The challenge of decommissioning offshore wind farms: how operators learn within the innovation system Dual Master’s Thesis

(1)

The challenge of decommissioning offshore wind farms: how

operators learn within the innovation system

Dual Master’s Thesis

Submitted to:

Dr. Jasper Veldman – University of Groningen

Dr. Jonathan Sapsed – Newcastle University Business School

Submitted by:

Matthias Lackinger

Operations Management (Dual Award) Student ID (Groningen): S4243994 Student ID (Newcastle): 190520739

(2)

II

Acknowledgment

My sincere thanks go to my first supervisor Jasper Veldman who supported me from the very beginning of this thesis with his expertise. He supported me when I had difficulties finding industry experts and challenged me with his constructive criticism. I am grateful for his feedback which was challenging and motivating. Jasper was always willing to answer my questions. In short, I am grateful for his extraordinary support.

I also thank Jonathan Sapsed for his support as my second supervisor. Jonathan’s comments served as a very good inspiration and he helped me explore further research areas.

I thank all my interview partners for taking the time to talk to me and giving me great insights into their work. This case study research would not have been possible without their cooperation and I now know how much a student can appreciate that.

Finally, I would like to thank all the lecturers and fellow students I have had the opportunity to meet over the past months. The program at Newcastle University and the University of Groningen was the right choice for me and I am grateful for the support I received along the way.

(3)

III

Abstract

(4)

IV

List of Figures

Figure 1: End-of-life options ... 5

Figure 2: Decommissioning is embedded within the sectoral system of innovation ... 11

Figure 3: Conceptual framework ... 17

List of Tables

Table 1: Interview overview ... 19

Table 2: Case companies overview ... 21

Table 3: Reliability and validity in the five case studies ... 22

Table 4: Cross-case comparison ... 38

(7)

1

1. Introduction

In recent years, thousands of offshore wind turbines have been built worldwide. So far, both governments and firms are mainly concentrating on how the expansion of offshore wind power can be fostered. However, with an expected physical lifespan of about 20-25 years (Davidsson et al., 2012; Lantz et al., 2013; Jensen, 2019) offshore wind farm decommissioning will play an essential role in the sector soon. An offshore wind farm decommissioning project is a costly, complex, and time-consuming activity. For example, to decommission the 160 turbines Gwynt y Môr offshore wind farm on the North Wales coast, the operator expects a 2-year execution period with estimated decommissioning costs of £ 400.000 per turbine (Drew, 2011). These estimations were made before the wind farm was even built, and only a few, much smaller nearshore wind farms have been decommissioned so far. In practice, there does not seem to be much consensus on the timing and execution of the decommissioning task, and best practices have yet to be developed as knowledge in this area is still scarce. However, knowledge, including practices, standardized procedures, tools, and technologies is necessary to make the dismantling of thousands of turbines cost-effective and environmentally sound. It is not known how actors prepare and learn about decommissioning although many challenges are known (e.g., Topham et al., 2019a). Clearly, innovation is needed, which does not come from one actor alone in a system where different actors, such as governmental agencies, university programs, or companies contribute to knowledge development. This explorative multiple case study focuses on such a ‘sectoral system of innovation’.

(8)

2 (Malhotra et al., 2019, p. 472). In addition, various innovation systems have already been analyzed in the literature (Frishammar et al., 2018) and policy recommendations for promoting learning activities have been proposed (e.g., Wieczorek and Hekkert, 2012). The contribution of this research work is the analysis of the innovation task of decommissioning, which is a liability for the asset owner, in combination with the used learning forms to address the issue within an innovation system. Thus, this thesis contributes to the existing literature about innovation systems and organizational learning. Furthermore, the multiple case study extends the existing literature on offshore wind energy by identifying important issues that need to be addressed concerning the decommissioning problem.

This research work demonstrates different approaches within the sectoral innovation system. Although cooperation in the offshore wind energy sector works well in principle, the long timeframe between the construction and decommissioning phase causes a split. On the one hand, offshore wind farm operators plan to think about the issue in detail only when the need arises, i.e., when end-of-life options are pending. On the other hand, other actors such as recycling firms or construction companies, are actively promoting innovation, as decommissioning will be a business opportunity for them. It was found in the various case studies that offshore wind farm operators are mainly gaining important knowledge by using their current turbines and by interacting with different actors of the innovation system. The third predominant learning form is learning by knowledge spillovers which seems to be of particular importance in the context of decommissioning as wind farm operators with relatively new offshore wind farms plan to benefit from the experience of earlier decommissioned wind farms of competitors. Moreover, offshore wind farm operators are hesitant to preemptively address the issue because there is still uncertainty about the timing and regulatory requirements. In this respect, it is interesting to note that decommissioning is rarely incorporated in the design phase of an offshore wind farm project.

(9)

(10)

4

2. Theoretical background

In recent years, a large number of papers about offshore wind energy have been published. However, particularly in the field of offshore wind farm decommissioning, there is still little literature available. The following section describes the current state of knowledge about offshore wind farm decommissioning and presents the theoretical background of learning within a sectoral system of innovation.

2.1. Offshore wind farm decommissioning

Few research works focus on the end-of-life phase when an asset needs to be decommissioned (Invernizzi et al., 2019) and particularly in the field of offshore wind farm decommissioning, there is still little literature available. However, offshore wind farm operators can benefit from a well-planned and well-executed decommissioning project because this can manage end-of-life costs, which can be significant. Learning about decommissioning can therefore reduce costs for the responsible offshore wind farm operators since uncertainties decrease as firms know more about what to expect at the end of a wind farm’s life.

