energy is released. Gasification: takes place when limited oxygen is supplied. Main product is a gas consisting of carbon monoxide and hydrogen that can be used for further applications. Pyrolysis: ‘burning in the absence of air’ (Heermann, Schwager and Whiting, 2001, p. E22) or, in other words, the thermochemical breakdown of biomass in the absence of air. End products can be a solid, liquid and gaseous fraction. End product composition is very much dependent on temperature, heating rate and residence time: production of charcoal requires a slow reaction at low temperature, while fast pyrolysis is a very fast process with high heating-rates that mainly results in liquid products.

We identified relevant classification and patent sets in the Fl2 dataset, from which we derived companies involved and periods of most relevance. We assessed interrelatedness by checking on co-classification and cross-citation. In addition, we cross-checked between literature and the datasets in order to clarify the role of companies involved in biomass pyrolysis. Results on co-classification (table 6.6) and cross-citation (table 6.7) show the

147

relevance of combustion and pyrolysis technologies, especially of fluidized bed combustion (Foster Wheeler and Ahlstrom) and waste incineration (Ebara). Patents on combustion are cited by Fl1 patents on fluid bed gasifiers over a broad time span (1984-2005), but mainly over the period 1990-1997. Patents on waste incineration are mostly cited over the period 1996-2004, which relates to the period of intense developments. Additional research showed that patents on fluidized bed combustion and incineration of waste are relatively often co-classified (22-38% of patents) and the same is true for patents on carbonization and cracking/pyrolysis (27-28% of the patents). In our dataset Exxon holds most patents on carbonization and cracking/pyrolysis. This made us wonder what the specific contribution was of companies working on biomass pyrolysis. We did a reverse search by identifying from literature some of the more prominent companies involved in biomass pyrolysis - Thermoselect, Pyrovac, Pyrotech, Alten, Enerkem, Ensyn (Beenackers and Bridgwater, 1989; Bridgwater, 1997; Heermann et al., 2001) - but none of them stands out in our datasets. We conclude that they had little influence.

Table 6.6 Co-classification of patents on different thermochemical processes within the Fl2 dataset, as percentage of total amount of patents in C10J 3/54-56.

IPC (sub) group & description

F23C 10 Combustion

fluidized bed

F23G 5 Incineration of w

aste or low

grade fuels

C10B 47-57 Carbonizing or coking process; destructive distillation

C10G 1 Cracking: production of liquid hydrocarbon m

ixtures from

wood and coal

total

patents 330 191 172 167

C10J 3 /54-56

Production of carbon monoxide and hydrogen from solid fuel by partial oxidation by Winkler technique / fluidization

385 12% 5% 9% 7%

Table 6.7 Citations made by patents on fluid bed gasification to patents on combustion or pyrolysis within the FL2 dataset, as percentage of total citations made.

Total

citations

Cited (sub) groups IPC subgroups Description

F23C 10 F23G 5

C10B 47-57

C10G 1

C10J 3 /54-56

Production of carbon monoxide and hydrogen from solid fuel by partial oxidation by Winkler technique / fluidization

2125 16% 12% 6% 5%

148

6.5.3 Feedstock A variety of gasifiers has been developed for different types of feedstock. It is not clear to what extent their developments are strongly related or not. Biomass is not clearly defined, it can refer to a wide variety of fuels. Patents hardly refer to biomass. They do refer to carbonaceous fuels (a naming hardly used in literature) – which is also an ill-defined category and seems to encompass a variety of biomass and waste fuels. Therefore we need to be more specific regarding feedstock. Wood is most applied in advanced gasifiers (Kirkels, 2014), but many other fuels received interest and some developed in specific niches; peat, black liquor, (municipal) waste, agricultural wastes, sludge and rice husk – see Kaltschmitt et al. (1998), Kirkels and Verbong (2011), and Reed and Gaur (2001). We selected several of these fuels that received significant interest – by our judgement. Peat gasification has received significant interest in the USA and Finland (a.o. by VTT) and was demonstrated in the Kemira Oy plant in Finland in the late-1980s (Kaltschmitt et al., 1998; Kirkels, 2014; Kirkels and Verbong, 2011). Black liquor is a highly-corrosive byproduct of the paper industry, a lignin-rich mixture of cooking chemicals and dissolved wood material. Development of black liquor gasification has been pursued by Asea Brown Boveri (ABB), Manufacturing and Technology Conversion International (MTCI) and Chemrec/Kvaerner (the latter is of the entrained flow type) (Kaltschmitt et al., 1998). Waste gasification is characterized by a variety of feedstock, technologies and manufacturers. A market study by Juniper Consultancy Services provides an overview of technologies and trends, see Heermann et al. (2001). Waste gasification technologies in general focus more on the production of manageable secondary waste products (solidified ash) than on energy recovery (Morris, Waldheim, Faaij and Stahl, 2005). The technology received a lot of attention in Japan and Germany (Kirkels and Verbong, 2011). We also did take into account coal – although obviously not a type of biomass - as fluid bed technology was originally developed for coal (Basu, 2006; Harmsen, 2000; Kaltschmitt et al., 1998; Longwell, Rubin and Wilson, 1995) and several of the biomass gasifiers were adaptations of coal gasifiers (IGT’s Renugas technology) or coal combustors (Lurgi) (Basu, 2006; Reed and Gaur, 2001). In literature there are claims of (potential) spillover between coal and biomass technology based on similarities in general gasifier designs, hot gas cleaning and (foreseen) applications (Babu, 1995; Babu and Whaley, 1992; Kirkels and Verbong, 2011). As there is no IPC classification relating to feedstock, we identified relevant patents by key word search in title and abstract (keywords as specified in table 6.8). From this we identified number of patents, relevant periods and companies. In addition, we checked to what extent patents refer to multiple feedstock.

149

Coal is most referred to in both datasets (respectively 76% and 71% of the patents that contain a reference to feedstock), followed by carbonaceous fuels (resp. 44% and 34%). Both are mainly referred to in the late-1970s and early-1980s by Exxon. Especially early on coal was the dominant feedstock: over the period 1976-1980 65% of Fl2 patents referring to feedstock did refer to coal; while over the period 2000-2008 this was only 31%. Other feedstock is much less referred to. The gasification of peat is mainly patented in the early-1980s by IGT; the gasification of black liquor is mainly patented during the 1990s by MTCI; and the gasification of municipal waste by Ebara. Biomass and wood are patented by a variety of companies, but these are not in line with relevant companies identified from literature – see section 4. Therefore we conducted a reversed search by company name in the Fl2 dataset16. It showed that these ‘biomass’ companies: a) do not refer more often to feedstock; and b) do refer mostly to coal and carbonaceous fuels.

Table 6.8 Reference in title/abstract to specific types of feedstock

Fl1 dataset (280 patents refer to feedstock) Fl2 dataset (1021 patents refer to feedstock)

Total patents

Dominant period

Dominant companies (# patents)

Total patents

Dominant period

Dominant companies (# patents)

Coal

212 1976-1984

Exxon (39) US DoE (13)

720 1978-1984

Exxon (55) US DoE (36) Shell (20)

Carbonaceous fuels

124 1976-1984

Exxon (29) IGT (7) US DoE (7)

351 1976-1990

Exxon (45) Texaco (28) US DoE (17)

Biomass

29 2003-2005

MTCI (7) Ebara (3)

93 1996-2003

MTCI (7) IGT (5)

Wood

23 1982-1989

Framatome (2) 68 --- Shell (3) Energy Products of Idaho (3)

Peat 18 1980-1982

IGT (4) 64 1980-1983

IGT (4)

Black liquor

10 1994-1997

MTCI (5) Combustion Engineering (3) Babcock & Wilcox (2)

31 1990-1997

MTCI (5) Combustion Engineering (3) Rockwell Int. Corporation (3) Babcock & Wilcox (3)

Municipal waste

7 --- Ebara (3) 30 Ebara (3) Ebara / Nikkiso (3) Agency Ind. Science & Tech. (3)

16 Company names: Ahlstrom, Batelle, Ebara, Foster Wheeler, Framatome, IGT, MTCI, Metallgesellschaft

150

We also checked on reference to multiple feedstock – see table 6.9. Half of the patents on carbonaceous fuels also refer to coal. A significant amount of patents on other feedstock also refers to coal and/or carbonaceous fuels. It is an indication that developments are strongly related. Only patents on black liquor stand out as a separate development. Given the strong overlap in subsets we did not construct a cross-citation matrix, as it would be hard to draw meaningful conclusions.

Table 6.9 Share of patents that refer to multiple feedstock in the FL2 dataset as percentage of patents in column 2.

# patents ref. to feedstock

coal carbonac. biomass wood peat black liquor

municipal waste

coal 720 --- 24% 4% 4% 5% 0% 2% carbonaceous 351 50% --- 4% 4% 5% 1% 1% biomass 93 32% 16% --- 27% 9% 5% 6% wood 68 41% 22% 37% --- 15% 6% 7% peat 64 55% 30% 13% 16% --- 0% 2% black liquor 31 0% 6% 16% 13% 0% --- 0% municipal waste 30 40% 17% 20% 17% 3% 0% ---

We conclude that although the term ‘biomass gasification’ demarcates a field in literature, this is not the case in US patents. Coal is the most important feedstock for fluid bed gasifiers. Instead of ‘biomass’ the term ‘carbonaceous fuels’ is often used. Many patents refer to multiple types of feedstock. Therefore fluid bed gasifiers qualify as fuel-flexible technology – a characteristic which is well recognized in literature. This is not true for patents on black liquor gasification that thereby qualify as a separate development. 6.5.4 End use A gasifier is configurational technology: the gasifier, gas cleaning unit and final conversion step together determine whether the plant runs efficient and problem-free. Different end use applications have been foreseen for biomass gasifiers: biofuels (e.g. Fischer-Tropsch diesel, ethanol and methanol); high-efficiency power generation using gas turbines and steam generators (IGCC); Synthetic Natural Gas (SNG); and in fuel cells (Kirkels and Verbong, 2011). Kirkels (2014) showed that the considered applications did influence the direction of technological development. He also suggests a possible indirect link between applications by the basic requirement to produce (ultra) clean syngas that can serve multiple applications.

151

For all applications mentioned above we identified relevant classification per technology17, selected the relevant patents from the Fl1 and Fl2 datasets and identified companies involved and periods of development, see table 6.10. It shows the strong relevance of steam technology between 1984 and 1996, a development that is dominated by Foster Wheeler. Fuel cell technology did have a strong influence over the 1998-2008 period. Despite the strong interest in methanol production and more recently biofuels this does not show in our datasets. The latter does not imply that these are not relevant. It might also be the case that these end-use technologies are not closely related to the development of biomass gasifiers, that they developed more independently; or that these technologies were already mature – e.g. Kirkels (2014) suggests that the conversion to methanol already was considered a proven technology in the early-1980s.

Table 6.10 Patents on end-use applications Fl1 dataset Fl2 dataset

End-use application & classification

Total patents

Dominant period

Dominant companies (# patents)

Total patents

Dominant period

Dominant companies (# patents)

Steam and steam turbines (F22B)

16 1991-1996

Foster Wheeler (8)

181 1984-1996

Foster Wheeler (51) Ahlstrom (14) Ebara (13)

Gas turbines (F02C) 13 --- --- 105 1976-2003

Texaco (7) General Electric (7)

Chemicals to electricity, fuel cells (H01M)

10 2003-2008

--- 94 1998-2005

Energy Research Corp. (9) General Motors (7) University of California (6) Motorola (6)

Alcohols (C07C 27&29) 4 1981-1982

Exxon (2) 42 1976-1989

Imperial Chemical Ind. (13) Metallgesellschaft (8) Phillips Petroleum Comp. (7)

Generation hydrocarbons from CO and H2 (C07C 1/00 – 1/12)

6 1976-1979

Exxon (3) 38 1976-1985

Union Carbide (4) Exxon (3)

Synthetic Natural Gas (C10L 3/06 & C10K 3)

7 --- --- 27 1976-1985

17 SNG is coupled to two subgroups: C10L 3/06 and C10K 3. Class C10K 3 is on modifying the chemical composition of combustible gases to produce an improved fuel while reducing the carbon monoxide content. As such, the class is relevant for SNG production, but might also be relevant for some of the other applications.

152

Table 6.11 Amount of patents that relates to multiple applications as percentage of patents in column 3 (=patents table 10 Fl2 + ego-1).

steam gas

turbine chemicals electricity

methanol / ethanol

hydro carbons SNG

F22B F02C H01M

C07C 27&29

C07C 1/00-1/12

C10L 3/06 & C10K 3

patents 543 374 219 121 165 133 steam and steam turbines

F22B 543 -- 24% 2% 1% 3% 3%

gas turbine F02C 374 36% -- 23% 7% 7% 7%

chemicals to electricity, fuel cells

H01M 219 5% 39% -- 9% 10% 14%

methanol/ethanol C07C

27&29 121 5% 23% 16% -- 26% 26%

generation hydrocarbons from CO and H2

C07C 1/00-1/12

165 11% 16% 13% 19% -- 30%

SNG C10L

3/06 & C10K 3

133 11% 21% 23% 23% 38% --

In order to reveal possible indirect links we selected all the patent sets related to different end-use applications (table 6.10 Fl2) and expanded these by snowballing (ego-1 network including both patents that cite or are being cited). We assume that by this procedure also patents on gas cleaning will be included; and that these are especially relevant patents as they might be shared among applications. We assessed the overlap between datasets, see table 6.11. The thermochemical applications (steam boilers and gas turbines) share a significant amount of patents; while the same is true for the chemical technologies (production of alcohol, SNG and hydrocarbons). However, another explanation might be simultaneous developments: SNG and hydrocarbon production that both received attention during the 1970s and early-1980s; gas turbine and chemicals-to-electricity technology that both developed from the late-1990s onwards. We checked three sets of related technologies in more detail: steam and gas turbines; gas turbines and fuel cells; and SNG and hydrocarbon production. At the junction between steam and gas turbine technologies mainly four companies operate: Foster Wheeler, Ebara, Metallgesellschaft and MTCI. Patents mainly relate to gasification, fluidized bed combustion and incineration of waste and low grade fuels. At the junction between gas turbine technology and fuel cells mainly MTCI is active and a wide variety of other companies. MTCI is known to have developed a steam-reforming technology that is capable of processing a wide spectrum of feedstock to produce hydrogen-rich, medium caloric value gas (Reed and Gaur, 2001). Patents

153

mainly relate to gasification, hydrogen production and fuel cells, e.g. large-scale stationary applications of fuel cells based on the gasification of coal. This explains the link with gas turbine technology. However, the broader set of patents related to fuel cells as reported on in table 6.10 also relates to automotive applications - which explains the involvement of car manufacturers like General Motors. At the junction between SNG and hydrocarbon production Exxon, Iron Carbide and TRW are active and a wide variety of other companies. Patents mainly relate to gasification, generation of hydrocarbons, production of hydrogen and the modification of the chemical composition of the gas. To conclude, patents on different end-use applications in our datasets are limited in number. They mainly share specific gasifier technology and not so much gas cleaning technology. 6.6 Technological trajectories Finally we identified for both datasets the technological trajectory by the approach followed by Verspagen (2007). In addition we assessed their neighborhoods – the part of the network that the top path relates to: developments that patents on the top path draw upon or contribute to; developments that therefore can be considered to be represented by the top path. For the largest components of both datasets18 we calculated the top paths using Verspagens Citpath software by both SPLC and SPNP leading to identical results. For the top paths we identified the companies involved, feedstock and processes (gasification or combustion), see table 6.12. We also assessed the part of the network which is close to the top path - the part of the network that is well represented by the top path. We decided to cut-off at distance 2 (ego-2 network of the top path or snowballing twice): it is still fairly close to the top path and it results in two sufficiently large datasets to compare on companies and technologies involved, see tables 6.13 and 6.14.19

18 Largest component is the largest set of connected patents in the dataset: for Fl1 75% of all patents (291 patents, mainly due to the presence of 83 isolates in the dataset – patents that don’t cite other patents nor get cited); for Fl2 99% of all patents are in the largest component. 19 We assessed the size of ego-networks at multiple distances (e.g. ego-1, ego-2, etc.) for both datasets: 11-14% of the patents are either at the top path or at distance 1; 39-40% at distance 2; and this gradually built up until at distance 5 almost the complete networks are covered.

154

Table 6.12 Top paths of the Fl1 (left) and Fl2 (right) dataset. Processes identified based on IPC classification, see section 6.5.2.

Feedstock identified from title/abstract, see section 6.5.3.

year patent company process feedstock year patent company process feedstock 1976 3932146 Exxon gas coal, carbonaceous

material 1976 3957459 Exxon gas coal 1976 3975168 Exxon gas coal, carb. material 1977 4026679 Stora gas carb. materials

1978 4094650 Exxon gas coal, carb. material 1978 4103646 EPRI comb coal 1982 4315758 IGT gas coal, carb. material 1981 4259911 CE coal 1990 4929255 Ahlstrom gas carb. materials 1984 4469050 York gas, comb -

1990 4969930 Ahlstrom gas, comb carb. materials 1986 4594967 FW comb - 1994 5284550 CE gas, comb black liquor 1987 4682567 FW comb - 1997 5624470 CE gas black liquor 1988 4741290 Steinmuller comb -

1990 4929255 Ahlstrom gas carb. materials 1990 4969930 Ahlstrom gas, comb carb. materials 1992 5095854 FW comb -

1994 5281398 Ahlstrom comb - 1996 5505907 Ahlstrom - biomass, peat, coal 1997 5620488 Ebara gas, comb, inc coal, (waste)

1999 5900224 Ebara gas coal, biomass, (waste) 2001 6283048 Ebara gas, inc coal, (waste) 2002 6401635 Corenso inc (liquid carton waste)

Table 6.13 Contrasting patents in the neighborhood of the Fl1 top path (left) with those that are not (right).

Fl1 ego-2: companies # patents Fl1 not ego-2: companies # patents EXXON 29 EXXON 23 IGT 11 USA DOE 10 USA DOE 7 METALLGESELLSCHAFT 8 BABCOCK & WILCOX 4 FOSTER WHEELER 7 AHLSTROM 4 EBARA 7 FOSTER WHEELER 4 CHEVRON 7 COMBUSTION ENGINEERING 4 EBARA, NIKKISO 7

Fl1 ego-2: classification sub

class count Fl1 not ego-2: classification

sub class

count

Carbon monoxide from solid fuels C10J 201 Carbon monoxide from solid fuels C10J 516 Chemical / physical process B01J 60 Chemical / physical process B01J 142 Combustion F23C 35 Production of hydrogen C01B 68 Regeneration pulp liquors D21C 15 Cracking of wood and coal C10G 60 Incineration of waste and low grade fuels F23G 12

Incineration of waste and low grade fuels F23G 59

155

Table 6.14 Contrasting patents in the neighborhood of the Fl2 top path (left) with those that are not (right).

