Managing long-term environmental aspects of wind turbines: a prospective case study
To be referred to as:
Andersen, Per Dannemand; Borup, M.; Krogh, T.
Managing long-term environmental aspects of wind turbines: A prospective case study.
Int. J. Technol. Policy Manag., Vol. 7, 2007, p. 339-354.
Per Dannemand Andersen1*, Mads Borup1, Thomas Krogh2
Risoe National Laboratory, P.O. Box 49, DK-4000 Roskilde, Denmark
1) System Analysis Department, 2) Wind Energy Department
*) Corresponding author:
ABSTRACT: This paper describes a method for mapping and mitigating the negative environmental impacts of wind turbines and provides an analysis of future removal and recycling processes of offshore wind turbines. The time horizon is up to 2050. The method is process-oriented and interactive with respect to the participation of the actors involved in this area. It recognizes the dynamic, uncertain and rapidly changing character of wind energy and deals systematically with the future removal and recycling of wind turbines and future wind turbine technologies. The method combines life-cycle assessment and technology foresight methods and integrates the perspectives of the present and the future.
Key words:
Prospective methodologies; Life cycle perspective; Technology foresight; Environment, Wind power.
Biographic notes:
Per Dannemand Andersen – M.Sc. (Mech. Eng.),PhD (MoT) - is head of the Technology Scenarios research programme at Risoe National Laboratory’s Systems Analysis Department. He has worked with energy technologies for more than two decades and his main areas of research are R&D strategy, technology foresight, technological innovation and the interaction between industry and science. He has headed and participated in several national and international studies of technology foresight and energy technology studies.
Mads Borup – M.Sc. (Eng.), PhD (STS: social studies of Science, Technology and Society) –is seniorscientist at the Technology Scenarios Research Programme at the Systems Analysis Department, Risoe National Laboratory. He works with science and technology policy, knowledge dynamics & practices, actor networks, technology foresight and strategies for integration of environmental aspects and sustainability in science and technological innovation. He has experience with areas as energy technology, nanotechnology, advanced communication technologies and environmental technology.
Thomas Krogh - B.Sc. (Mech. Eng.) – is a research engineer at the Wind Energy Department at Risø National Laboratory. He has been working with various fields in wind energy since 1998 – loads, mechanical and structural design of wind turbines in connection with research projects, certification and consultancy services.
Managing long-term environmental aspects of wind turbines: a prospective case study
1. INTRODUCTION
Wind turbines are one of the most environmentally sound technologies for producing electricity. However, the removal phase of the life-cycle of wind turbines has been identified as a blind spot in analyses of the environmental impacts of wind-power systems. Most previous impact analyses have focused primarily on the operational phase of the wind-turbine life-cycle and in some cases also on the production and mounting phases. Because the wind-turbine industry is relatively young, there is only a limited amount of practical experience of the removal and recycling of wind turbines, particularly in respect of offshore wind turbines, a relatively recent phenomenon. It is likely to take more than twenty years before a substantial amount of practical experience regarding the dismantling, separation, recycling, disposal etc. of wind-power systems is gained. Nonetheless major steps in the definition and planning of these processes in the removal phase have been taken implicitly in recent years. The design of offshore wind farms and turbines, as well as regulatory demands and organizational responsibilities, are to a considerable extent determining the structure and processes of the removal phases many years ahead. A systematic combination of both present and future perspectives that reflects the rapid development of wind power and the continuously changing character of wind-turbine technologies will be needed here to address this question.
The study being reported here focuses on the removal phase in respect of novel aspects connected with the offshore localization of wind turbines. The study has developed an interactive and process-oriented method for investigation of the environmental impacts of wind turbines and for finding ways to reduce negative impacts. It tries to establish an exchange of knowledge between the removal phases and the design phases of new wind turbines.
The approach taken in the study was to integrate the life-cycle assessment method (LCA) with the method of technology foresight (TF). The approach was process-oriented and to a considerable degree analysed the potential for reducing the hazardous environmental impacts of wind turbines questioning terms of establishing an interaction between the relevant actors of the area.
Offshore wind farms are planned in many countries in northern Europe around the North Sea and in the Baltic Sea. Important aspects of developments in wind-power systems and technologies are of an international character, and it is our understanding that the method and conclusions of the research project are of general relevance. However, it should be noted that the research project has been carried out primarily in the Danish context, including the understandings of offshore wind turbines, developments in wind power, and the organizational and regulatory practices that apply here.
The paper starts with a short introduction to the approach taken in the study, followed by the results and considerations in each phase of the study. Finally, the paper offers conclusions regarding the lessons learned and the future handling of the environmental aspects of wind turbines.
2. THE INTERACTIVE APPROACH
Over the last ten to twenty years, the method of life-cycle assessment (LCA) has become established as a way of analysing the environmental impacts of particular products (for an introduction, see Weidema, 1997).
Traditionally the core of LCA has been a detailed analysis of the processes of the entire life-cycle of the product, from the early phases of the design process to the demolition and recycling of the product materials after end use. It is assumed that the product is known and fully specified. In principle, all materials, resource consumptions, discharges and environmental impacts, including all sub-processes, should be assessed. In practice, there is always a limit on how much information is included and a tendency to focus on some parts and processes more than others.