Fl2 ego-2: companies # patents Fl2 not ego-2: companies # patents FOSTER WHEELER 66 EXXON 65 EXXON 29 TEXACO 42 A. AHLSTROM 26 USA, DoE 31 IGT 26 SHELL OIL COMPANY 20 EBARA, NIKKISO 20 SRP 687 PTY 19 TEXACO 20 METALLGESELLSCHAFT 18 METALLGESELLSCHAFT 20 PHILLIPS PETROLEUM 16

Fl2 ego-2: classification sub

class count Fl2 not ego-2: classification

sub class

count

Carbon monoxide from solid fuels C10J 659 Carbon monoxide from solid fuels C10J 648 Combustion F23C 598 Chemical / physical process B01J 623

Chemical / physical process B01J 493 Separation of gasses and purification

B01D 456

Incineration of waste and low grade fuels

F23G 362 Cracking of wood and coal C10G 399

Boilers and steam generation F22B 203 Production of hydrogen C01B 383

Although we expect both paths to represent the technological trajectory of fluidized bed gasifiers, both paths only share two patents and relate to rather different developments. Apparently fluid bed gasification is not characterized by a strong dominant development, as else we would expect this development to be present in both top paths. Both top paths show the relevance of coal gasification and of combustion technology, but they also differ in many ways. The Fl1 top path represents a patchwork of developments: Exxon’s catalytic and steam-blown gasification of coal to produce methane under high pressure; an improvement on IGT’s U-gas technology (pressurized coal gasification under high temperature); both Ahlstrom patents relate to circulating fluidized bed technology; and Combustion Engineering that patented atmospheric CFB gasification of black liquor (recovery boiler). The top path neighborhood (table 6.13) shows that the development is also related to Babcox Wilcox’s efforts on black liquor gasification, but not to development efforts by Ebara and Metallgesellschaft. The Fl2 top path is much more detailed (encompassing more patents) as a result of the larger underlying patent set that showed higher connectivity and it shows a more narrow focus of developments. The top path and its neighborhood (table 6.12 and 6.14) shows the presence of Foster Wheeler, Ahlstrom, IGT and Metallgesellschaft, but it also includes patents by Framatome and Studsvik – all companies that are known for their contribution to biomass gasification. The top paths reflect their involvement over time: early on Exxon holds many patents (1976-1982); from 1983 to 1997 several companies are involved, although between 1990 and 1997 Foster

156

Wheeler is dominant; and from 1999 to 2005 Ebara is dominant (in line with findings in previous sections). The neighborhood strongly relates to fluidized bed combustion, incineration of waste and low grade fuels, and to boilers and steam generation. The non-ego-2 dataset, the part of the network that is not directly related to the top path, shows a dominant presence of Exxon, Texaco and Shell. Citations show that they strongly draw upon each other’s work. Metallgesellschaft is also present in this dataset – the company held a broad technology portfolio (Heermann et al., 2001). This part of the network strongly relates to purification of gasses and exhaust gas treatment; cracking of solid fuels, carbonizing and cocking (mainly before 1986 by Exxon); and patents related to different end-use applications (fuels cells, gas turbine technology, biofuels- related to subclass C07C) - although it is not clear whether they are related to the Exxon developments or constitute a rather separate development. To conclude, both top paths show limited similarity. The Fl1 top path reveals a patchwork of developments. The Fl2 top path shows a more narrow focus and incremental innovation and therefore represents better the characteristics of a technological trajectory. 6.7 Conclusions and discussion 6.7.1 How did fluidized bed gasifiers develop? We studied technological development in detail by patent landscaping (section 4) and by specifically assessing the interrelatedness of technological developments between subsets of patents (section 5). Based on our findings we reconstructed a timeline of dominant developments, see figure 6.3. Especially early developments, 1976-1985, are characterized by high patenting intensity. These developments are dominated by Exxon that mainly worked on coal conversion (gasification, cracking and destructive distillation). The period 1981-1997 is characterized by co-evolution of biomass gasification and combustion, mainly by Foster Wheeler and Ahlstrom. Especially gasification patents over the period 1990-1997 draw upon earlier combustion patents. This development is strongly related to steam turbine technology and to some extent also to gas turbine technology. It shows strong links with the later development of waste gasifiers and incineration, which is dominated by Ebara (1998-2002). Most recently interest shifts towards fuel cells and hydrogen production. The influence of other technologies is more modest and restricted to specific periods: cracking or carbonization of coal; fixed bed gasifiers (mainly influencing developments in 1984-1988 and 2004-2005); and entrained flow gasifiers (after 1996). Most important end-use applications include steam turbines, gas turbines and fuel cells. Patents on end-use technologies mainly relate to similar fields: thermochemical

157

conversion technologies (steam turbines, gas turbines) or chemical technologies (production of alcohol, SNG, hydrocarbons). With respect to feedstock we showed that instead of ‘biomass’ the term ‘carbonaceous fuels’ is often used in patents. Coal is the most referred to feedstock. Many patents refer to multiple types of feedstock, e.g. coal and carbonaceous fuels. This is also the case for patents that are held by companies that are recognized in literature for their contribution to biomass gasification. While literature seems to have evolved around the field of ‘biomass gasification’, we conclude that patenting evolved around ‘fuel-flexible fluid bed technology’. Only black liquor gasification stands out as a somewhat separate development.

Figure 6.3 Overview of developments in fluid bed technology based on this research. Year represent years of patent publications, which on average over our datasets has a time lag of

three years with the priority date. 6.7.2 Influence on technological trajectories The next question to answer is how interrelated developments influenced the emerging and unfolding of trajectories. We followed the Verspagen (2007) approach to reconstruct technological trajectories. For the FL1 datasets that only contains patents on fluid bed gasification the top path revealed a patchwork of subsequent developments (by Exxon, IGT, Ahlstrom and CE) which did not had a strong technological focus. The Fl2 dataset - that also includes patents on interrelated technologies - shows a top path that mostly reflects development work by Foster Wheeler, Ahlstrom and Ebara. It has a much more narrow technological focus as it mainly relates to gasification, combustion and incineration. That combustion and incineration are central in this top path confirm the importance of these interrelated technological development for the field, while the interaction between combustion

1976 1985 1995 2005 2008

1976-1985 Exxon coal gasification

pyrolysis: cracking, destructive distillation

2003-2008 fuel cells hydrogen

production

1981-1997 combustion, gasification

steam generation (1991-1996) Foster Wheeler, Ahlstrom and others

1980-1982 IGT

peat

1994-1997 black liquor

MTCI, CE, B&W

1998-2002 waste

incineration Ebara

158

and gasification patents on the top path suggest that the path creation was a co-evolutionary process – which is confirmed by the findings in section 5. The conclusions of the previous section add to this in several ways: it shows that developments on the top path relate to and can be considered a representation of a broader patent set; it shows parallel developments which are not represented by the top path; and it shows to what extent these parallel developments are related or not. It also shows that initial conditions are constructed, as different developments draw upon different prior technologies. For example, the Fl2 top path builds upon early patents on combustion technology, while at that time the efforts by Exxon on coal conversion were dominant. Based on these findings we argue that there are different types of top paths, see figure 6.4: path 1 relates to a homogenous technological development and as such shows strong focus and incremental innovation; path 2 also has a strong focus but draws upon two interrelated technologies; while path 3 links different technological developments. Fl1 top path is a type 3: it connects developments that subsequently were patented over time, all interrelated, but also showing a rather different technological focus. The Fl2 top path is a type 2: it starts with a co-evolution of air blown CFB combustion and gasification and subsequently includes technologies of waste incineration and gasification – technologies that are closely related, as we showed. By its stronger technological focus and longer path it shows a better representation of incremental innovation, which is an essential characteristic of technological trajectories.

Figure 6.4 Top path as representation of technological trajectories can relate to more homogeneous technological development (1), or to more heterogeneous developments by

connecting several clusters (parts of the network that show high connectivity) thereby relating to different technologies (2, 3).

This distinction of different types of top paths has both methodological and theoretical relevance. It is methodologically relevant as it shows that the choice for a specific analytical lens and scope (choice on system boundaries) determines which

1

2

3

159

developments are reconstructed. This is of special relevance in case of interrelated developments. It also holds a theoretical relevance, as it shows two different mechanisms of path creation: it shows how engineers for the development of technologies draw upon similar technologies, but that these developments are embedded in a much broader technological context. We made this explicit for our top paths. The FL2 top path relates to a distinctive technological developments; its neighborhood (ego-2) relates to 40% of the patents – not only those held by the dominant companies in the top path; and within 5 steps of citing/being cited the complete network is covered, including loosely related developments by Exxon, Texaco and Shell. This is more broadly known as the small world phenomenon. We realize that this pattern reflects as much the process of path creation as well as the nature of this specific technology – characterized by a large variety of technologies and applications and a high level of technological interrelatedness. One could argue that the large influence of loosely related technologies is mainly due to the immaturity of the technology of biomass gasification; that the technology is still in a phase of exploration and variation. To check upon this we did two quick scans on more mature technologies: one on entrained flow gasification that is used for coal (subgroups C10J 3/46-48) and one on fluid bed combustion that is used for a variety of feedstock (F23C 10 – as far as covered by our Fl4 dataset). We did so for the period 1976-2003, as well as for the period 1990-2003 – over the latter period both technologies can be considered mature (Harmsen, 2000; Watson, 1997)20. Our findings show also for these technologies high levels of interrelatedness. Therefore we conclude that this is a characteristic of the field and not just due to the immaturity of technology. The policy implication of our findings is that the focus should not be too much on a specific technology, but more on broader technological fields of interrelated technologies: support for a broader field might better ensure continuous developments; and technological proximity might provide a good starting point for uptake of a technological trajectory, as well as provide synergies for further development. However, we do see practical constraints. We showed that initial conditions are constructed: the large efforts on coal entrained flow gasification in the

20 Over the whole period entrained flow patents show only 10% internal citations, 55% citations outside the dataset and 35% citation < 1976. For patents after 1990 these were respectively 2% internal, 26% outside and 72% prior to 1990. Over the whole period fluidized bed combustion patents show 37% internal citations, 41% citations outside the dataset and 22% citations < 1976. For patents after 1990 these were respectively 15% internal, 5% outside and 79% prior to 1990.

160

early 1980s hardly influenced the development of fluid bed technology at that time; but did influence developments after 1996. It illustrates the complexity of understanding (or even more difficult: building upon) this interrelatedness. 6.7.3 Discussion Our initial concern was that the findings would be affected by our choice for US patents while trying to uncover global developments. Our findings over this research confirm the validity of this approach: most leading manufacturers known from literature could be traced back. However, the question whether the overrepresentation of US efforts in the patent set did effect the network structure remains unanswered. Finally, the operationalization by Verspagen involves a rather narrow focused reconstruction of a path – a sequence of interrelated patents. However, while discussing technological trajectories Nelson and Winter (1982, p258/259) emphasize the incremental innovations taking place around a dominant design; and Dosi (1982, p154) sees progress as the movement of multi-dimensional trade-offs among relevant technological variables. The shared focus and interrelatedness of developments in clusters of patents suggest that also these clusters qualify as technological trajectories. Paths hold the advantage of a clear storyline of how developments unfold; but it remains unclear what developments they actually represents – see figure 6.4. Clusters are much more difficult to understand, as they have a complex internal structure and include a variety of related developments. Their strength is that by definition they do have a technological focus. Also, the larger number of patents in a cluster and their citations allow to better represent and understand the relatedness or even overlap with other technological developments. Acknowledgements I would like to thank Floortje Alkemade for her constructive feedback; Önder Nomaler for sharing his interest in network analysis and practical support; Bart Verspagen, Koen Frenken and two anonymous reviewers for constructive discussions and comments on a prior version of this chapter.

161

162

Chapter 7

Conclusions The aim of this thesis is to understand the development of biomass gasification and the dynamics of this process. In this chapter I draw conclusions and answer the research questions I posed in chapter 1. The first question is how did the technology of biomass gasification develop and how can this development be explained, what were the underlying driving forces? The second research question is, did this result in emerging trajectories and what did a path dependency and a path creation approach contribute? In section 7.1 I combine my findings to characterize the development of biomass gasification and conclude on the underlying driving forces. In section 7.2 I present the emerging trajectories in biomass gasification and discuss the contribution of the various approaches. Section 7.3 described methodology issues and the implications of the research.

163

7.1 How did biomass gasification develop over time? 7.1.1 How has biomass gasification developed? This is the main question that I address: in chapter 3 I focus on the broad field of biomass gasification with an overview of technologies, applications and their status; in chapters 4 and 6 I focus on the technological development of advanced gasifiers. Here I discuss the dynamics of biomass gasification, the technologies and applications, and finally the development of advanced gasifiers. Initial efforts originate from the late 1970s in response to the concerns raised by the oil crises. Both fixed bed gasifiers as well as advanced gasifiers were considered for development and application. The advanced gasifiers made it possible for the first time to consider biomass as a modern energy carrier, for large scale and advanced applications. Interest lasted till the mid-1980s. By that time the oil prices had dropped and stabilized. Renewed interest started in the 1990s, as shown by scientific papers, conference publications, and development and demonstration efforts. This continued to grow after 2003, demonstrated by an increase in patents (see chapters 3 and 5). Mainly the US, European countries and the EU were involved, although after 2000 Japan and China became strong contenders. The development effort covered the exploration and application of a variety of technologies, see table 7.1 and figure 7.1. With the benefit of hindsight, I conclude that certain technologies were suited for specific feedstock and applications–although this was not always recognized upfront. What was considered the preferred technology was time and context dependent. This variety of technologies proved to be interrelated: by sharing a technology and knowledge base, sometimes involving the same actors, by sharing networks and the same policy sphere. Advanced gasifiers were the figurehead in the field‒their foreseen application raised high expectations and resulted in significant policy support over the years. Their development and demonstration were thanks to transnational efforts. The focus shifted over time: from methanol production in the late 1970s and early 1980s, to IGCC during the 1990s, and biofuel production after 2003. Efforts to develop and demonstrate gasifiers for methanol production and IGCC evolved over waves of interest, showing hype-disappointment dynamics.

164

Table 7.1 Biomass gasification: a summary (see chapter 3).

Preferred technology

• Updraft (small, mainly heat) • Downdraft (small, mainly power)

• Circulating fluid bed (large) • Entrained flow (large, fuels & chemicals)

Main applications (niches)

• Heat • Combined heat and power • Co-combustion • Rural electrification / developing

countries

• IGCC (research) • Fuels & chemicals (research) • Waste gasification

Scale 0.05 – 10s MWth Dominant suppliers

Multiple

Dominant countries

USA, Finland, Sweden, Germany, Austria; Japan (waste); China, India (small scale)

Figure 7.1 Overview of developments in fluid bed technology, as used for larger scale

and more advanced gasifiers (see chapter 6) The interest in biofuels grew more gradually but also extended to larger actor groups and had greater momentum. Initially, in the late 1970s and early 1980s, researchers, governments, and companies were involved in gasification. Over the 1990s, the community broadened and started to involve more heterogeneous actors, including utilities and feedstock companies. After 2000, the automotive and petrochemical industries got involved, the latter mainly in first generation biofuels and biochemical conversion routes. Other gasification technologies developed parallel to these dominant trends, but to a lesser extent: waste gasification, black liquor gasification,

1976 1985 1995 2005

1976-1985 Exxon coal gasification

pyrolysis: cracking, destructive distillation

2003-2008 fuel cells hydrogen

production

1981-1997 combustion, gasification

steam generation (1991-1996) Foster Wheeler, Ahlstrom & others

1980-1982 IGT

peat

1994-1997 black liquor

MTCI, CE, B&W

1998-2002 waste

incineration Ebara

165

and gasifiers in combination with fuel cells. While the literature (and the field) evolved around biomass gasification (see chapters 3, 4, 5), engineering efforts focused on fuel-flexible fluid bed technology (chapter 6). Overall I conclude that the development of biomass gasifiers is diversified, involving many different technologies, designs, and manufacturers. Companies typically use gasifiers for specific feedstock and applications. Over longer time periods this has not led to a dominant design. I qualify it as a niche development–a phase in the development of a technology characterized by continuous development, demonstration, and limited diffusion. I conclude that the focus on specific applications (methanol production, IGCC and biofuels) provided momentum and direction in the search for technologies to be applied and problems to be solved. Especially early on in each period, a wide variety of technologies was promoted (variation). Only over time did the focus became stricter and efforts aligned when governments pre-selected technologies to support and pursue in their RD&D programs. In the course of the 1980s and the 1990s, this resulted in emerging trajectories, which I discuss in greater detail in section 7.2. 7.1.2 How can these developments be explained? I have described a variety of driving forces. Some are external and beyond the control of stakeholders involved in the development of biomass gasification. These forces are often not specific to this technology, but generated a window of opportunity for a variety of technologies. Some are strongly linked to technological functionality and maturity while others relate to the social dimensions of development: visions and hype-disappointment mechanisms. These various driving forces are discussed below. Largely exogenous forces had a very strong influence and include: • Socio-economic and environmental challenges High oil prices, concerns regarding

depletion of oil, and climate change concerns. Over each period, these challenges were the strongest drivers as the timing and momentum resulted in commitments to the general field of biomass gasification and specifically more advanced applications. These challenges also brought guidance, because the engineering community translated them into technical requirements (chapter 4). The challenges changed over time as concerns regarding oil prices and depletion fluctuated and concerns regarding climate change grew. Prioritizing which challenges to address and how to address them shifted over time. Both strongly affected long-term developments due to discontinuation and refocusing of efforts.

166

• National policies National governmental support was important for development. Gasifiers are mainly developed by countries with a large feedstock base, with industries that can develop and produce gasifiers or that have beneficial end-use markets–and preferably a combination of all these (chapters 2, 3). Efforts by the US, EU, Finland, and Sweden aligned: they simultaneously supported RD&D on gasifiers for methanol production, IGCC and biofuel production. This provided momentum to the field and it defined a dominant paradigm to work within (chapter 5). In addition, governments actively supported innovation processes (e.g. by formulating or supporting visions; supporting networking activities and thereby sharing learning experiences). Although the power plant suppliers and process industry typically serve an international market, I show that governments mainly support national product champions. This resulted in minor variations within the dominant paradigm at the global field level (chapter 4).

• (Lack of) competition from other energy technologies The loss of support for oil based applications during the 1970s, and for coal power and nuclear power in the late 1980s and early 1990s provided the window of opportunity for biomass gasification. This reversed around 2000 and started to affect developments: the competition from natural gas based combined cycles, wind turbines and (co-) combustion of biomass made it unattractive to continue pursuing the biomass IGCC route (chapters 2, 3, 5). Although competition is often perceived as a market mechanism, restrictions in RD&D budgets also forced policy makers to decide which technologies to support and their focus changed over time. In the case of EU policy, this initially supported biomass gasification in the late 1970s and early 1980s. The focus shifted to biomass pyrolysis in the late 1980s and early 1990s, then back to biomass gasification in the mid-1990s. I also showed a strong shift in support for broader technological fields: from biochemical conversion technologies over the 1980s towards thermochemical conversion technologies in the 1990s (chapter 2).

Forces related to the technology of biomass gasification and its development include: • Functionality of technology Biomass gasification allowed for large scale and

advanced applications. Consequently for the first time, biomass was considered as a modern energy carrier. High efficiency contributed to IGCC’s promise in the 1990s while it is also an important argument in the preference for second generation over first generation biofuels (chapter 5).

• Flexibility as enabling technology The nature of biomass gasification as enabling technology in combination with the variety of selection environments showed to be important for the long term continuation of its development: the great variety of considered feedstock, technologies and end-use applications combined with the

167

great variety of selection environments over different countries and sectors secured continuous and long-term development.

• Maturity of technology and technological progress The maturity of the technology mainly influenced expectations in the late 1980s, greatly reinforced by the progress in biomass gasification and fluid bed combustion. The rising interest in biofuels around 2003 was not based on the maturity of gasification technology. Prior developments had focused on IGCC, but at that time the momentum was in a phase of disappointment. The field tried to adapt and refocus on biofuels, but this came with a reorientation of different technologies and a set-back in technological maturity, with much more focus on first generation biofuels and biochemical conversion routes.

• Spillover effects Occasionally the argument was used of potential spillover (e.g. from coal gasification or fixed-bed biomass gasification). The promise of spillover effects contributed to the IGCC promise, as several sources refer to the maturity of biomass combustion technology and the spillover potential (chapter 5). In my patent-citation network analysis (chapter 6), the actual spillover effects are explicit, illustrating that the entire field is strongly linked to various broader developments (fluid bed technology, combustion) and that the developments within biomass gasification are also linked. I show that these high levels of interrelatedness were not (only) due to the immaturity of the technology, but were characteristic of the field. Similar, more mature technologies (coal gasification, biomass combustion) also show high levels of interrelatedness. The patent-citation analysis shows that the development of advanced biomass gasifiers in the late 1970s (chapters 4 and 5) was based on a much broader development effort in fluid bed technology focusing on the conversion of coal (chapter 6). Developments during the 1990s focused on IGCC (chapter 4). Both Foster Wheeler and Ahlstrom strongly contributed to this development, building on their combustion technologies. Combustion technology proved to have a strong impact on these specific developments in biomass gasification as well as on the technological field, even though the patent set was defined around gasifier technology. The top path of the gasification-based patent set suggests that both co-evolved. Later developments in waste conversion utilized these gasification and combustion technologies (chapter 6). This thesis confirms the well-known fact that technologies benefit most from closely related technologies, those that show technological proximity. However, they come with a disadvantage that is less well articulated: these same technologies are likely to become their closest competitor. The development of biomass gasification and biomass combustion demonstrated this interaction. The phase of expectation dynamics showed to be relevant: during hype, the focus is on

168

the positive consequences of spillover effects from related technologies, whereas during disappointment, there is more focus on the competition from the same technologies.

Forces related to the social dimensions of developing biomass gasifiers include: • Alignment to broader technological visions Several visions were shared by broader

groups of actors, highlighting greater social momentum. The vision of methanol fuel was building on natural gas and coal gasification technologies, while the promise of biofuels was strongly based on first generation biofuels and progress in biochemical conversion routes (chapter 5). Developments in biomass gasification were able to hitch a ride, benefiting from the legitimacy and support created by these broader visions.

• Hype-disappointment mechanisms Explanatory factors include social mechanisms such as novelty preference, bandwagon effects, competitive threats of losing the emerging market or the risk of technological lock-out, and access to first-hand learning. Other factors are the interaction of different levels of expectations (micro, meso and macro) and mechanisms that show increasing returns to scale (learning curve effect). For learning curve effects it can take a long time before they have a significant effect, as they depend on the realization of multiple plants that can take several years. It is the expectation that this effect will become relevant that contributes to hype (chapter 5).

The application of biomass gasifiers frequently required adaptation of both the technology as well as the selection environment. With respect to technology this included a more fundamental study of phenomena; the testing of the configurational technology in demo plants; and the adaptation to other feedstock and applications. With respect to the selection environment it required adaptation of rules and regulations (e.g. permits, emission standards, etc.), perceptions, practices, and market structures (governments forcing power industries to look for alternatives; strict regulations on and support for second generation biofuels; creating biomass supply chains). This complicated the processes of adoption and diffusion. 7.2 Did this result in emerging trajectories? The second research question is: did the development result in emerging trajectories and what did a path creation and a path dependency approach contribute to their reconstruction? I already partially answered this question: overall no dominant design or clear trajectory emerged, but the development of advanced biomass gasifiers

169

showed emerging trajectories over different periods. I reconstructed these trajectories using two different approaches: the Geels and Raven (2006) approach that focusses on path creation as a result of shared expectations (chapter 4); and the Verspagen (2007) approach based on dominant paths in patent-citation networks (chapter 6). In addition, I detailed the dynamics in expectations (chapter 5) to better understand the structuring process and studied real time and interrelated developments in the patent set (chapter 6). Below I discuss the results of both the path creation and the path dependency approaches and how they help to trace the emerging trajectories. I do so by evaluating the relevant characteristics of technological trajectories: dominant design, incremental innovation, product sequences, shared expectations and/or routines, and self-reinforcing processes contributing to lock-in. In chapter 4 I applied the Geels and Raven approach to path creation for advanced biomass gasifiers. The large socio-economic and environmental problems of the time influenced expectations and provided momentum and direction to development. Especially governments are sensitive to this broader problem context and became highly involved. The engineering community translated issues into technological requirements. Consequently differing trajectories emerged: over the 1980s the focus was on methanol production by exploring oxygen-blown and indirectly-heated gasifiers; in the 1990s the focus was on IGCC for renewable power production by applying air-blown fluidized bed gasifiers. Developments after 2000 with the focus on biofuel production saw the exploration of multiple concepts. Early on in each period a wide variety of technologies was considered in the literature (variation). Over time the focus became narrower, when governments pre-selected technologies to pursue in their RD&D programs. I was able to identify preferred manufacturers and designs by the end of each period. However, over longer time periods this did not lead to dominance or a strong competitive advantage. Shifts in the socio-economic context, and thereby the expectations, not only resulted in a regular shakeout of less fit companies. Premium technologies and companies also struggled to survive. The Geels and Raven approach emphasizes shared expectations and focus, in line with the concept of dominant paradigms. There is less focus on product sequences. In the case of biomass gasification, the long development times, especially for demo plants, also contributed. There is limited focus on characterizing technological developments (incremental or radical innovation) and the cumulative nature of knowledge. Different parallel and subsequent developments were identified, although it is not always clear how these developments are related. Studying the dynamics of expectations (chapter 5) allowed me to detail and explain the self-reinforcing mechanisms that lead to

170

temporal structuring at an early stage of technological development, through expected progress or anticipated cost reductions. In chapter 6 I study patent-citation networks, following the Verspagen approach to reconstruct emerging trajectories from a path dependency perspective. My focus on patents for fluidized bed technologies is because these are strongly related to advanced gasifiers. The top path on fluid-bed gasifiers indicates that the development over time is best characterized as a patchwork of interrelated developments. I defined a broader dataset to take into account wider interrelated developments, including combustion and pyrolysis. The top path of this dataset shows that the Foster Wheeler and Ahlstrom technology is dominant. With a narrower focus on developments, I qualify this as a technological trajectory. This development draws on and co-evolves with combustion technology, influencing the subsequent development of waste incinerators and gasifiers. The full patent set reveals broader parallel developments and shifts in focus on technological developments. By drawing on patents, the Verspagen approach leans towards the engineering efforts, more in line with Dosi, and Nelson and Winter. The approach has a strong focus on product sequences and dominant paths. I added a real-time perspective that shows the shared focus at different times and a relational view to reveal how different developments are related. Expectations, stakeholder perspectives and broader contextual factors remain beyond the scope. The ‘path perspective’ shows some tension with the fluidity of technological developments that can be (loosely) related, as the structure of the patent-citation networks shows. An explicit check on incremental or radical innovation is lacking–this is left to the interpreting scholar to judge. To conclude, both approaches reveal emerging trajectories, but of a different nature. They vary not only due to different theoretical starting points (respectively the path creation and path dependency perspectives), but also because of the differences in methodologies and data sets. Both methods appear to be complementary. The Geels and Raven approach emphasizes path creation and the social dimension, while the Verspagen approach emphasizes path dependency and the engineering dimension. I show that combining both approaches can reveal a variety of developments such as early structuring leading to emerging trajectories, shifts in focus, and I trace how specific developments are related to prior and later technological developments. The case study also reveals this complementarity. The Verspagen approach shows that the 1990s’ IGCC developments were based on prior art of fluid bed combustion

171

already developed in the early 1980s. The Geels and Raven approach shows this was not the focal point in the 1980s, as the field of biomass gasification concentrated on exploring advanced concepts for producing methanol fuel. Finally, the broader examination of the patent set reveals that this was initially embedded in and based on a much broader effort to develop fluid bed technology, especially for the gasification and conversion of coal. 7.3 Discussion I close by discussing the methodological issues and some practical and policy implications of my research. To link the path creation and the path dependency perspectives, I made sure that both approaches reflected a real-time as well as a longitudinal perspective. Thus I expanded the analysis of top paths in patent-citation networks with analysis of real-time developments and their interrelatedness. This supports the Verspagen approach, as it shows how the top paths relate to or even draw on broader technological developments. However, I argue that this is equally important for the Geels and Raven approach: reconstructing real-time developments from patents can serve as cross-examination of trends, albeit limited to the development of the technology, while patent-citation analysis can clarify how different developments are related. For the Geels and Raven approach, I identified multiple technological developments and claims of (expected) spillover. However, these are often general and ill-substantiated claims, not mentioning the extent of the anticipated spillage, nor revealing the extent to which these expectations are shared by actors in the field. Patent-citation analysis enables a bottom-up reconstruction of spillover using a formalized approach. This specific example shows the trade-off between both approaches: either reconstruct the potential effects of spillover according to expressed claims and expectations, in line with a path creation perspective that might be significantly inaccurate; or reconstruct from a patent-citation network, which is more accurate, but is by definition in retrospect and therefore cannot strictly support a path creation perspective. Combining multiple approaches conflicts with the view of theoretical purists who argue that path creation and path dependency are different and strictly separate viewpoints on technological developments. For example, Garud, Kumaraswamy and Karnøe (2010) oppose the mixing of approaches, as it mixes ontologies. In their view on path creation, earlier technological developments are only relevant to the extent

172

that they influenced actors’ expectations at a certain time. They suggest “it would be productive to use real time notes made by the actors themselves” (p.770). Despite extensive research in biomass gasification, I have not found actors who made such notes extensively and continuously over the past 40 years. I imagine that other sources (internal company documents, policy documents) might add detail by strengthening the actor-specific and time-specific perspective, but I also see practical and methodological constraints on how to reconstruct from these the global and long-term field level developments. I did draw on the literature, science and technology indicators, and patent analyses in line with other scholars’ practices. To some extent this resulted in a mixing of ontologies. While the path creation perspective emphasizes how expectations or intentions guide developments that result in articles or patents, in practice the latter are used to reconstruct expectations and their dynamics in retrospect. Most articles and all patents are outputs of a development process and are biased towards ‘winning’ technologies. Alternative technologies that were considered but not pursued are not (as extensively) represented; nor are unsuccessful efforts that never reached the patenting stage. Thus, despite my reconstruction of ‘real-time developments’ in chapters 4, 5 and section 6.4, I can imagine that I have not met the strict standards that Garud et al. set for reconstructing a path creation perspective. I had to sacrifice some of this perspective to come up with a better reconstruction and understanding of field level developments. Trade-offs as discussed above are characteristic for a mixed-method approach as applied throughout my study. Quantitative indicators enabled me to trace developments long after the initial hype had died down and the focus shifted to other themes, while also indicating the momentum and amount of RD&D efforts. However, quantitative indicators suffered from several methodological restrictions, especially for tracing long term, global and interrelated developments. The qualitative approaches strongly support the trend analysis as they contain a lot of data on trends–albeit fragmented. In addition, the qualitative approaches allowed for a more contextualized reconstruction. Combining both in a mixed-method approach offers a reality check, on the one hand for quantitative research by showing how specific datasets and algorithms relate to practice, and on the other hand for qualitative research by correcting for subjectivity and cherry picking through literature selection or interpretation by the scholar. As methodologies and datasets do not always go well together, a reflexive approach is required. This way restrictions and incompatibilities can be revealed and potentially overcome.

173

Many of the recent contributions to Innovation Sciences seem to focus on serving national policy making. The inherent risk is that each barrier in an innovation process is reformulated as an opportunity, a problem to be overcome by improving policy making. I argue that a technology-focused study like this one provides a different perspective on lessons for steering and policy making: the focus is not only on what one can steer, but also on what one cannot steer. As such, this case brings a somewhat sobering perspective, as many of the forces are beyond the direct control of (national) policy making or limit the room to maneuver. While innovation scientists come up with detailed insights and recommendations, this thesis suggests that policy making is a rough and pragmatic process that typically:

• captures the window of opportunity resulting from large socio-economic and environmental challenges;

• adapts to the socio-politically relevant context of the time (e.g. the selective attention to debates on biofuels, ignoring well-known concerns) and strongly links energy-from-biomass with agricultural policy;

• matches policy with a promising technology–preferably one that will benefit the national economy and has at that time specific advantages over other technologies, for example the (expected) technological functions or maturity;

• aligns efforts with those of others, joins the bandwagon (e.g. the strong alignment between EU and US policies based on competitive threat mechanisms as well as broad and cumulative learning and learning curve effects);

• strives for and works towards market introduction and technology diffusion, but cannot count on them (despite many technologies and long and intensive RD&D efforts, biomass gasification is still in a niche development phase); should try to capture additional benefits from RD&D like community formation, learning, etc.

I highlight another policy dilemma prompted by the evolutionary nature of technological development: going with the momentum and aligning with dominant developments (selection, path creation, contribution to self-reinforcing developments) versus variation and non-conformity. My study case shows that momentum and alignment provided direction and scale, but variation and small scale developments proved important at times of discursive shifts. For instance Finland strongly contributed to the initial international interest in IGCC with its Jalo program and Choren started working on biofuels while the dominant focus was on IGCC, which proved to be important for later developments. Amid early exploration and niche development phases, both made a significant contribution to long term developments.

174

Three questions remain to be answered in this thesis: was the development of biomass gasification a failure; what are the prospects for the future; and what can I learn from this case with respect to innovation patterns? In section 7.1 I provide an overview of current applications. The technology did find some application, both in Western countries and in South-East Asia, but mainly with limited diffusion and for less advanced applications. This can be qualified as a limited success. The developments of high-end applications did not live up to the lofty expectations – as was shown in chapters 4-6. These applications are still in a phase of development and demonstration, depending on governmental support and hardly showing diffusion. However, in chapter 5 I showed that these high expectations are not necessarily accurate predictions of future developments, nor are they meant to be. Rather they are voiced in order to raise support for the development of a technology. As such the development of biomass gasification can be considered a success, as it continued to receive attention and support over the past 40 years. It resulted in exploration of a variety of technologies and application domains, and in continued learning and networking. All these are in support of the development and (future) application of the technology. Did the overall development provide value for money? In other words, did developments justify the RD&D budgets invested? That is a hard question to answer in general; and not a question that can be answered based on this research. Partially the answer will also be depending on the future success of the technology. This brings me to the second question: what are the prospects for the future? First a word of caution: this research did not focus on the prospects for the future; and the future of this technology is open and uncertain. But this thesis does offer a characterization of the technological development of biomass gasification on which I can draw. I conclude that a further niche development is likely for this technology in the near future: its developments remains strongly dependent on governmental support and shielding, while by its potential and flexibility it remains an attractive technology to pursue for specific applications. The big promises that are currently voiced will contribute to continued interest: the promise of bio-refineries; the promise of large-scale second generation biofuel production; and the promise of a bio-based economy. The final question is what I can learn from this case with respect to innovation patterns. The variety of possible applications of biomass gasifiers suggests that at each moment in time ‘matchmaking’ could have taken place, a process in which academics and manufacturers working on a specific technology connect to governments and selection environments that fit best their interest. However, in practice most development efforts were more localized, supported by national

175

policies or at best at European or North-American scale. I explain that by the immaturity of the technology that dependent on national innovation policies. This is not a universal pattern of development as the development of similar technologies show. For example, in the case of coal gasification the Dutch multinational Shell developed coal gasification by RD&D in amongst others Germany and USA, as these countries had the research facilities and support for clean coal technology (both countries have a strategic interest in coal, while Shell has strong finances and knowledge a.o. by the FT synthesis in South Africa). This seems to suggest that in the case of biomass gasification there was not one manufacturer or research group that stood out with respect to technology or competences over national product champions. Another example is the case of biomass combustion. Leading combustion technologies were developed outside large governmental programs (Ahsltrom/FW and Lurgi). Over the 1980s these European technologies have been introduced first in the US and captured the US market, which was specifically advantageous due to PURPA regulations for independent power producer. Later these technologies found a market in Scandinavia and Asia as well. This pattern can likely be explained by the maturity and competitiveness of the combustion technology in comparison to the biomass gasification technology. This thesis shows that for biomass gasification patterns of development are complex: the technology developed over multiple related niches; over multiple selection environments; and in addition these also changed over time. This complexity might be an interesting topic for future research on understanding dynamics in socio-technical developments and transitions.