These limitations can become a problem if the LCA is narrowly considered a question of obtaining the results and figures that come from the process. If instead the LCA can be considered a process of learning in which the people and organizations involved in the process acquire knowledge about the product and its environmental impacts, the limitations will often appear less severe. Often one of the important outcomes of a LCA is the knowledge the actors obtain about which parts of the life-cycle process are uncertain, and which parts there is no precise information about. This point is supported by the observation that it is usually very difficult to understand the results of LCAs carried out by others. To make proper sense of the results, a detailed understanding of the background to the figures is needed.
These considerations are, of course, widely acknowledged by the LCA community, and over the years, prospective elements and learning/decision elements have increasingly been introduced into LCA (Christiansen, Horup, Jensen 2001; Pesonen et al. 2000). The literature therefore now deals with two approaches to LCA, namely retrospective and prospective LCA. The two approaches have been defined as follows: “Retrospective LCA is defined by its focus on describing the environmentally relevant physical flows to and from a life cycle and its subsystems. Prospective LCA is defined by its aim to describe how environmentally relevant flows will change in response to possible decisions” (Ekvall, Tillman & Molander, 2005).
A number of recent studies and reviews have been carried out on methodologies and approaches for future-oriented analyses and policy-making in science and technology (Kuhlmann et al., 1999; Holtmannspötter & Zweck, 2001, Porter et al, 2004). Such reviews of methodologies for policy-planning in science and technology often discuss three distinct approaches: science and technology foresight, technology forecasting, technology assessment, and science and technology policy evaluation.
Definitions of science and technology foresight can be found in many references, but there is a general consensus that foresight is concerned with the impact of technological developments on society, the focus being on the identification of broad future trends and the socio-economical aspects of emerging technologies. A broad cross-societal dialogue is the most central trait of foresight exercises.
Technology forecasting is also concerned with emerging technologies and their implications, but compared with science and technology foresight, it involves less dialogue between the various stakeholders, if any. Where foresight is often linked with public policy forecasting, it is often associated with business strategy..Technology forecasting focuses primarily on technological and economical aspects, and techno-economical development is a central issue in the prospective use of LCA.
Technology assessment, like science and technology foresight, deals with the impact of new and emerging technologies on society, and cross-societal dialogue is essential. Since technology assessment tends to focus on the risks of technologies and their secondary implications for society, the examination of norms and values is important.
Technology foresight and future-oriented strategic analyses apply a variety of standard research methodologies and approaches that have been adapted from standard social-science research methodologies. One group of tools is labelled trend analysis tools and includes methods such as extrapolations of time series, S-curve analysis, analogies, experience curves etc. Such trend analysis tools are often less applicable for prospective analysis where there are high uncertainty, long-time horizons and changes in technology or market situation. We are then left with what is often referred to as judgemental methodologies, such as focus groups or interviews with experts, Delphi surveys, scenario writing, etc. (Jantsch, 1967; Millet and Horton, 1991; Porter et al. 2004). From the above considerations, it can be seen that LCA and TF complement each other, suggesting that an integration of the two approaches will prove fruitful in dealing with the environmental aspects of rapidly changing areas of technology.
The project was carried out in six stages. Stage 1 comprised an analysis of the environmental effects of state-of-the-art wind turbines through an LCA with a focus on manufacturing and decommissioning. This stage relied on other studies of the environmental effects of wind energy. In stage 2 a mapping of current trends in wind power technologies and concepts was carried out, the aim being to obtain an overview of trends affecting how wind turbines will be designed in the future. Stage 3 took the form of an expert panel brainstorm on drivers for the development of wind-turbine technology, the focus being on factors that could not easily be extrapolated. Stage 4 consisted of a Delphi survey on the future of wind-energy technology, asking questions based on the first three stages. In stage 5 the characteristics of future wind-turbine technology (2020-2030) were drawn up. Finally, a panel with experts on decommissioning, recycling, waste handling, etc., was used as brainstorming session on present and future (2050) environmental aspects after the machinery is decommissioned.
3. LIFE-CYCLE ASSESSMENTS OF OFFSHORE WIND TURBINES
Several projects have shown that wind turbines are an environmentally sound technology for producing electricity. Wind energy has very low environmental impacts (Kummel et al, 1997; ELFOR, 2000; Tryfonidou and Wagner, 2004). Other studies have carried out LCAs on offshore wind farms (Hassing, 2001; Properzi and Herk-Hansen, 2002). These studies conclude that the greatest environmental impacts come from three main sources:
- bulk waste from the tower and foundations (e.g. from steel production), even though a high percentage of the steel is recycled
- hazardous waste from components in nacelle (e.g. from steel alloys)
- greenhouse effects (e.g. from steel production and surface treatment)
These results indicate that further analysis should take into account changes in the materials used in the tower and the foundations, as well as changes to nacelle components (changes in overall concept, use of gearless designs, use of power electronics, etc). In comparing offshore and onshore wind farms, the role of the cables becomes central (Hassing, 2002). Many kilometres of sea cables are used, which involve heavy and complex constructions that can resist the harsh environment of the sea. There is a considerable environmental impact from the production of the cables, which, since they consist of many different materials, are difficult to handle in the dismantling phase.