176

References Albert, M., Avery, D., Narin, F., & McAllister, P. (1991). Direct validation of citation counts as indicator of

industrially important patents. Research Policy, 251-259. Alcácer, J., Gittelman, M., & Sampat, B. (2009). Applicant and examiner citations in U.S. patents: an overview

and analysis. Research Policy, 415-427. Alkemade, F., & Suurs, R. (2012). Patterns of expectations for emerging sustainable technologies. Technological

Forecasting & Social Change, 448-456. Allgeier, H., Caratti, G., & Sandberg, O. (1995). Towards a European BIO-ENERGY strategy. 8th European

biomass conference, (pp. 11-19). Armistead, T., Schwieger, R., & Reina, P. (2002). Boiler technology pursues goals of efficiency and lower

emissions-circulating-fluidized-bed equipment pushes size limits. Power & industial power plants. Asplund, D., Sahrman, K., & Solantausta, Y. (1980). The energy from wood and research on wood energy in

Finland. In B.-e. Council (Ed.), Bioenergy 80: world congress and exposition (pp. 311-314). Washington: Bio-energy Council.

Asveld, L., Est, R. v., & Stemerding, D. (. (2011). Getting to the core of the bio-economy: A perspective on the sustainable promise of biomass. The Hague: Rathenau Instituut.

Auken. (1996). Opening remarks at the 9th European bioenergy conference. 9th European biomass conference, (pp. 5-8).

Babu, S. (1999). Thermal gasification of biomass. IEA Bioenergy task 20. Babu, S. (2005, March). Observations of the current status of biomass gasification. Retrieved April 2013, from

IEA Bioenergy: http://www.ieabioenergy.com/LibItem.aspx?id=184 Babu, S. P. (1995). Thermal gasification of biomass technology developments: end of task report for 1992-1994.

Biomass and bioenergy, 271-285. Babu, S. P., & Whaley, T. P. (1992). IEA Biomass thermal gasification project. Biomass and bioenergy, 299-306. Bacovsky, D., Dallos, M., & Wörgetter, M. (2010). Status of 2nd generation biofuels demonstration facilities in

June 2010. IEA Bioenergy Task 39. Bain, R. (2011). United States country report, IEA Bioenergy, task 33. Golden, Colorado: NREL. Bain, R. L. (1993). Electricity from biomass in the United States: status and future directions. Bioresource

Technology, 86-93. Bain, R., Amos, W., Downing, M., & Perlack. (2003). Biopower technical assessment: state of the industry and

technology. Golden, Colorado: NREL. Baker, E., Brown, M., Moore, R., Mudge, L., & Elliott, D. (1986). Engineering analysis of biomass gasfier product

gas cleaning technology. Richland: Pacific Northwest Laboratory. Bakker, S., & Budde, B. (2012). Technological hype and disappointment: lessons from the hydrogen and fuel cell

case. Technology Analysis & Strategic Management, 549-563. Bakker, S., Van Lente, H., & Meeus, M. (2011). Arenas of expectations for hydrogen technologies. Technological

Forecasting & Social Change, 152-162. Bakker, S., van Lente, H., & Meeus, M. (2012). Dominance in the prototyping phase - The case of hydrogen

passenger cars. Research Policy, 871-883. Banerjee, R. (2006). Comparison of options for distributed generation in India. Energy Policy, 101-111. Basberg, B. (1987). Patents and the measurement of technological change: a survey of the literature. Research

Policy, 131-141. Basu, P. (2006). Combustion and gasification in fluidized beds. Boca Raton: Taylor & Francis Group. Beenackers, A. (2001). The main conclusions from the first world conference and exhibition on biomass from

energy and industry, Sevilla (2000). 1st World conference on biomass for energy and industry, (pp. lxi-lxv). Beenackers, A., & Bridgwater, A. (1989). Gasification and pyrolysis of biomass in Europe. Pyrolysis and

gasification (pp. 129-157). Elsevier Science Publishers.

177

Beenackers, A., & Maniatis, K. (1984). Conclusions of the workshop on evaluation of themochemical gasification reactors for biomass. In A. Bridgwater (Ed.), Thermochemical processing of biomass (pp. 319-325). London: Buttersworths.

Beenackers, A., & Maniatis, K. (1993/1994). Gas cleaning in electricity production via gasification of biomass: conclusions of the workshop. In A. Bridgwater (Ed.), Advances in thermochemical biomass conversion 1992 (pp. 540-546). Glasgow: Blackie Academic & Professional.

Beenackers, A., & Maniatis, K. (Eds.). (1998). Biomass & Bioenergy, 15(3), 193-273. Beenackers, A., & van Swaaij, W. (1983). Methanol from wood, a state of the art review. Energy from biomass,

2nd EC conference 1982 (pp. 782-798). Brussels and Luxembourg: ECSC. EEC, EA EC / Applied Science Publishers.

Beenackers, A., & van Swaaij, W. (1984). Gasification of biomass, a state of the art review. In A. Bridgwater (Ed.), Thermochemical processing of biomass (pp. 91-136). London: Butterworths.

Beenackers, A., & van Swaaij, W. (1985). The biomass to synthesis gas pilot plant programme of the C.E.C.; a first evalution of its results. Energy from biomass, 3rd EC conference – Venice (Italy), 1985, (pp. 120-145).

Beenackers, A., & van Swaaij, W. (1986). Introduction to the biomass to synthesis gas pilot plant programme of the C.E.C. and a first evaluation of results. In A. Beenackeers, & W. van Swaaij (Eds.), Advanced gasification. Solar energy R&D in the European Community, energy from biomass. (Vols. Series E, volume 8, pp. 2-27). Brussels and Luxembourg: Commission of the European Communities.

Beenackers, A., & van Swaaij, W. (1990). Gasification of biomass. Need for further R&D. Biomass for energy and industry, 5th EC conference 1989 (pp. 2.524-2.533). Brussels and Luxembourg: Commission of the European Communities / Elsevier Applied Science.

Bekkers, R., & Martinelli, A. (2012). Knowledge positions in high-tech markets: trajectories, standards, strategies and true innovators. Technological Forecasting and Social Change, 1192-1216.

Bettini, V. (1994). Keynote Address. 7th European biomass conference, (pp. 9-12). Beurskens, L., & Hekkenberg, M. (2011). Renewable energy projections as published in the national renewable

energy action plans of the European member states. Petten: ECN. Bhattacharya, S. (2005). Biomass gasification in Asia. In H. Knoef, Handbook Biomass Gasification (pp. 162-180).

Enschede: BTG. Bhupatiraju, S., Nomaler, Ö., Triulzi, G., & Verspagen, B. (2012). Knowledge flows - Analyzing the core literature

of innovation, entrepreneurship and science and technology studies. Research Policy, 1205-1218. Bio-energy Council. (1980). Bioenergy 80: world congress and exposition. Washington: Bio-energy Council. Biomass energy foundation. (2008, February). Woodgas database producers. USA. Biondi, L., & Galli, R. (1992). Technological trajectories. Futures, 580-592. Blomkvist, G., Liinanki, L., Lindman, N., & Sjöberg, S.-O. (1983). Swedish gasification research and process

development at RIT. Energy from biomass, 2nd EC Conference 1982 (pp. 896-900). Brussels and Luxembourg: Commission of the European Communities / Applied Science Publishers.

Boerrigter, H., & Rauch, R. (2005). Syngas production and utilisation. In H. Knoef, Handbook biomass gasification (pp. 211-230). Enschede: BTG.

Boissonnet, G., Boudet, N., Seiler, J.-M., & Duplan, J.-L. (2005). Process simulation and comparison of several biomass gasification routes for the production of Fischer-Tropsch liquids. Biomass for energy, industry and climate protection, 14th EC conference, (pp. 588-591).

Bole, T., & Londo, M. (2009). Improving the investment climate of second generation biofuels. 17th European Biomass Conference and Exhibition, Hamburg (pp. 54-61). ETA-Renewable Energies.

Borup, M., Brown, N., Konrad, K., & van Lente, H. (2006). The sociology of expectations in science and technology. Technology Analysis & Strategic Management, 285-298.

BP. (2011). BP Statistical Review of World Energy. Brandon, O., & Kinsey, D. (1984). Implementation of technologies for the thermochemical processing of

biomass. In A. Bridgwater (Ed.), Thermochemical processing of biomass, (pp. 307-311). London: Buttersworth.

Brandon, O., King, G., & Kinsey, D. (1984). The role of thermochemical processing in biomass exploitation. In A. Bridgwater (Ed.), Thermochemical processing of biomass (pp. 11-34). London: Buttersworths.

178

Breault, R. W. (2008, april 24). DOE´s Gasification Program Overview. Retrieved October 2009, from Clean Air Council: http://www.cleanair.org/coalconversation/pres/Breault.pdf

Bridgwater, A. (1984). The thermochemical processing system. In A. Bridgwater (Ed.), Thermochemical processing of biomass (pp. 35-52). Birmingham / London: University of Aston / Butterworths.

Bridgwater, A. (1990a). A survey of thermochemical biomass processing activities. Biomass, 22, 279-292. Bridgwater, A. (1990b). Biomass pyrolysis technologies. 5th European biomass conference, (pp. 2.489-2.496). Bridgwater, A. (1993). A preliminary study on the technical and economic feasibility of wood gasfication for

power generation. Knowle: Consultants on process engineering Ltd. Bridgwater, A. (1995). The technical and economic feasibility of biomass gasification for power generation. Fuel,

631-653. Bridgwater, A. (1997). Fast pyrolysis of biomass in Europe. Biomass gasification and pyrolysis. State of the art

and future prospects (pp. 53-67). Newbury: CPL press. Bridgwater, A., & Beenackers, A. (1985). Research priorities in thermal conversion technology. 3rd European

biomass conference, (pp. 247-262). Bridgwater, A., & Beenackers, A. (1990). Gasification and pyrolysis of biomass in Europe. 5th European biomass

conference, (pp. 2.827-2.830). Bridgwater, A., & Bolhàr-Nordenkampf, M. (2005). Economics of biomass gasification. In H. Knoef, Handbook

biomass gasification (pp. 321-343). Enschede: Biomass Technology Group. Bridgwater, A., & Evans, G. (1993). An assessment of thermochemical conversion systems for processing

biomass and refuse. Aston University / DK Teknik. Bridgwater, A., Beenackers, A., Sipila, K., Zhenghong, Y., Chuangzhi, W., & Li, S. (1999). An Assessment of the

Possibilities for Transfer of European Biomass Gasification Technology to China. EC THERMIE Publication No. 13563. Retrieved 2009, from http://ec.europa.eu/energy/renewables/studies/doc/bioenergy/final_report_for_publication.pdf

Bridgwater, T. (1992). Thermochemical and biochemical biomass conversion activities. Biomass and Bioenergy, 2(1-6), 307-318.

Brown, N. (2003). Hope against hype - accountability in biopasts, presents and futures. Science Studies, 3-21. Budde, B. (2015). Hopes, hypes and disappointment: on the role of expectations for sustainability transitions. Budde, B., Alkemade, F., & Matthias Weber, K. (2012). Expectations as a key to understanding actor strategies

in the field of fuel cell and hydrogen vehicles. Technological Forecasting & Social Change, 1072-1083. Burdon, I. (2003). The future for advanced thermal conversion technologies - the commercial perspective. In A.

Bridgwater (Ed.), Pyrolysis and gasification of biomass and waste, proceedings of an expert meeting 2002 (pp. 19-21). Berk: CPL Press.

Carrasco, J. (2005). Summary of the conference results. 14th European biomass conference, (pp. liii-lvi). Chapman, D. (2012, September 2). Warnings ignored in Range Fuels debacle. Retrieved August 2014, from The

Atlanta Journal Constitution: http://www.ajc.com/news/business/warnings-ignored-in-range-fuels-debacle/nR2H8/

Chartier, P. (1981). Prospects for energy from biomass in the European Community. 1st European biomass conference, (pp. 22-33).

Chartier, P. (1983). Summing up and evaluation of biomass prospects. 2nd European biomass conference, (pp. vii-ix).

Chartier, P. (1990). Biomass fore energy and industries: fifth European conference (Lisbon) executive summary. 5th European biomass conference, (pp. 2.1169-2.1175).

Chartier, P. (1992). General synthesis of the conference. 6th European biomass conference, (pp. 1406-1414). Chartier, P. (1994). Synopsis of sessions and prospects for biomass. 7th European biomass conference, (pp. 363-

367). Chartier, P., & Hall, D. (1981). Summary report of the co-chairmen of the conference. 1st European biomass

conference, (pp. 2-6). Chartier, P., & Palz, W. (1980). The second European Community programme on energy from biomass. Bio-

energy '80: world congress and exposition (pp. 321-325). Washington: Bio-Energy Council.

179

Chartier, P., Beenackers, A., & Grassi, G. (Eds.). (1995). Biomass for energy, environment, agriculture and industry. Proceedings of the 8th European biomass conference 1994. Oxford: Pergamon / Elsevier Science Ltd.

Chartier, P., Ferrero, G., Henius, U., Hultberg, S., Sachau, J., & Wiinblad, M. (Eds.). (1996a). Biomass for energy and the environment. Proceedings of the 9th European biomass conference 1996. Oxford: Pergamon / Elsevier Science Ltd.

Chartier, P., Ferrero, G., Henius, U., Hultberg, S., Sachau, J., & Wiinblad, M. (1996b). Preface. 9th European biomass conference, (pp. v-vi).

CHESF; CIENTEC, CVRD, ELECTROBAS, Shell Brasil, Shell International Petroleum Company. (1995). Brazilian BIG-GT demonstration project (WBP). Technology selection - final report. .

China Patent Information Center. (2009). China patent database. Retrieved May 2009, from http://search.cnpat.com.cn/Search/EN/

Ciferno, J. P., & Marano, J. J. (2002). Benchmarking Biomass Gasification Technologies for Fuels, Chemicals and Hydrogen Production. US Department of Energy, NETL.

Clarke, F., Strub, A., Ghose, T., & Berger, B. (1981). Foreword. 1st European biomass conference, (pp. v-vi). Cobb, J. T. (2007, November). Survey of Commercial Biomass Gasifiers. 2007 AIChE annual meeting. Retrieved

from University of Pittsburgh, Department of Chemical and Petroleum Engineering. Colitti, M., & Baronti, P. (1981). Energy policies of industrialized countries. Energy, 233-262. Colombo, U. (1994). Keynote Address. 1st European biomass conference, (pp. 3-8). Connor, M., Diebold, J., & Sjöström, K. (1997). Workshop on the fundamentals of thermochemical biomass

conversion. In A. Bridgwater, & D. Boocock (Ed.), Developments in thermochemical biomass conversion (pp. 1617-1620). London: Blackie Academic & Professional / Chapman & Hall.

Costello, R., & Finnell, J. (2002). Biomass in the U.S. Federal Government: activities and strategies. Twelfth European biomass conference. Biomass for energy, industry and climate protection. (pp. 10-16). Florence: ETA Florence and WIP Munich.

Dallas, I., McLellan, R., Scanlon, K., & Smith, D. (1986). The gasification of wood using the oxygen donor process. In B. A.A.C.M., & W. van Swaaij (Ed.), Advanced gasification. Methanol Production from Wood - Results of the EEC Pilot Programme (pp. 115-172). Dordrecht: D. Reidel Publishing Company.

Dallemand, J. (2009). 17th biomass conference technical report. 17th European biomass conference. De Lange, H., & Barbucci, P. (1998). The Thermie Energy Farm project. Biomass and Bioenergy, 219-224. de Sampaio Nunes, P. (1995). The role of biomass in the European energy policy - Keynote address. 8th

European biomass conference, (pp. 20-29). De Santi, G., Dallemand, J., Ossenbrink, H., Grassi, A., & Helm, P. (Eds.). (2009). 17th European biomass

conference. From research to industry and markets. Florence: ETA Florence Renewable Energies. Diebold, J., & Stevens, D. (1989). Progress in pyrolysis and gasification of biomass: an overview of research in

the United States. In G. Ferrero, K. Maniatis, A. Buekens, & A. Brigwater (Ed.), Pyrolysis and Gasification, Luxembourg 1989 (pp. 14-27). London and New York / Brussels and Luxembourg: Elsevier Applied Science / EEC.

Dignum, M. (2013). The power of large technological visions. The promise of hydrogen energy (1970-2010). 's-Hertogenbosch: BOXPress.

Dorca Duch, A., & Huertas Bermejo, J. (2008). Biomass gasification. The characteristics of technology development and the rate of learning. Chalmers University of Technology. Göteborg, Sweden: Chalmers University of Technology. Retrieved June 9, 2009, from http://www.chalmers.se/ee/SV/forskning/forskargrupper/miljosystemanalys/publikationer/ee/SV/forskning/forskargrupper/miljosystemanalys/publikationer/pdf-filer/2008/downloadFile/attachedFile_13_f0/2008-16.pdf?nocache=1222948096.81

Dosi, G. (1982). Technological paradigms and technological trajectories: a suggested interpretation of the determinants and directions of technical change. Research Policy, 147-162.

Dosi, G., Orsenigo, L., & Sylos-Labini, M. (2005). Technology and the economy. In N. J. Smelser, & R. Swedberg, Handbook of Economic Sociology (pp. 678–702). Princeton, New Jersey & New York: Princeton University Press & Russell Sage.

180

Downs, A. (1972). Ups and downs with ecology - the 'issue-attention cycle'. The Public Interest, 28, 38-50. Drift, A. v. (2001). CFB-vergassers voor biomassa, een overzicht van bestaande installaties. Petten: ECN. Drift, A. van der; Boerrigter, H.; Energie Onderzoekcentrum Nederland (ECN). (2006). Synthesis gas from

biomass for fuels and chemicals. Petten: ECN. Durand, H. (1981). The French bioenergy programme. Energy from biomass, 1st EC Conference 1980, (pp. 826-

837). E4tech. (2009). Review of technologies for gasification of biomass and wastes. Retrieved from

http://www.e4tech.com/reports/review-of-technologies-for-gasification-of-biomass-and-wastes/ EBC European Biomass Conference. (1980-2010). EBTP European Biofuels Technology Platform. (2013). Biomass to Liquids (BtL). Retrieved July 2014, from

http://www.biofuelstp.eu/btl.html Edouard, L. (1986). The relationship between coal and nuclear power for electricity production: a strategy for

electricity production in France. Energy, 1293-1297. Elliot, P. (1993). Biomass-energy overview in the context of Brazilian biomass-power demonstration.

Bioresource Technology, 13-22. Elliott, D., & Maggi, R. (1997). Workshop report: catalytic processes. In A. Bridgwater, & D. Boocock (Ed.),

Developments in thermochemial biomass conversion (pp. 1626-1630). London: Blackie Academica & Professional / Chapman & Hall.

ETA / WIP. (2007). The policy bioenergy debate. Can bioenergy deliver the required market penetration? 15th European biomass conference, (pp. lxxiii-lxxxix).

EurObserv'ER. (2008, december). Solid biomass barometer. Primary energy production of solid biomass in the European Union in 2006. Systèmse Solaires, le journal des énergies renouvelables(188), pp. 69-84.

EuropaBio. (2008-2011). Bio-based economy. Retrieved 2011, from http://www.bio-economy.net European Commission. (1997). Energy for the future, renewable sources of energy. White paper for a

Community strategy and action plan. Luxembourg: European Commission. European Patent Office. (2009). Advanced search. Retrieved 2009, from European Patent Office - Espacenet

1998-2008: http://ep.espacenet.com/advancedSearch European Patent Office. (2015). EPO Worldwide Patent Statistical Database (PatStat). Retrieved from

http://www.epo.org/searching/subscription/raw/product-14-24.html Eurostat. (2009). Eurostat - agricultural products. In: Agricultural statistics, main results 2007-2008.

Luxembourg: Office for Official Publications of the European Communities. Eurostat European Commission. (2011). Energy. Retrieved 2011, from

http://epp.eurostat.ec.europa.eu/portal/page/portal/energy/introduction Eurostat, Roubanis, N., Dahlström, C., & Noizette, P. (2010). Renewable energy statistics. Eurostat. European

Union. EWAB Programme Novem. (2001). Gasifier Inventory: inventory of biomass gasifier manufacturers and

installations. EWAB/Novem. Exxon Research and Engineering Company. (1978). Exxon Catalytic Coal Gasification Process Predevelopment

Program. Final Project Report. U.S. Department of Commerce. Faaij, A. (2006a). Modern biomass conversion technologies. Mitigation and Adaptation Strategies for Global

Change, 343-375. Faaij, A. (2006b). Bio-energy in Europe: changing technology choices. Energy Policy, 322-342. Faaij, A. (2007). Potential Contribution of Bioenergy to the World's Future Energy Demand. IEA Bioenergy. Faaij, A., Meuleman, B., & Ree, R. v. (1998). Long term perspectives of biomass integrated gasification with

combined cycles. Costs and efficiency and a comparison with combustion. Utrecht: EWAB / NOVEM. Faber, A., & Frenken, K. (2009). Models in evolutionary economics and environmental policy: toward an

evolutionary environmental economics. Technological forecasting & social change, 462-470. Faber, A., van den Bergh, J., Idenburg, A., & Oosterhuis, F. (2005). Survival of the Greenest. Evolutionary

economics and policies for energy innovation. 6th Conference of the European Society for Ecological Economics. Lissabon.

181

Fabry, R., & Ferrero, G. (1989). Full-scale demonstration projects of the European Community in the field of pyrolysis, gasification and carbonisation of biomass and waste. In G. Ferrero, K. Maniatis, A. Buekens, & A. Bridgwater (Ed.), Pyrolysis and gasification (pp. 405-410). London and New York / Brusssels and Luxembourg: Elsevier Applied Science / EEC.

Fabry, R., & Ferrero, G. (1990). Implementation and results of the commission's energy demonstration programme. 5th European biomass conference, (pp. 2.1158-2.1164).

Fabry, R., & Goudeau, J.-C. (1987). Thermal conversion of biomass and waste by combustion and pyrolysis/gasification, CEC directorate-general for energy demonstration programme. 4th European biomass conference, (pp. 274-282).

Fairclough, N. (2003). Analysing discourse, textual analysis for social research. Oxon: Routledge. FAO. (2006). Global Forest Resources Assessment 2005. Rome: FAO. FAO, Food and Agriculture Organization of the United Nations. (1986). Wood gas as engine fuel. FAO. Fenn, J., & Raskino, M. (2008). Mastering the hype cycle. How to choose the right innovation at the right time.

Boston, Massachusetts: Harvard Business Press. Fjällström, T. (2004). Highlights on the technical programme. 2nd World Conference on biomass for energy,

industry and climate protection, (p. xxviii). Fleck, J. (1994). Learning by trying: the implementation of configurational technology. Research Policy, 637-652. Foley, G., & Barnard, G. (1983). Biomass gasification in developing countries. London: International Institute for

Environment and Development IIED. Fontana, R., Nuvolari, A., & Verspagen, B. (2009). Mapping technological trajectories as patent citation

networks. An application to data communication standards. Economics of Innovation and New Technology, 311-336.

Foster Wheeler. (2009). Pioneering CFB technology. Retrieved from Foster Wheeler website. Foster Wheeler Energia Oy. (n.d.). Leading the world in CFB. Frishammer, J., Söderholm, P., Bäckström, K., Hellsmark, H., & Ylinenpää, H. (2015). The role of pilot and

demonstration plants in technological development: synthesis and directions for future research. Technology Analysis & Strategic Management, 1-18.

Garud, R., & Karnøe, P. (2001). Path creation as a process of mindful deviation. In R. Garud, & P. (. Karnøe, Path dependence and creation (pp. 1-40). Mahwah: Lawrence Erlbaum Associates.

Garud, R., Kumaraswamy, A., & Karnøe, P. (2010). Path dependence or path creation? Journal of Management Studies, 760-774.

Geels, F. (2002). Technological transitions as evolutionary reconfiguration processes: a multi-level perspective and a case study. Research Policy, 1257-1274.

Geels, F. W., & Schot, J. (2010). The dynamics of transitions: a socio-technical perspective. In J. Grin, J. Rotmans, & J. Schot, Transitions to sustainable development (pp. 9-101). New York / Oxon: Routledge.

Geels, F., & Raven, R. (2006). Non-linearity and expectations in niche-development trajectories: ups and downs in Dutch biogas development (1973-2003). Technology Analysis & Strategic Management, 375-392.

Geels, F., Pieters, T., & Snelders, S. (2007). Cultural enthusiasm, resistance and the societal embedding of new technologies: psychotropic drugs in the 20th century. Technology Analysis & Strategic Management, 145-165.

Geisler, E. (1994). Key output indicators in performance evaluation of research and development organizations. Technological forecasting and social change, 189-203.

Ghosh, D., Sagar, A., & Kishore, V. (2004). Scaling up biomass gasifier use: applications, barriers and interventions. Washington D.C.: The World Bank.

Goidich, S. J., Hyppanen, T., & Kauppinen, K. (1999, August). CFB boiler design and operation using the INTREX heat exchanger. 6th International Conference on Circulating Fluidized Beds. Würzburg. Retrieved from Website of Foster Wheeler.

Grace, J. R., & Bi, H. (1997). Introduction to circulating fluidized beds. In J. Grace, A. Avidan, & T. Knowlton (Eds.), Circulating Fluidized Beds (pp. 1-20). London: Chapman & Hall.

Grassi, B., & Utria, B. (2004). Overview of world wide activities in international co-operation (round table). 2nd World Conference on biomass for energy, industry and climate protection, (pp. 71-74).

182

Grassi, G. (1988). The European R&D programme, strategy for the future. In A. Bridgwater, & J. Kuester (Ed.), Research in thermochemical biomass conversion (pp. 16-30). Essex: Elsevier Science Publishers.

Grassi, G. (1990). Definition dún programme biomasse. 5th European biomass conference, (pp. 1.39-1.48). Grassi, G. (1992). 6th European conference on biomass. 6th European biomass conference, (pp. 27-29). Grassi, G. (1993). Overview of thermochemical conversion R&D activities of the CEC. In A. Bridgwater (Ed.),

Advances in thermochemical biomass conversion, Interlaken 1992 (pp. 52-62). Glasgow / London: Blakie Academic & Professional / Chapman & Hall.

Grassi, G. (1994a). Introductory remarks. 7th European biomass conference, (p. 2). Grassi, G. (1994b). Strategy for biomass implementation. 6th European biomass conference, (pp. 263-271). Grassi, G., & Bridgwater, A. (1990). Biomass for energy, industry and the environment. A strategy for the future.

Brussels / Luxembourg: Commission of the European Communities / Edizioni Esagono - Sesto San Giovanni. doi:EUR 12897 EN

Grassi, G., & Bridgwater, A. (1993). The opportunities for electricity production from biomass by advanced thermal conversion technologies. Biomass and Bioenergy, 339-345.

Grassi, G., & Pirrwitz, D. (1983). The European Community's research and development programme for energy from biomass. 2nd European biomass conference, (pp. 659-663).

Grassi, G., Collina, A., & Zibetta, H. (Eds.). (1992a). Biomass for energy, industry and environment, 6th E.C. conference 1991. Brussels and Luxembourg: Commission of the European Communities / Elsevier Applied Science.

Grassi, G., Collina, A., & Zibetta, H. (1992b). Preface. 6th European biomass conference, (p. v). Grassi, G., Delmon, B., Molle, J.-F., & Zibetta, H. (Eds.). (1987a). Biomass for energy and industry, 4th E.C.

conference 1987. Brussels and Luxembourg: Commission of the European Communities / Elsevier Applied Science.

Grassi, G., Delmon, B., Molle, J.-F., & Zibetta, H. (1987b). Preface. 4th European biomass conference, (p. v). Grassi, G., Gosse, G., & dos Santos, G. (Eds.). (1990a). Biomass for energy and industry, 5th E.C. conference

1989. Brussels and Luxembourg: Commission of the European Communities / Elsevier Applied Science. Grassi, G., Gosse, G., & dos Santos, G. (1990b). Preface. 5th European biomass conference, (pp. v-vii). Griliches, Z. (1990). Patent statistics as economic indicators: a survey. Journal of Economic Literature, 1661-

1707. Gunawardhana, T. (1983). Producer gas 1982. A collection of papers on producer gas with emphasis on

applications in developing countries. Stockholm: Beijer Institute of ecological economics. Hajer, M., & Versteeg, W. (2005). A decade of discourse analysis of environmental politics: achievements,

challenges, perspectives. Journal of Environmental Policy & Planning, 175-184. Hall, B., Jaffe, A., & Trajtenberg, M. (2001). The NBER patent citations data file: lessons, insights and

methodological tools. Cambridge: NBER. Hall, D. (1982). Editorial, biomass for energy in Europe. Biomass, 239-244. Hall, D. (1983). Food versus fuel, a world problem? 2nd European biomass conference, (pp. 43-62). Hall, D. (1997). Biomass energy in industrialised countries - a view of the future. Forest ecology and

management, 17-45. Hall, D., & Coombs, J. (1985). Round tables. 3rd European biomass conference, (pp. 236-239). Hall, D., & House, J. (1995). Biomass energy in Western Europe in 2050. Land use policy, 37-48. Hall, D., & Palz, W. (1985). Preface. 3rd European biomass conference, (p. v). Hall, D., Grassi, G., & Scheer, H. (Eds.). (1994a). Biomass for energy and industry, 7th E.C. conference 1992.

Brussels and Luxembourg: Commission of the European Communities / Ponte Press. Hall, D., Grassi, G., & Scheer, H. (1994b). Preface. 7th European biomass conference, (p. v). Hall, S. (2001). Foucault: Power, Knowledge and Discourse. In M. Wetherell, S. Taylor, & S. Yates, Discourse

theory and practise (pp. 72-81). London: Sage Publications. Hamelinck, C., & Bain, R. (2003). Syngas for synfuels - workshop report. In A. Bridgwater (Ed.), Pyrolysis and

gasification of biomass and waste, proceedings of an expert meeting 2002 (pp. 693-696). Berks: CPL Press. Hardy, C., Harley, B., & Philips, N. (2004, spring). Discourse analysis and content analysis: two solitudes?

Qualitative Methods Newsletter, pp. 19-22.

183

Harhoff, D., Scherer, F., & Vopel, K. (2003). Citations, family size, opposition and the value of patent rights. Research Policy, 1343-1363.

Harmsen, R. (2000). Forces in the development of coal gasification (PhD thesis). Utrecht: Universiteit Utrecht. Hayes, R. (1989). Overview of thermochemical conversion of biomass in Canada. In G. Ferrero, K. Maniatis, A.

Buekens, & A. Bridgwater (Ed.), Pyrolysis and gasification (Luxembourg, 1989) (pp. 28-39). Brussels and Luxembourg: Commission of the European Communities / Elsevier Applied Science.

Heermann, C., Schwager, F., & Whiting, K. (2001, September). Pyrolysis & Gasification of Waste: A Worldwide Technology & Businesss Review. Juniper Consultancy Services Ltd.

Hellsmark, H. (2010). Unfolding the formative phase of gasified biomass in the European Union. Göteborg: Chalmers University of Technology.

Hellsmark, H., & Jacobsson, S. (2012). Realising the potential of gasified biomass in the European Union - policy challenges in moving from demonstration plants to a larger scale diffusion. Energy Policy, 507-518.

Hellsmark, H.; Chalmers University of Technology. (2007). The formative phase of biomass gasification. Retrieved June 2008, from http://web.fu-berlin.de/ffu/veranstaltungen/salzburg2007/Hans%20Hellsmark.pdf

Henley, J. P. (2007, march 23). Overview of China's lead in coal gasification implementation. Retrieved sept 08, 2009, from Website Gasification Technologies Council: http://www.gasification.org/Docs/Legis/2007%20China%20Roundtable/Senate%20RT%20Review%20DOW%20JPH%202007%2003%2023.pdf

Hervouet, V. (2009). The European Biofuels Technology Platform Highlights. 17th European Biomass Conference and Exhibition, Hamburg, Germany. ETA-Renewable Energies.

Hirst, N. (2007). Address from the International Energy Agency - IEA. 15th European biomass conference, (pp. lxii- lxiii).

Hodam, R., Williams, R., & Lesser, M. (1982). Engineering and economic characteristics of commercial wood gasifiers in North America. Hodam & Associates, Inc. Golden, Colorado: SERI, Solar Energy Research Institute.

Hogan, E. (1992). Thermochemical conversion R&D activities in Canada. Biomass for energy, industry and environment; 6th EC conference 1991 (pp. 624-628). Brussels and Luxembourg: ECSC, EEC, EAEC / Elsevier Science Publishers.

Hogan, E. (1993). Overview of Canadian thermochemical biomass conversion activities. Advances in thermochemical biomass conversion, 1992 Interlaken Switzerland (pp. 15-25). Glasgow: Blackie Academic or Professional.

Hoogwijk, M. (2004). On the global and regional potential of renewable energy sources. Utrecht: Universiteit Utrecht.

Hughes, E. (1998). Advanced biomass: technology characteristics, status and lessons learned. Palo Alto, California: EPRI.

Hummon, N., & Doreain, P. (1989). Connectivity in a Citation Network: the Development of DNA Theory. Social Networks, 39-63.

Hung, S.-W., & Wang, A.-P. (2010). Examining the small world phenomenon in the patent citation network: a case study of the radio frequency identification (RFID) network . Scientiometrics, 121-134.

Huttunen, S. (2009). Ecological modernisation and discourses on rural non-wood bioenergy production in Finland from 1980 to 2005. Journal of Rural Studies, 239-247.

IEA. (2010). Key world energy statistics. IEA. IEA. (2011). Technology roadmap biofuels for transport. IEA. IEA International Energy Agency. (2015). World Energy Outlook 2015. Paris: OECD/IEA. Intellectual Property India. (2008). Public search for patents. Retrieved 2009, from Controller General of Patents

Designs and Trademarks: http://ipindia.nic.in/patsea.htm Jäger-Waldau, A., & Ossenbrink, H. (2004). Progress of electricity from biomass, wind and photovoltaics in the

European Union. Renewable and sustainable energy reviews, 157-182. Järvenpää, H., & Mäkinen, S. (2008). An empirical study of the existence of the hype cycle: a case of DVD

technology. Engineering Management Conference. IEEE.

184

Johansson, T. (1993). Biomass-based power generation in Europe. On the technical and economical feasibility of wood production, and wood gasification for gas turbine power generation. Lund, Sweden: University of Lund.

Johnson, R. B., Onwuegbuzie, A., & Turner, L. (2007). Toward a definition of mixed methods research. Journal of Mixed Methods Research, 112-133.

Johnson, R., & Onwuegbuzie, A. (2004). Mixed methods research: a research paradigm whose time has come. Educational Researcher, 14-26.

Jun, S.-P. (2012). A comparative study of hype cycles among actors within the socio-technical system: with a focus on the case study of hybrid cars. Technological Forecasting & Social Change, 1413-1430.

Juniper Consultancy Services Ltd. (2007). Advanced Conversion Technology (Gasification) For Biomass Projects. Kaltschmitt, M., & Bridgwater, A. (Eds.). (1997). Biomass gasification and pyrolysis, state of the art and future

prospects. Brussels and Luxembourg / Newbury: European Commission / CPL Scientific Limited. Kaltschmitt, M., Rösch, C., & Dinkelbach, L. (1998). Biomass Gasification in Europe. Luxembourg: Office for

official publications of the European Communities. Karki, M., & Krishnan, K. (1997). Patent citation analysis: a policy analysis tool. World Patent Information, 269-

272. Kavalov, B., & Peteves, S. (2005). Status and perspectives of biomass-to-liquid fuels in the European Union.

Luxembourg: Office for Official Publications of the European Communities. Kemp, R., Rip, A., & Schot, J. (2001). Constructing transitions paths through the management of niches. In R.

Garud, & P. (. Karnoe, Path dependence and creation (pp. 269-299). Mahwah, NJ: Lawrence Erlbaum Associates.

Kirkels, A. (2012). Discursive shifts in energy from biomass: a 30 year European overview. Renewable and Sustainable Energy Reviews, 4105-4115.

Kirkels, A. (2014). Punctuated continuity: the technological trajectory of advanced biomass gasifiers. Energy Policy.

Kirkels, A. (2016). Biomass boom or bubble? A longitudinal study on expectation dynamics. Technological Forecasting & Social Change, 83-96.

Kirkels, A., & Verbong, G. (2011). Biomass gasification: still promising? A 30 year global overview. Renewable and sustainable energy reviews, 471-481.