There are some uncertainties and limitations regarding LCA studies of offshore wind turbines: in particular, the specific processes of, for example, the dismantling and recycling processes, which are crucial for the environmental profile, are not known. Many material recovery processes are not included in the analyses. Another example of the limitations involved is that the area use and the impacts on the landscape, which are generally considered critical environmental impacts elsewhere, are not included. In general, it is the experience of these LCAs that it is difficult to obtain data for the analysis, but also and difficult to delimit the data collection itself. The conclusion therefore is that iterative data collection processes should be used. Through the learning process of the iterative data collection, where the analysis should be directed is defined, and more reliable data and knowledge about the uncertainties of the analysis and of the wind turbines are produced. It is argued that close co-operation with the wind-turbine producer is necessary.
4. ANALYSIS OF TRENDS IN WIND TURBINES
Key information for a life-cycle analysis is related to the types and masses of materials used in the technology or product in question. The amount of material to be recycled in the future relates, of course, to the expected future installation of wind turbines over the years to come. Several organisations have carried out studies of such scenarios, two types of which can be drawn up. The extrapolative or predictive type of scenario draws on elements from forecast techniques. Normative scenarios often reflect more radical discontinuities and may be combinations of technological possibilities and political ambitions or targets. With regard to descriptive scenarios, the most comprehensive study is the annual World Market Update and Forecast from the consultancy BTM Consult (BTM Consult, 2005). These forecasts are based on manufacturers’ information and order stocks, existing national policies, programmes and targets, etc.Other descriptive scenarios have been published by the International Energy Agency (IEA, 2002),who as late as 2001 estimated the cumulated global installations of wind turbines to be 34 GW by 2010 and 67 GW by 2020. As actual installations by the end of 2004 accounted for 48 GW, the IEA projections seem to be quite conservative, even unreliable. The World Energy Council has formulated two comprehensive predictive scenarios for world energy supplies in 2020: a Current Policy Scenario, and an Ecologically Driven Scenario (WEC, 1993). These scenarios also contain expectations concerning the cumulative installation of wind power. With regard to normative scenarios, several organisations and consultancies have formulated forecasts for the future utilisation of wind power. One such normative scenario has been formulated by two non-governmental organisations, the European Wind Energy Association and Greenpeace (EWEA/Greenpeace, 2003), whose study analysed the potential for providing world electricity systems with 12% of electricity supplied from wind-power plants. In this scenario, it is suggested that wind power will expand to 234 GW by 2010, 1260 GW by 2020 and to 2517 GW by 2030.
2004 / 2010 / 2020 / 2030World cumulated installations / Realised
BTM: 48 GW / IEA: 34 GW
BTM: 137 GW
WF12: 234 GW / IEA: 67 GW
WEC(cp): 180 GW
WEC(ed): 474 GW
BTM: 539 GW
WF12: 1260 GW / WF12: 2571GW
Table 1. Scenarios and projections for future cumulated wind turbine installations globally. Sources: IEA: World Energy Outlook (2002), WF12: WindForce 12 (2001), E21: BTM: BTM Consult (2005). WEC (cp): WEC (1993) current policy scenario, WEC (ed.): WEC (1993) ecologically driven scenarios.
The average size of commercially sold wind turbines has increased dramatically since the beginning of modern wind-turbine utilisations in the early 1980s. By this time, a typical wind turbine had a rated power of 55 kW. In 2004 the largest commercially available machines are in the size range of 3–4.5 MW, almost a hundred times larger than twenty years ago. If this development is extrapolated into the future, new machines are expected to be in the size range of 10 MW by 2010. If this historical trend is extrapolated further until 2030, 100 MW wind turbines would appear to be the norm. However, design experts expect technical development to stop at about 10 MW. Therefore, the study settled on 10 MW as the typical size of wind turbines in the timeframe of 2020-2030.
Similar track-record data are available for the development of materials and main components. Extrapolated linear (or other) estimates for future wind turbines can then be made. Figure 2 gives the example of masses of blades.
Figure 2. Extrapolation of the relationship between blade mass (in kg) and blade length (in metres). Based on historical data for the two wind-turbine blade manufacturers, Vestas Wind Systems A/S and LM Glasfiber A/S. Extrapolation using LM data.
Based on such extrapolations of current trends, a sketch of what a wind turbine would look like in 2020-2030 can be drawn up, as in Table 2.
Overall size / Rated capacityRotor diameter / 10 MW
160 m
Blades / Blade length
Blade weight / 75 m
20 t
Nacelle and hub / Total mass
- Gearbox mass
- Generator mass
- Hub mass / 400 t
- 85 t
- 45 t
- 75 t
Tower / Tower height
Tower mass / 130 m
800 t
Table 2. Estimated key figures for a 10-MW wind turbine by 2020-2030 based on extrapolations of current trends.