Klass, D. (1985). Energy from biomass and wastes: a review and 1983 update. Resources and Conservation, 157-239.

Klass, D. (1987). Energy from Biomass and Wastes: 1985 Update and Review. Resources and Conservation, 7-84. Klass, D. L. (1995). Biomass energy in North American policies. Energy Policy, 1035-1048. Kliman, M. L. (1983). Methanol, natural gas, and the development of alternative transportation fuels. Energy,

859-870. Knoef, H. (1996). Biomassavergassing-vastbed; fase 2: systeemdefinitie en conceptueel ontwerp. Utrecht: EWAB

Programme Novem. Knoef, H. (Ed.). (2005a). Handbook Biomass Gasification. Enschede: BTG biomass technology group. Knoef, H. (2005b). Status of small-scale biomass gasification and prospects. In Handbook biomass gasification

(pp. 38-75). Enschede: BTG. Kolb. (2011). Country report Germany for IEA task 33 Thermal gasification of biomass (final 3). Koljonen, J., Kurkela, E., & Wilén, C. (1993). Peat-based HTW-plant at Oulu. Bioresource Technology, 95-101. Konrad, K. (2006). The social dynamics of expectations: the interaction of collective and actor-specific

expectations on electronic commerce and interactive television. Technology Analysis & Strategic Management, 429-444.

Konrad, K., Markard, J., Ruef, A., & Truffer, B. (2012). Strategic responses to fuel cell hype and disappointment. Technological Forecasting & Social Change, 1084-1098.

Koornneef, J., Junginger, M., & Faaij, A. (2007). Development of fluidized bed combustion - an overview of trends, performance and cost. Progress in Energy and Combustion Science, 19-55.

Kopetz, E., Weber, T., Palz, W., Chartier, P., & Ferrero, G. (Eds.). (1998). Biomass for energy and industry. 10th European conference and technology exhibition 1998. Rimpar: C.A.R.M.E.N.

185

Kurkela, E. (1989). Status of peat and biomass gasification in Finland. Biomass, 287-292. Kurkela, E., & Kurkela, M. (2009). BIGPOWER, Advanced Biomass Gasification for High Efficiency Power. VTT.

VTT. Kurkela, E., Stahlberg, P., Laatikainen, J., & Simell, P. (1993). Development of simplified IGCC-processes for

biofuels: supporting gasification research at VTT. Bioresource Technology, 37-47. Kwant, K. W. (1998). Status of Biomass Gasification in countries participating in the IEA Bioenergy gasification

activity. IEA Bioenergy Gasification. Kwant, K. W. (2004). Conference summary. 2nd World Conference on biomass for energy, industry and climate

protection, (pp. l-lii). Kwant, K. W., & Knoef, H. (2004). Status of gasification in countries participating in the IEA and GasNet activity. Kyritsis, S., Beenackers, A., Helm, P., Grassi, A., & Chiaramonti, D. (Eds.). (2001). 1st World conference on

biomass for energy and industry. 2000. London: James & James Ltd. Lagendijk, V., & Verbong, G. (2012). Setting the stage for the energy transition. In G. Verbong, & D. Loorbach

(Eds.), Governing the Energy Transition (pp. 24-50). New York: Routledge. Lamers, P., Hamelinck, C., Junginger, M., & Faaij, A. (2011). International biofuel trade - a review of past

developments in the liquid biofuel market. Renewable and sustainable energy reviews, 2655-2676. Lancia, F. (2010a). T-lab software, Tools for text analysis. Lancia, F. (2010b, September). User's manual T-Lab - tools for text analysis. Retrieved from T lab:

http://www.tlab.it/en/presentation.php Larson, E., Svenningsson, P., & Bjerle, I. (1989). Biomass gasification for gas turbine power generation. In T.

Johansson, B. Bodlund, & R. Williams (Eds.), Electricity. Efficient end-use and new generation technologies, and their planning implications (pp. 697-739). Lund: Lund University Press.

Lehrer, N. (2010). (Bio)fueling farm policy: the biofuels boom and the 2008 farm bill. Agric. Hum. Values, 427-444.

Lente, H. v. (2014, September). Genres of Innovation: storylines of theories of technology. Lightner, V. (2009, June). U.S. Department of Energy, Biomass Program, growing America's energy future.

Retrieved December 2012, from Biomass Research and Development: http://www.usbiomassboard.gov/pdfs/doe_budget_overview_biomass_program_valri_lightnerjg.pdf

Liu, H., Ni, W., Li, Z., & Ma, L. (2008). Strategic thinking on IGCC development in China. Energy Policy, 1-11. Longwell, J., Rubin, E., & Wilson, J. (1995). Coal: energy for the future. Prog. Energy Combust. Sci., 21, 269-360. Lundqvist, R. (1993). The IGCC demonstration plant at Värnamo. Bioresource Technology, 49-53. Maniatis, K. (1998). Overview of EU Thermie gasification projects. Power production from biomass III,

gasification and pyrolysis R&D&D for industry (pp. 9-34). Espoo: VTT Energy. Maniatis, K. (2001). Progress in Biomass Gasification: An Overview. In A. Bridgwater, Progress in

Thermochemical Biomass Conversion (pp. 1-32). Oxford: Blackwell. Maniatis, K. (2005). Message from IEA bioenergy. 14th European biomass conference, (pp. xlix-l). Maniatis, K. (2007). Opening of the closing session and summary of the conference. 15th European biomass

conference, (pp. lxv-lxvi). Maniatis, K., & Ferrero, G. (1995). The Thermie targeted projects on biomass gasification. Proceedings second

biomass conference of the Americas: energy, environment, agriculture, and industry (pp. 576-585). Golden: NREL.

Maniatis, K., & Kotronaros, A. (2002). The bioenergy policy of the European Commission. Twelfth European biomass conference. Biomass for energy, industry and climate protection. (pp. 57-61). Florence: ETA Florence and WIP Munich.

Maniatis, K., & Millich, E. (1998). Energy from biomass and waste: the contribution of utility scale biomass gasification plants. Biomass and Bioenergy, 195-200.

Maniatis, K., Grimm, H.-P., Helm, P., & Grassi, A. (Eds.). (2007a). 15th European biomass conference. From research to market deployment. Florence: ETA Renewable Energies and WIP Renewable Energies.

Maniatis, K., Grimm, H.-P., Helm, P., & Grassi, A. (2007b). Foreword. 15th European biomass conference, (p. liv).

186

Maniatis, K., Guiu, G., & Riesgo, J. (2003). The European Commision perspective in biomass and waste thermochemical conversion. In A. Bridgwater (Ed.), Pyrolysis and gasification of biomass and waste, proceedings of an expert meeting 2002 (pp. 1-18). Berks: CPL Press.

Maniatis, K., Papadoyannakis, M., & Segerborg-Fick, A. (1998). Biomass and waste: themochemical conversion activities in EC programmes. In E. Kopetz, T. Weber, W. Palz, P. Chartier, & G. Ferrero (Ed.), Biomass for energy and industry (pp. 264-267). Rimpar: C.A.R.M.E.N.

Maniatis, K., Rensfelt, E., & Solantausta, Y. (1997). Conclusions of the workshop 'Applications for thermochemical processes'. Developments in thermochemical biomass conversion (pp. 1631-1635). London: Blackie Academic & Professional / Chapman & Hall.

Martin, J.-M. (1996). Energy technologies: systemic aspects, technological trajectories, and institutional frameworks. Technological Forecasting and Social Change, 81-95.

Martins Coelho, J. C. (1988). Second international producer gas conference, March 1985. Stockholm: Beijer Institute of ecological economics.

McCormick, K., & Kaberger, T. (2007). Key barriers for bioenergy in Europe: Economic conditions, know-how and institutional capacity, and supply chain co-ordination. Biomass & bioenergy, 443-452.

McGowin, C. R., & Wiltsee, G. A. (1996). Strategic analysis of biomass and waste fuels for electric power generation. Biomass and Bioenergy, 167-175.

Miles, T., & Miles, T. j. (1989). Overview of Biomass Gasification in the USA. Biomass, 163-168. Millich, E. (2001). Opening session, key note speech. 1st World conference on biomass for energy and industry,

(pp. xlv-xlvi). Mina, A., Ramlogan, R., Tampubolon, G., & Metcalfe, J. (2007). Mapping evolutionairy trajectories: applications

to the growth and transformation of medical knowledge. Research Policy, 789-806. Mitchell, C., Bridgwater, A., Stevens, D., Toft, A., & Watters, M. (1995). Technoeconomic assesment of biomass

to energy. Biomass and Bioenergy, 205-226. Mokyr, J. (1990). The lever of richnes. New York: Oxford University Press. Morris, M., Waldheim, L., Faaij, A., & Stahl, K. (2005). Status of large-scale biomass gasification and prospect. In

H. Knoef, Biomass gasification. Narin, F., & Olivastro, D. (1988). Technology indicators based on patents and patent citations. In A. van Raan

(Ed.), Handbook of Quantitative Studies of Science and Technology. North Holland: Elsevier Publishers. Neavel, R., Knights, C., & Schulz, H. (1981). Exxon Donor Solvent Liquefaction Process [and Discussion].

Phylosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences. Negro, S. (2007). Dynamics of Technological Innovation Systems. Utrecht: Universiteit Utrecht. Nelson, R., & Winter, S. (1982). An Evolutionary Theory of Economic Change. Cambridge (Mass.): Belknap Press. Neuendorf, K. (2004, spring). Content analysis - a contrast and complement to discourse analysis. Qualitative

Methods Newsletter, pp. 33-36. Niele, F. (2005). Energy. Engine of Evolution. Amsterdam: Elsevier. Oberling, D. F., Obermaier, M., Szklo, A., & Lèbre La Rovere, E. (2012). Investments of oil majors in liquid

biofuels: the role of diversification, integration and technological lock-ins. Biomass and Bioenergy, 270-281.

OBP Office of the Biomass Program. (2005). A strong energy portfolio for a strong America. USA: DoE. OECD. (2009). The Bioeconomy to 2030: designing a policy agenda. OECD. OECD/IEA. (2011). R&D Statistics. Retrieved 2011, from IEA International Energy Agency:

http://www.iea.org/stats/rd.asp OECD/IEA. (2014). 2014 Key world energy statistics. Paris: International Energy Agency. OECD/IEA. (2015). R&D Statistics. (2014 edition). Retrieved July 2015, from IEA International Energy Agency:

http://www.iea.org/stats/rd.asp Office of the Biomass Programme. (2009). Biomass. Multi Year Programme Plan. U.S. Department of Energy. Oppenheim, C., & Allen, J. (1979). The overlap of U.S. and Canadian patent literature with journal literature.

World Patent Information, 77-80. Overend, R. (1980/1981). Thermochemical routes to gaseous and liquid fuels. 1st European biomass conference,

(pp. 481-484).

187

Overend, R. (1995). Session D2: Chemicals and products - thermochemistry. In P. Chartier, A. Beenackers, & G. Grassi (Ed.), Biomass for energy, environment, agriculture and industry. Proceedings of the 8th European biomass conference 1994. (pp. 185-189). Oxford: Pergamon / Elsevier Science Limited.

Overend, R., & Chum, H. (1993/1994). Thermochemical conversion research in the U.S. Department of Energy biofuels and power programs. In A. Bridgwater (Ed.), Advances in thermochemical biomass conversion 1992 (pp. 26-47). London: Blackie Academic & Professional / Chapman & Hall.

Paappanen, T., Leinonen, A., & Hillebrand, K. (2006). Fuel peat industry in EU, summary report. VTT. Palmberger, B. (1980). Bioenergy 80: world congress and exhibition (p. 343). Washington: Bio-energy Council. Palonen, J., Lundqvist, R., & Stahl, K. (1996). IGCC technology and demonstration. In K. Sipilä, & M. Korhonen

(Ed.), Power production from biomass II with special emphasis on gasification and pyrolysis R&DD (pp. 41-54). Espoo: VTT.

Palz, W. (2002). Keynote speech. Biomass, the energy of the past and the energy of the future. 12th European biomass conference, (pp. xxx-xxxii).

Palz, W. (2007). Flashback to the conference and announcement of the next European biomass conference. 15th European biomass conference, (p. lxx).

Palz, W., Chartier, P., & Hall, D. (Eds.). (1981). Energy from biomass, 1st E.C. Conference 1980. Brussels and Luxembourg: Commission of the European Communities / Applied Science Publishers.

Palz, W., Coombs, J., & Hall, D. (Eds.). (1985). Energy from biomass, 3rd E.C. Conference 1985. Brussels and Luxembourg: Commision of the European Communities / Elsevier Applied Science Publishers.

Palz, W., Sptizer, J., Maniatis, K., Kwant, K., Helm, P., & Grassi, A. (Eds.). (2002). Twelfth European biomass conference. Biomass for energy, industry and climate protection. 2002. Florence: ETA Florence and WIP Munich.

Pease. (2007). The policy bioenergy debate: can bioenergy deliver the required market penetration? 15th European biomass conference, (pp. lxxi-lxxxix).

Pettersson, E. (1980). Bio-Energy in Sweden. Bioenergy 80: world congress and exposition (pp. 20-22). Washington: Bio-energy Council.

Pinch, T. J., & Bijker, W. E. (1984). The Social Construction of Facts and Artefacts: Or How the Sociology of Science and the Sociology of Technology Might Benefit Each Other. Social Studies of Science, 399-441.

Piterou, A., Shackley, S., & Upham, P. (2008). Project ARBRE: lessons for bio-energy developers and policy makers. Energy Policy, 2044-2050.

Raldow, W. (2004). Keynote speech, European Commission. Second world biomass conference. Biomass for energy, industry and climate protection. (pp. xxx-xxxi). Florence: ETA Florence and WIP Munich.

Raven, R. (2006). Towards alternative trajectories? Reconfigurations in the Dutch electricity regime. Research Policy, 581-595.

Reed, T. (1980a). A survey of biomass gasification. Volume III - Current technology and research. Golden: Solar Energy Research Institute.

Reed, T. (1980b). Biomass gasification: yesterday, today and tomorrow. Bioenergy 80 (pp. 190-193). Washington: Bioenergy Council.

Reed, T., & Gaur, S. (2001). A survey of biomass gasification 2001: gasifier projects and manufacturers around the world. Biomass Energy Foundation.

Reed, T., & Miles, T. (1988). Gasification workshop. In A. Bridgwater, & J. Kuester (Ed.), Research in thermochemical biomass conversion (Phoenix, 1988) (pp. 1189-1191). London and New York: Elsevier Applied Science.

Reed, T., & others. (1979). Survey of biomass gasification. Volume I - Synopsis and executive summary. Golden: Solar Energy Research Institute.

REN21 Renewable Energy Policy Network for the 21st Century. (2014). The First Decade: 2004-2014. Paris: REN21.

Rensfelt, E. (1984). Practical achievements in biomass gasification. In H. Egneus, & A. Ellegard (Ed.), Bioenergy 84. I, pp. 174-204. London: Elsevier Applied Science Publishers.

188

Rensfelt, E. (1991). Gasification for power productions. In G. Grassi, & I. Bertini (Ed.), 1st European forum on electricity production form biomass and solid wastes by advanced technologies (pp. 65-79). Brussels and Luxembourg: Commission of the European Communities. doi:EUR 14689 EN

Rensfelt, E., & Gobel, B. (2003). Strategies for development and implementation of biomass gasification. Pyrolysis and gasification of biomass and waste (pp. 681-686). Newbury / Burmingham: CPL Press / Aston University.

Rensfelt, E., Morris, M., & Waldheim, L. (2004). Project ARBRE, UK - A wood fuelled combined-cycle demonstration plant. Biomasse-Vergasung - Internationale Tagung Leipzig, Oktober 2003 (pp. 204-224). Münster: Landwirtschaftsverlag.

Rip, A., & Kemp, R. (1998). Technological change. In S. Rayner, & E. Malone (Eds.), Human choice and climate change (Vol. 2, pp. 327-399). Columbus: Batelle Press.

Rösch, C., & Kaltschmitt, M. (1998). Economics. In M. Kaltschmitt, C. Rösch, & L. Dinkelbach (Eds.), Biomass gasification in Europe (pp. 183-196). Brussels / Luxembourg: European Commission.

Rösch, C., & Kaltschmitt, M. (1999). Energy from biomass - do non-technical barriers preven an increased use? Biomass and bioenergy, 347-356.

Rösch, C., Kaltschmitt, M., & Dinkelbach, L. (1998). Research, development and demonstration. In M. Kaltschmitt, C. Rösch, & L. Dinkelbach (Eds.), Biomass gasification in Europe (pp. 159-182). Brussels / Luxembourg: European Commission.

Ruef, A., & Markard, J. (2010). What happens after a hype? How changing expectations affected innovation activities in the case of stationary fuel cells. Technology Analysis & Strategic Management, 317-338.

Salo, K. (1998). Kotka Ecopower IGCC-project, the attempt to transfer the Biocycle project to Finland. Biomass and bioenergy, 225-228.

Salo, K., & Keränen, H. (1996). Biomass IGCC. In K. Sipilä, & M. Korhonen (Ed.), Power production from biomass II with special emphasis on gasification and pyrolysis R&DD (pp. 23-39). Espoo: VTT.

Salo, K., & Patel, J. (1997). Integrated gasification combined cycle based on pressurized fluidized bed gasification. In A. Bridgwater, & D. Boocock (Ed.), Developments in thermochemical biomass conversion (pp. 994-1005). London: Blackie Academic & Professional / Chapmann & Hall.

Sandelowski, M. (2014). Unmixing mixed-methods research. Research in Nursing & Health, 3-8. Sandén, B. A. (2004). Technology path assessment for sustainable technology development. Innovation:

management, policy & practice, 316-330. Sandén, B. A., & Jonasson, K. M. (2005). Variety creation, growth and selection dynamics in the early phases of a

technological transition. The development of alternative transport fuels in Sweden 1974-2004. Göteborg: Chalmers University of Technology.

Scheer, H. (1995). Towards a strategy for development and propagation of bio-energy in the single European market. 8th European biomass conference, (pp. 5-10).

Schiefelbein, G. (1989). Biomass thermal gasification research: recent results from the United States DOE's research program. Biomass, 145-159.

Schmid, J. (Ed.). (2008). 16th European biomass conference & exhibition. From research to industry and markets. Florence: ETA Florence Renewable Energies.

Schmidt, A. (1992). Biomass projects in Europe. 6th European biomass conference, (pp. 95-100). Schot, J., & Geels, F. W. (2007). Niches in evolutionary theories of technical change. A critical survey of the

literature. Journal of Evolutionary Economics, 605-622. Schwager, J., & Whiting, K. (2003, Oct 14). Progress towards commercialising waste gasification: A worldwide

status report. Retrieved Sept 8, 2009, from Website Gasification Technologies Council: http://www.gasification.org/Docs/Conferences/2003/22SCHW.pdf

Sengers, F., Raven, R., & Venrooij, A. v. (2010). From riches to rags: biofuels, media discourses and resistance to sustainable energy technologies. Energy Policy, 5013-5027.

Shand, R., & Bridgwater, A. (1984). Fuel gas from biomass: status and new modelling approaches. In A. Bridgwater (Ed.), Thermochemical processing of biomass (pp. 229-254). London: Butterworths & Co (Publishers) Ltd.

189

Shore, J., & Hackworth, J. (2007). Trends and transitions in the diesel market, NPRA meeting 2007. Retrieved July 2014, from U.S. Energy Information Administration: http://www.eia.gov/pub/oil_gas/petroleum/presentations/2007/npra2007/npra2007.ppt

Sims, R. E., Mabee, W., Saddler, J. N., & Taylor, M. (2010). An overview of second generation biofuel technologies. Bioresource Technology, 1570-1580.

Sipilä, K. (1993). New power-production technologies: various options for biomass and cogeneration. Bioresource Technology, 5-12.

Sipilä, K., & Korhonen, M. (Eds.). (1993a). Bioresource Technology, 46, 1-188. Sipilä, K., & Korhonen, M. (1993b). Preface. Bioresource Technology, 1. Sipilä, K., & Korhonen, M. (Eds.). (1996). Power production from biomass II with special emphasis on gasification

and pyrolysis R&DD. Espoo: VTT. Sipilä, K., & Korhonen, M. (Eds.). (1998). Power production from biomass III; Gasification and pyrolysis R&D&D

for industry. Espoo: VTT. Sipilä, K., & Rossi, M. (Eds.). (2002). Power production from waste and biomass IV; advanced concepts and

technologies. Espoo: VTT. Sipilä, K., Kurkela, E., & Solantausta, Y. (1993/1994). New options for biomass-based power production by IGC.

A Finnish national research programme - JALO. In A. Bridgwater (Ed.), Advances in thermochemical biomass conversion 1992 (pp. 77-86). Glasgow: Blackie Academic & Professional.

Sjunneson, L., Carrasco, J., Helm, P., & Grassi, A. (Eds.). (2005a). 14th European biomass conference. Biomass for energy, industry and climate protection. 2005. ETA Florence and WIP Munich.

Sjunneson, L., Carrasco, J., Helm, P., & Grassi, A. (2005b). Foreword. 14th European biomass conference, (p. xli). Smil, V. (1999). Energies. An illustrated guide to the biosphere and civilization. Cambridge, Massachusetts: MIT

Press. Smith, A., Stirling, A., & Berkhout, F. (2005). The governance of sustainable socio-technical transitions. Research

Policy, 1491-1510. Snepvangers, J., & Stork, J. (1987). Kolenvergassing anno 1987. Sittard - Netherlands: NEOM Nederlandse

Energie Ontwikkelings Maatschappij BV. Solantausta, Y., & Beenackers, A. (1989). Workshop 3 - Gasification technology and economics. In G. Ferrero, K.

Maniatis, A. Buekens, & A. Bridgwater (Ed.), Pyrolysis and gasification (pp. 396-398). Brussels and Luxembourg: Elsevier Science Publisher / Commission of the European Communities.

Solantausta, Y., Bridgwater, A., & Beckman, D. (1995). Feasibility of power production with pyrolysis and gasification systems. Biomass and Bioenergy, 257-269.

Sorda, G., Banse, M., & Kemfert, C. (2010). An overview of biofuel policies across the world. Energy Policy, 6977-6988.

Sperling, D., & DeLuchi, M. (1989). Is methanol the transportation fuel of the future? Energy, 469-482. Spitzer, J. (2002). Amsterdam 2002 - what have we accomplished? 12th European biomass conference, (pp.

xxxvii-xxxix). Spitzer, J. (Ed.). (2010). 18th European biomass conference and exhibition. From research to industry and

markets. Florence: ETA Florence Renewable energies. Stahl, K., Waldheim, L., Morris, M., Johnsson, U., & Gardmark, L. (2004). Biomass IGCC at Värnamo, Sweden -

Past and future. GCEP Energy Workshop. California: Stanford University. Stassen, H. E. (1995). Small scale biomass gasifiers for heat and power, a global review. Washington: The World

Bank. Stassen, H., & van Swaaij, W. (1982). Application of biomass gasification in developing countries. Energy from

biomass, 2nd E.C. Conference (pp. 705-714). Brussels and Luxembourg: Commission of the European Communities / Applied Science Publishers.

Steinert, M., & Leifer, L. (2010). Scrutinizing Gartner's hype cycle approach. Technology Management for Global Economic Growth. IEEE.

Stelson, T. (1980). U.S. Government programs in bio-energy. Bioenergy 80, Atlanta (pp. 22-24). Washington, D.C.: Bio-energy Council.

190

Sterzinger, G. J. (1994). Integrated gasification combined cycle and steam injection gas turbine powered by biomass joint-venture evaluation. Golden, Colorado: NREL.

Stevens, D. (1992). Biomass conversion: an overview of the IEA Bioenergy Agreement Task VII. Biomass and Bioenergy, 213-227.

Stevens, J. (1994). Review and Analysis of the 1980-1989 Biomass Thermochemical Conversion Program. Golden, Colorado: NREL.

Ström, E., Liinanki, L., Sjöström, K., Rensfelt, E., Waldheim, L., & Blackadder, W. (1985). Gasification of biomass in the MINO-process. In H. Egneus, & A. Ellegard (Ed.), Bioenergy 84 (pp. 57-64). London: Elsevier Applied Science Publishers Ltd.

Strub, A. (1984). The Commission of the European Communities R&D Programma 'Energy from Biomass'. In A. Bridgwater (Ed.), Thermochemical processing of biomass (pp. 1-10). Butterworths.

Strub, A. (1985). The energy from biomass programme of the Commission of the European Communities. 3rd European biomass conference, (pp. 3-5).

Strub, A., Chartier, P., & Schleser, G. (Eds.). (1983a). Energy from biomass, 2nd E.C. Conference 1982. Brussels and Luxembourg: Commission of the European Communities / Applied Science Publishers.

Strub, A., Chartier, P., & Schleser, G. (1983b). Preface. 2nd European biomass conference, (p. v). Synthetic Fuels Associates. (1982). Technical evaluation of wood gasification. Palo Alto, California. Tanaka, R., & Johnson, P. (2005, June). Alternatives to landfill: an overview of Japan's incineration policies and

technologies for handling municipal solid waste. Science and Innovation Section, British Embassy Tokyo. Thomson Reuters. (2010, April). Web of Science Coverage Expansion. Retrieved 04 2011, from Thomson Reuters

Community: http://community.thomsonreuters.com/t5/Citation-Impact-Center/Web-of-Science-Coverage-Expansion/ba-p/10663

Thomson Reuters. (2014, February). Web of Science. Thomson Reuters. (2015a). Derwent Innovation Index. Thomson Reuters. (2015b, 9 30). Derwent Innovations Index. Retrieved from Thomson Reuters website:

http://thomsonreuters.com/en/products-services/scholarly-scientific-research/scholarly-search-and-discovery/derwent-innovations-index.html

Thomson Reuters. (2015c, July). Web of Science. Thomson Reuters. (2015d, 09 30). Web of Science core collection. Retrieved from Thomson Reuters website:

http://thomsonreuters.com/en/products-services/scholarly-scientific-research/scholarly-search-and-discovery/web-of-science-core-collection.html

Trajtenberg, M. (1990). A penny for your quotes: patent citations and the value of innovations. RAND Journal of Economics, 172-187.

Turkenburg, W. C. (2000). Renewable Energy Technologies. In UNDP, World Energy Assessment. Energy and the challenge of sustainability. (p. 266). New York: UNDP.

U.S. Department of Energy (DoE). (april 1992). Electricity from biomass, a development strategy. DoE. U.S. Department of Energy (DoE) National Energy Technology Laboratory (NETL). (2000). Gasification,

Worldwide Use and Acceptance. SFA Pacific. U.S. Department of Energy (DoE), National Energy Technology Laboratory (NETL) . (2005). Gasification, Current

Industry Perspective, Robust Growth Forecast. World Survey Results 2004. NETL. U.S. Department of Energy (DoE), National Energy Technology Laboratory (NETL). (2007a). Gasification World

Database 2007, Current Industry Status, Robust Growth Forecast. NETL. U.S. Department of Energy (DoE), National Energy Technology Laboratory (NETL). (2007b, September).

Gasification - Gasification database [Exces XLS file]. Retrieved May 2008, from NETL, the ENERGY lab: http://www.netl.doe.gov/technologies/coalpower/gasification/database/database.html

U.S. Department of Energy (DoE), National Energy Technology Laboratory. (2007c). Gasification Technologies Project Portfolio. Pittsburgh: NETL.

Ulmanen, J. H., Verbong, G. P., & Raven, R. P. (2009). Biofuel developments in Sweden and the Netherlands. Protection and socio-technical change in a long term perspective. Renewable and Sustainable Energy Reviews, 1406-1417.

UNDP. (2004). World Energy Assessment: Overview 2004 update. New York: UNDP.

191

USDA United States Department of Agriculture. (1983). A biomass energy production and use plan for the United States 1983-1990. Washington, D.C.: USDA.

USPTO United States Patent and Trademark Office. (1976-2008). Retrieved 2009-2015, from Patent full text and full image database: http://patft.uspto.gov/

USPTO United States Patent and Trademark Office. (1976-2013). Retrieved July 2014, from Patent full text and full image database: http://patft.uspto.gov/

van Lente, H. (1993). Promising technology. The dynamics of expectations in technological developments. Delft: Eburon.

van Lente, H. (2012). Navigating foresight in a sea of expectations: lessons from the sociology of expectations. Technology Analysis & Strategic Management, 769-782.

van Lente, H., & Rip, A. (1998). The rise of membrane technology: from rhetorics to social reality. Social Studies of Science, 221-254.

van Lente, H., Spitters, C., & Peine, A. (2013). Comparing technological hype cycles: towards a theory. Technological Forecasting & Social Change, 1615-1628.

van Swaaij, W. (1981). Gasification - the process and the technology. Energy from biomass 1980 (pp. 485-495). London: Applied Science Publishers.

van Swaaij, W., Fjällström, T., Helm, P., & Grassi, A. (Eds.). (2004a). Second world biomass conference. Biomass for energy, industry and climate protection. 2004. Florence: ETA Florence and WIP Munich.

van Swaaij, W., Prins, W., & Kersten, S. (2004b). Strategies for the future of biomass for energy, industry and climate protection. 2nd World Conference on biomass for energy, industry and climate protection, (pp. xxxix-xlii).

van Thuijl, E., Roos, C., & Beurskens, L. (2003). An overview of biofuel technologies, markets and policies in Europe. Petten: ECN.

Verbong, G., & Loorbach, D. (2012). Governing the Energy Transition. New York: Routledge. Verbong, G., Christiaens, W., Raven, R., & Balkema, A. (2010). Strategic Niche Management in an unstable

regime: biomass gasification in India. Environmental Science & Policy, 272-281. Verbong, G., Geels, F., & Raven, R. (2008). Multi-niche analysis of dynamics and policies in Dutch renewable

energy innovation journeys (1970-2006): hype-cycles, closed networks and technology-focused learning. Technology Analysis & Strategic Management, 555-573.

Verspagen, B. (2007). Mapping technological trajectories as patent citation networks: a study on the history of fuel cell research. Advances in Complex Systems, 93-115.

Verspagen, B. (n.d.). Citpath software. Waldheim, L. (May 2012). IEA Biomass Agreement Task 33, Country report Sweden 2012. Waldheim Consulting. Walter, D. (1988). Biofuels and municipal waste technology, an overview of the DOE program. In A. Bridgwater,

& J. Kuester (Ed.), Research in thermochemical biomass conversion (Phoenix, 1988) (pp. 10-15). London and New York: Elsevier Applied Science.

Watson, W. (1997). Constructing success in the electric power industry: combined cycle gas turbines and fluidised beds (PhD thesis). Sussex: SPRU, University of Sussex.

Watts, R., & Porter, A. (1997). Innovation Forecasting. Technological Forecasting and Social Change, 25-47. WCED World Commission on Environment and Development. (1987). Our common future [Brundtland report].

Oxford / New York: Oxford University Press. Wikipedia contributors. (2011a, May 18). Price of petroleum. Retrieved May 2011, from Wikipedia, the free

encyclopedia: http://en.wikipedia.org/w/index.php?title=Price_of_petroleum&oldid=429767547 Wikipedia contributors. (2011b, April 28). Food vs. Fuel. Retrieved 2011, from Wikipedia, the free encyclopedia:

http://en.wikipedia.org/w/index.php?title=Food_vs._fuel&oldid=426304070 Wikipedia contributors. (2014, Januari). Range Fuels. Retrieved July 2014, from Wikipedia, The Free

Encyclopedia: http://en.wikipedia.org/w/index.php?title=Range_Fuels&oldid=591804228 Wilén, C., & Kurkela, E. (1998). Biomass gasification in Finland. In M. Kaltschmitt, C. Rösch, & L. Dinkelbach

(Eds.), Biomass gasification in Europe (pp. 113-132). Brussels / Luxembourg: European Commission. Williams, L. (1981). Opening speech. 1st European biomass conference, (pp. 14-18).

192

Williams, R., & Larson, E. (1989). Expanding roles for gas turbines in power generation. In T. Johansson, B. Bodlund, & R. Williams (Eds.), Electricity. Efficient end-use and new generation technologies, and their planning implications (pp. 503-553). Lund: Lund University Press.

Williams, R., & Larson, E. (1993). Advanced gasification based biomass power generation. In T. Johansson, H. Kelly, A. Reddy, & R. Williams (Eds.), Renewable energy. Sources for fuels and electricity. (pp. 729-785). Island Press.

WIPO. (n.d.). International Patent Classification. Retrieved 2015, from http://www.wipo.int/classifications/ipc/en/

World Energy Council. (2013). World Energy Resources, 2013 survey. World Energy Council. Wu, C., Yin, X., Yuan, Z., Zhou, Z., & Zhuang, X. (2010). The development of bioenergy technology in China .

Energy, 4445–4450. Yoon, B., & Park, Y. (2004). A text-mining-based patent network: analytical tool for high-technology trend.

Journal of High Technology Management Research, 37-50. Zschach, U., von Cramon-Taubadel, S., & Theuvsen, L. (2010). Public interpretations in the discourse of

bioenergy - A qualitative media analyis. Berichte Uber Landwirtschaft, 502-512.

193

194

Appendix A Gasifier types Gasification has a long history (Harmsen, 2000; Knoef, 2005a). It was used for production of town gas in the 19th and 20th centuries. Coal gasification was explored in situations of oil shortages (Germany in World War II, South Africa facing an international oil boycott under apartheid) and in response to the oil crises of the 1970s. The universal gasifier that is able to handle different types of feedstock and serve different applications does not exist. Several generic types of gasification technologies have been developed and demonstrated, each offering certain advantages and disadvantages with respect to handling feedstock, plant size, gas quality, reliability, efficiency, etcetera – see for a description for example Kaltschmitt et al. (1998), Knoef (2005a), E4tech (2009) and Bacovsky et al. (2010). Gasifiers can be classified in different ways:

• Choice of the gasifying agent (air, oxygen or steam) • Heat supply (autothermal or directly heated gasifiers in which the heat

required for the gasification reaction is provided by partial combustion of the biomass; or allothermal or indirectly heated gasifiers in which the heat is provided by an external source, such as the circulation of inert material or steam)

• The pressure in the gasifier. Pressurized gasification is more complex and costly, but it “provides higher throughputs, with larger maximum capacities, promotes hydrogen production and leads to smaller, cheaper downstream cleaning equipment. Furthermore, since no additional compression is required, the syngas temperature can be kept high for downstream operations and liquid fuels catalysis.” (E4tech, 2009, p3)

• The temperature range in which the gasifier is operated • How the biomass is fed into the gasifier and is moving around in it.

The most common generic designs are fixed bed gasifiers (either updraft or downdraft gasifiers), entrained flow gasifiers, fluidized bed gasifiers (bubbling, circulating or dual fluidized beds), and plasma gasifiers – see table A1. Of course a wider variety of gasifier designs exist, including combinations of these basic design.

195

Table A1 Main gasifier types. Note that biomass particles are shown in green, and bed material in blue. Reproduced by courtesy of E4tech from “Review of technology for the

gasification of biomass and wastes” (June 2009, p4/5), see http://www.e4tech.com/reports/review-of-technologies-for-gasification-of-biomass-and-wastes/

Updraft fixed bed • The biomass is fed in at the top of the gasifier, and the air,

oxygen or steam intake is at the bottom, hence the biomass and gases move in opposite directions

• Some of the resulting char falls and burns to provide heat • The methane and tar-rich gas leaves at the top of the gasifier,

and the ash falls from the grate for collection at the bottom of the gasifier

Downdraft fixed bed • The biomass is fed in at the top of the gasifier and the air, and

oxygen or steam intake is also at the top or from the sides, hence the biomass and gases move in the same direction

• Some of the biomass is burnt, falling through the gasifier throat to form a bed of hot charcoal which the gases have to pass through (a reaction zone)

• This ensures a fairly high quality syngas, which leaves at the base of the gasifier, with ash collected under the grate

Entrained flow (EF) • Powdered biomass is fed into a gasifier with pressurized

oxygen and/or steam • A turbulent flame at the top of the gasifier burns some of the

biomass, providing large amounts of heat, at high temperature (1200-1500°C), for fast conversion of biomass into very high quality syngas

• The ash melts onto the gasifier walls, and is discharged as molten slag

Bubbling fluidized bed (BFB) • A bed of fine inert material sits at the gasifier bottom, with

air, oxygen or steam being blown upwards through the bed just fast enough (1-3m/s) to agitate the material

• Biomass is fed in from the side, mixes, and combusts or forms syngas which leaves upwards

• Operates at temperatures below 900°C to avoid ash melting and sticking. Can be pressurized

196

Table A1 –continued. Main gasifier types.

Circulating fluidized bed (CFB) • A bed of fine inert material has air, oxygen or steam blown

upwards through it fast enough (5-10m/s) to suspend material throughout the gasifier

• Biomass is fed in from the side, is suspended, and combusts providing heat, or reacts to form syngas

• The mixture of syngas and particles are separated using a cyclone, with material returned into the base of the gasifier

• Operates at temperatures below 900°C to avoid ash melting and sticking. Can be pressurized

Dual fluidized bed (Dual FB) • This system has two chambers – a gasifier and a

combustor • Biomass is fed into the CFB / BFB gasification chamber,

and converted to nitrogen-free syngas and char using steam

• The char is burnt in air in the CFB / BFB combustion chamber, heating the accompanying bed particles

• This hot bed material is then fed back into the gasification chamber, providing the indirect reaction heat

• Cyclones remove any CFB chamber syngas or flue gas • Operates at temperatures below 900°C to avoid ash

melting and sticking. Could be pressurized

Plasma • Untreated biomass is dropped into the gasifier, coming

into contact with an electrically generated plasma, usually at atmospheric pressure and temperatures of 1,500-5,000°C

• Organic matter is converted into very high quality syngas, and inorganic matter is vitrified into inert slag

• Note that plasma gasification uses plasma torches. It is also possible to use plasma arcs in a subsequent process step for syngas clean-up

197

198

Appendix B Demo plants In this thesis I draw upon the concept of demonstration plants, as these play a crucial role in bringing a technology to the market. Especially in chapter 4 I draw upon an inventory of pilot and demonstration plants for the reconstruction of technological trajectories of biomass gasification. The overview of demo plants contributes to this reconstruction as the overview can be considered a representation of technological knowledge; it allows to identify heterogeneous actors involved in financing, building and operating these plants; the plants themselves can be considered an indicator of success – not all technologies are demonstrated; and the plants serve as a proof-of-principle for configurational technologies by showing the (un)successful interaction of different components. However, the terminology used for describing demo plants is not standardized. Therefore I will discuss here more extensively the role of demo plants; the terminology applied; and how this was dealt with in this thesis. Recently, Frishammer, Söderholm, Bäckström, Hellsmark and Ylinenpää (2015) provided a review on the role of pilot and demonstration plants (PDPs) in technological development – on which I will draw. They argue that pilot and demonstration plants play a central role in process industries, like chemical industry, drug industry, energy industry and in steel making. These plants bridge between knowledge creation on the one hand and industrial application and commercial adoption on the other. The role of PDPs have been picked up by different streams of literature: engineering and natural science research, technology and innovation management, and innovation systems. Each of these literature streams comes with a different goal and different system boundaries to apply. But there are also many commonalities with respect to the role of PDPs: they serve upscaling, uncertainty reduction, and learning. However, the overall conclusion is that the role of PDPs in innovation management and policy is far from sufficiently understood. An important observation by Fishammer et al. is that PDPs are not a homogenous phenomenon and as such can serve different goals. Some are more closely related to the R&D stage while others are more closely to market application. “As we move from the R&D stage towards diffusion the system boundaries become wider, and there is a need to focus not only on technical but also on institutional and market-related uncertainties.” (Frishammer, 2015, p12) To address the inhomogeneous nature of PDPs often a more detailed classification is used to categorize existing plants – and thereby come to an understanding of the different technologies and the market. Two possible classifications are shown in table B.1 and B.2, as used in status reports on biomass and waste gasification. Over chapter

199

4 and 5 I applied the terminology from table B.1 by making a distinction between pilot, demonstration and commercial plants. The process of classifying plants can be challenging and is to some extent open for contestation. First, it is based on multiple characteristics and a specific plant might not show a coherent profile over all of them – scoring different on several characteristics. Second, full and transparent information is often lacking. Laboratory or pilot projects might be realized ‘in house’ with limited public disclosure of test results or status. While for demonstration and semi-commercial units it is important to connect to the market: they can be considered show-cases constructed to attract attention. As such, manufacturers might underplay costs and problems of these plants. In addition, there might be an information asymmetry between more and less successful plants. In my experience especially the latter are less disclosed.

Pilot facility Facility, which does not operate continuously; facility not embedded into an entire material logistic chain; only the feasibility of selected technological steps is demonstrated; the product might not be marketed.

Demonstration facility Facility demonstrating the capability of the technology for continuous production (operated mainly continuously); the facility covering the entire production process or embedded into an entire material logistic chain; the product is being marketed; facility may not be operated under economical objectives.

Commercial facility Facility operated continuously with high level of availability; facility operated under economical objectives; the product is being marketed.

Table B.1 Terminology of pilot and demonstration plants as applied by Bacovsky et al. (2010, p.9)

Bench scale / conceptual

Laboratory scale pre-pilot testing. This category is also used when a process has been significantly modified such that an earlier demonstration is not applicable.

Pilot First small scale development unit. Demonstrator Semi-commercial installation that is being commissioned or is already

operating as a first reference installation. Frequently operated by the developer themselves and used for further refining the process.

Semi-commercial A full scale commercial plant has been commissioned and supplied to a customer or is operating satisfactorily under a ‘Build, Own, Operate’ (BOO) contract & the company is pursuing further opportunities.

Fully commercial Multiple full scale plants have been delivered to more than one customer and these have operated satisfactorily for more than two years.

Table B.2 Terminology of pilot and demonstration plants as applied by Heerman et al. (2000/2001, pE.276)

200

Summary ‘The Green Struggle’ Energy is crucially important for our economy and society. Currently fossil fuels dominate the energy portfolio, causing political and social concerns regarding availability and environmental impact. Renewable energy technologies are gaining headway, but in the short term, their application remains limited. A faster uptake of renewables requires an understanding of how technologies develop and diffuse. Evolutionary scholars like Dosi (1982) and Nelson and Winter (1982), argue that technological developments are highly selective and path dependent, evolving over technological trajectories. This phenomenon is down to shared routines: engineers in a technological field work in more or less the same direction. Quasi-evolutionary scholars argue that this perspective largely neglects the agency by a broader variety of actors involved in the innovation and diffusion process. To underline this different perspective, these scholars coined the term path creation. Path dependency and path creation are often studied using different methodologies. I argue that to understand emerging technological trajectories, we need to analyze the factors that lead to path creation as well as those that contribute to the further development of technological trajectories–and therefore should be studied using both methodologies. That is why I studied the case of biomass gasification in depth. The availability of biomass is huge. Gasification technology offers the opportunity to apply this biomass for high-end applications and has understandably attracted a great deal of interest over the past four decades. My study aims to understand developments in biomass gasification and its dynamics. I posed two research question: 1) How did the technology of biomass gasification develop and how can this be explained; and 2) Did this result in emerging trajectories and what do a path creation and path dependency perspective contribute to their reconstruction? I draw on an evolutionary perspective of technological change, considering variation and selection processes and the competition between technologies at different levels of aggregation. In addition, reconstructing emerging trajectories requires a longitudinal perspective in order to capture multiple product sequences and a global perspective, as developments were carried out by a transnational field. The development of biomass gasifiers is embedded in a broader technological context of energy-from-biomass technologies and renewables in general. Chapter 2 describes how I applied a discourse analysis of RD&D to these broader developments and policy

201

on energy-from-biomass. It shows the changes over time with respect to considered renewable, feedstock, conversion technology, and application as well as supporting arguments. This enabled me to not only compare developments in biomass gasification with those in the closely related technologies of biomass combustion and biomass pyrolysis, but also create a timeline with the key drivers of developments. I reconstructed the discourse in Western Europe between 1980 and 2010 by studying open literature and European Biomass Conference papers. I also conducted a quantitative content analysis of conference titles. Four different discourses emerged based on differentiation into scale and knowledge intensity, but also relating to feedstock and conversion technology. In this way, the complex developments can be structured and understood as shifts between and within discourses. This is especially relevant as each discourse involves a different policy arena and different actors. In chapter 3 I present an overview of the (dynamics in) worldwide development of biomass gasification as part of the more general field of gasification, which is dominated by entrained flow coal gasifiers. Based on literature study and trend analysis of science and technology indicators, my findings show that both biomass and coal gasification are subjected to similar driving forces. However, biomass and coal have different characteristics. Consequently the different technologies applied for each one on a practical and industrial level limit the linkage between both developments. I identified two periods of large momentum. The first began in the mid-1970s and lasted until the mid-1980s. There was great interest in coal gasification, dominated by the US. The second period relates more to biomass gasification. Momentum took off in the mid-1990s and has been rising since. Efforts in Europe dominated this period although China and Japan were catching up. The technology has been successfully applied in a few niche markets, but largely remains confined to research, development and demonstration niches. High-end applications like Integrated Gasification Combined Cycles (IGCC) and transport fuels are gaining ground in research and development. However, biomass gasification is not yet mature enough for widespread application in the market. It is still in a stage of variation and there is no dominant design yet. In most markets it cannot compete with other technologies like natural gas based processes and biomass combustion. Thus I conclude that it is unlikely to break through in the short term. A gradual niche development seems much more probable. From chapters 2 and 3 I conclude that advanced biomass gasifiers received strong transnational support, while developments in small-scale fixed-bed gasifiers were less driven by a transnational field and not well represented by patents and scientific articles. Thus in chapter 4 I turn to tracing the emerging technological trajectory of

202

advanced biomass gasifiers by studying Geels and Raven (2006), whose approach conceptualizes the emergence of trajectories as the result of shared expectations between actors. It requires a real-time perspective of expectations, revealing how stakeholders at specific times perceived drivers, problems, opportunities, solutions, and technologies. Combining this perspective with a longitudinal study allowed me to study the emergence of trajectories based on a rich body of literature, mainly conference contributions, and an overview of demonstration plants. Over the late 1970s and early 1980s, the focus was on methanol production, mainly by indirectly heated or oxygen blown gasifiers. During the 1990s, the key focus was IGCC, for which air blown fluid bed gasifiers were considered (both atmospheric and pressured; both BFB and CFB). After 2003, the focus shifted to the production of biofuels, considering various technologies. The most significant factor in the growing interest and emergence of trajectories was the socio-economic context. This context offered windows of opportunity and direction to developments. Changes in this context resulted in paradigm shifts and discontinuities in developments, which in turn demonstrated a change in considered end-products and technologies, as well as the companies involved. Further impact came from firm-specific differences, like the focus on specific feedstock, scale and more recently type of biofuels to produce. These minor differences were strengthened by the national focus of supportive policies, as well as specific focus on multiple technologies in US and EU policies. The technology’s long-term development probably benefitted from this variation and from niche developments. However, the field had a hard time overcoming the technological challenges and discontinuities ensuing from changes in socio-economic context. While chapter 4 reveals a strongly aligned focus of developments and timing of support, it does not fully explain these phenomena: the role of various stakeholders remains under-exposed; nor does it fully capture the dynamics in expectations, including self-reinforcing mechanisms. To address these aspects, I make use of the literature on the sociology of expectations. In chapter 5 I trace the dynamics in field-level expectations by performing a longitudinal study, covering the transnational interest. This work is based on the same body of literature as in chapter 4, using science and technology indicators to strengthen the validity of my reconstruction of dynamics. This required differentiating between different levels of expectations, time frames, and actors involved. I reconstructed two hype-disappointment cycles: the first focused on methanol production during the early 1980s, the second on IGCC during the 1990s. A further period after 2003 focusing on biofuels, showed a more gradual development and a much greater momentum. This approach offers insight into how developments over the different periods are related. I discuss how expectation

203

dynamics are represented in graphs and influenced by underlying mechanisms. In addition, I reflect on the benefits of a mixed-method approach beyond cross-examination and empirical complementarity: it offers a reality check for quantitative research by showing how specific datasets and algorithms relate to practice; moreover, a reality check for qualitative research that corrects for subjectivity and cherry picking through literature selection or selective interpretation by the scholar. In chapter 6 I reconstruct the technological trajectory for advanced gasifiers–just like in chapter 4, but with a different approach. This is based on Verspagen’s approach (2007) that identifies in retrospect the technological trajectories (called top paths) in patent-citation networks. I did so for patents on fluid bed technology, which over the years has been the dominant technology among advanced biomass gasifiers. In order to link these trajectories to broader developments and a path creation perspective, I complemented Verspagen’s approach with a real-time overview of broader developments in the patents set, along with an analysis of how different developments are related. It showed that the whole technological field is characterized by highly interrelated technological developments. While prior chapters indicate that the social field evolved around biomass gasification, the findings in this chapter show that the technological development evolved around fuel-flexible fluid-bed technology. The dominant top path shows co-evolution with biomass combustion technology and later on waste incineration/gasification. The top path mainly includes patents by Foster Wheeler, Ahlstrom and Ebara. In chapter 7 I draw overall conclusions and discuss implications. I summarize the findings in previous chapters regarding technologies applied, end-use applications, dominant manufacturers and countries involved, and overall dynamics. These developments can be explained by different driving forces: largely exogenous forces (socio-economic and environmental challenges, national policies, competition between technologies); technology related forces (functionality of technology, maturity of technology and technological progress, spillover effects); and forces related to social spheres (alignment to broader technological vision, hype-disappointment mechanisms). I show that developments in advanced biomass gasifiers did result in emerging trajectories. Both the Geels and Raven approach to path creation through shared expectations as well as the Verspagen approach to path dependency by analyzing patent-citation networks reveal these emerging trajectories. But trajectories with a difference, both in empirical findings and the nature of developments. The Geels and Raven approach emphasizes shared expectations and shared focus, in line with the

204

idea of a dominant paradigm. There is less focus on product sequences and the cumulative nature of technology and knowledge development. Although we can identify parallel or subsequent developments in the literature, their interrelationship is not always clear. The Verspagen approach has a strong focus on product sequences and dominant paths. There is some tension between the linearity of this path perspective and the more complex network structure of technological developments, as the underlying patent-citation networks show. Expectations and broader contextual factors remain out of scope. I argue that the differences in both approaches not only reflect the contrast in theoretical starting points (path creation versus path dependency), but also in the methodologies applied, including the underlying data. I conclude that both methods are complementary. Combining the approaches to studying emerging trajectories means you have to extend both approaches to cover real-time developments as well as longitudinal trends. In the discussion section I defend the value of my mixed-method approach in this study because it provides cross-examination, complementarity, and a reality check for the various methods applied. I also address the criticism that mixing path dependency and path creation perspectives is mixing ontologies and should therefore be avoided.

205

206

Curriculum Vitae Arjan F. Kirkels (1972, Maarheeze) is since 2008 lecturer in the section of Technology, Innovation and Society at Eindhoven University of Technology (TU/e). His areas of professional interest are the development of renewable energy technologies in a socio-economical context, a combination of industrial ecology, feasibility studies and innovation sciences. Since 2008 he started working on a PhD project on the development and trajectories of biomass gasification of which the results are presented in this thesis. After finishing high school in 1991 at Philips van Horne in Weert, Arjan studied Technology and Society at Eindhoven University of Technology. He wrote his thesis on the optimization of electricity production at a municipal waste incinerator (2000). From 1995-2008 he worked as lecturer at TU/e for the center Technology for Sustainable Development (TDO). He has been involved in setting up and teaching of more than 10 courses regarding sustainability and environmental issues, sustainable product design (design for environment), design of sustainable energy systems for the built environment, energy analysis, and industrial ecology – for which he also edited and wrote several readers used at TU/e and HAS Den Bosch. Over the years he supervised over 30 multidisciplinary student groups and many bachelor and master thesis projects, working on topics in the field of technology and sustainability. He participated in networks on sustainability in higher education - on which he also published, he was co-lecturing workshops in Indonesia and was guest lecturer at Fontys and at HOVO - higher education for older people. Publications for this thesis • Kirkels, A. (2012). Discursive shifts in energy from biomass: a 30 year European overview.

Renewable and Sustainable Energy Reviews, 4105-4115. (chapter 2 this thesis) • Kirkels, A., & Verbong, G. (2011). Biomass gasification: still promising? A 30 year global

overview. Renewable and sustainable energy reviews, 471-481. (chapter 3 this thesis) • Kirkels, A. (2014). Punctuated continuity: the technological trajectory of advanced

biomass gasifiers. Energy Policy. (chapter 4 this thesis) • Kirkels, A. (2016). Biomass boom or bubble? A longitudinal study on expectation

dynamics. Technological Forecasting & Social Change, 83-96. (chapter 5 of this thesis)

207

Other publications • Drinkwaard, W., A. Kirkels, H. Romijn (2010). A learning-based approach to

understanding success in rural electrification: Insights from Micro Hydro projects in Bolivia. Energy for sustainable development, 232-237.

• Boer, A., A. Kirkels (2009) Hivos study biomass gasifiers in developing countries. • A.F. Kirkels, A.M.C. Lemmens, F.L.P. Hermans, D.A.A. van Noort, A.H.M. Siepe (2002).

Curriculum Greening at Eindhoven University of Technology. Proc. EMSU. Reprinted in: Teaching Sustainability at universities: towards curriculum greening.

• A. Kirkels (2002) Werktuigbouwkunde en Duurzame Ontwikkeling, duurzame ontwikkeling in opleidingen werktuigbouwkunde, een verkenning. Published in the series ‘Vakreviews Duurzame Ontwikkels’, UCM/KUN, 2002, ISBN 90 77004-09-2

• A. Kirkels (2000). Analyse en optimalisatie energiehuishouding bij AVR-AVIRA afvalverbranding te Duiven. TU/e, Eindhoven, juli 2000 (M Sc thesis).

• D.A.A. van Noort, P.P.A.J. van Schijndel, A.F. Kirkels, J.M.N. van Kasteren, A.M.C. Lemmens (1999). A new multimedia course on sustainable technology: integration with technical curricula. Proc. ENTREE'99, 311-324.

• A. Kirkels, H. Kok, M. Olijdam (1999) Recycle yourself, how to start a recycling project at school. Kok Advice & Research in cooperation with Agro Group, Eindhoven (Netherlands) / Bialystok (Poland).

Textbooks • Reader ‘From industrial ecology to cradle-to-cradle’ • Reader ‘Analysis tools. Design for a sustainable future’ • Reader ‘Energy analysis for industry and the built environment’ • Reader ‘Technology and sustainability’

208

EEiinnddhhoovveenn UUnniivveerrssiittyy ooff TTeecchhnnoollooggyy

SScchhooooll ooff IInnnnoovvaattiioonn SScciieenncceess