In the end-of-life phase, the offshore wind farm operators have three decision options (Figure 1): (1) life-cycle extension (i.e., the continued operation of the turbines beyond the actual planning date), (2) repowering (i.e., continued operation after replacing individual components), or (3) decommissioning (i.e., complete dismantling and end of the operation). Especially with the last two options, which require at least a partial deconstruction, many other decisions have to be made (Figure 1). For example, the number and type of chartered vessels affect the costs and required time for decommissioning. Several failure modes (e.g., cracks in the blades, or corrosion in the tower) influence the end-of-life time and the basis for decision-making (Luengo and Kolios, 2015).

(11)

5 McCulloch, 1996) decommissioning and related costs are often seen as a necessary evil because firms have no economic incentive to remove their offshore assets. Invernizzi et al. (2019) discuss how decommissioning of nuclear power plants can create value for the responsible firm. Although they conclude from their interview study that value management “has potential to improve the performance of decommissioning projects” (Invernizzi et al., 2019, p. 680), the interviewed experts had discordant opinions about what “value” even means in the context of a decommissioning project. Value for the firm which has the decommissioning obligation could mean carrying out the decommissioning project as efficiently as possible in terms of finance or time, but social aspects or environmental considerations can also add value.

Furthermore, offshore decommissioning is a challenging task. Decommissioning projects in the offshore oil and gas industry are growing rapidly in the UK, the Netherlands, and Norway, for example (OGUK, 2019). Although the offshore wind sector can adopt many of the oil and gas techniques while operating under the challenging offshore site conditions (Smith and Lamont, 2017), operational processes common to oil and gas asset decommissioning cannot be readily transferred to the wind energy sector because of the vastly different requirements. In contrast to oil and gas where often one heavy rig needs to be decommissioned, a wind farm consists of dozens of identical entities spread over several square kilometers, which makes logistics and vessel utility more important. Nevertheless, multiple factors are less complex for offshore wind farm decommissioning. The risk of pollution of the marine

Figure 1: End-of-life options

1. Life-cycle extension 2. Repowering 3. Decommissioning

▪ Timing (with the end of the operating license or already earlier?)

▪ Scope of action (What needs to be replaced or what needs to be done for decommissioning?)

▪ Use of materials (re-use, recycling, disposal)

▪ Technical procedures (different options in terms of costs, environment, resources)

▪ Inhouse or outsourced

(12)

6 environment is low in contrast to the oil and gas industry because fewer chemicals or hydrocarbons can leak out. Moreover, most wind farms are built relatively close to shore and not over such a great depth of water as most oil rigs (Smith and Lamont, 2017). However, it must be noted that this will also change in the near future, as many floating offshore wind projects are underway to enable the operation in deep waters (WindEurope, 2018). Whether this type of construction is easier to decommission than the most widespread monopiles is not yet clear and depends mainly on the mooring on the seabed and on how much dismantling work has to be done offshore.

Some wind farm operators are also active in the offshore oil and gas sector and may use parts of their decommissioning knowledge for their wind energy business. However, many wind farm operators have no experience at all. Although there is more experience with decommissioning onshore wind turbines, there are still open questions, related to, for example, optimal removal and recycling methods (Jensen, 2019; Topham et al., 2019b). Nevertheless, offshore wind farm decommissioning cannot be equated with onshore wind turbine decommissioning as conditions at sea are very different (Ortegon et al., 2013). The offshore environment, such as the dependence on weather conditions or consideration of marine life makes offshore operations more complex in general. The power supply connection infrastructure or the required vessel resources further complicate the offshore decommissioning task. Topham et al. (2019a) describe four key challenges regarding the decommissioning of offshore wind farms: the unspecific regulatory framework, the right decommissioning procedures, logistic issues (especially vessels’ availability), and environmental impacts. Their findings, however, are based on the limited academic literature and press articles so that this interview study offers direct insights into the sectoral innovation system.

(13)

7 distance to shore, for example. In addition, the various actors in the sectoral innovation system must cooperate and learn together because the decommissioning task is not solved by one actor alone. Wind farm operators must communicate their plans to seaports, or blade manufacturers must pass on the composition of the materials used to recycling companies.

Stentoft et al. (2016) state that innovation and collaboration in the offshore wind energy supply chain can reduce the cost of energy of offshore wind. The authors also conclude from the academic literature that there is a “lack of collaboration across the [offshore wind energy] supply chain” (Stentoft et al., 2016, p. 152). However, their work is purely theoretical and although decommissioning is included in their framework, the offshore wind innovation system is not studied with a qualitative or quantitative approach. Thus, it remains unknown what the reasons for a ‘lack of collaboration’ are.

(14)

end-of-8 life moment is known with limited accuracy (Topham and McMillan, 2017). The uncertainty and a lack of knowledge make it essential that the offshore wind farm operators learn about possible solutions in order to increase planning reliability and the success of the decommissioning project.

Vasconcelos Gomes et al. (2018) analyze in their case study how entrepreneurs manage collective uncertainties in an innovation ecosystem where uncertainty affects not only the innovator but the whole innovation ecosystem.1 Although firms in their study “initially tended to focus on […] individual uncertainties” (Vasconcelos Gomes et al., 2018, p. 177) before dealing with issues that were relevant for multiple actors, they show that entrepreneurs manage collective uncertainties, for example, through common learning experiments (e.g. workshops) to mitigate uncertainty. In the research work of Perminova et al. (2008) regarding the phenomenon of uncertainty in projects, the authors state that reflective learning and sensemaking are “key elements in managing uncertainty” (Perminova et al., 2008, p. 78) because they can help the firm to react flexibly to changes. The more the offshore wind farm operators learn about decommissioning, the broader the knowledge base to counter uncertainties will be. Other actors in the innovation system, such as recycling firms or logistics providers are also interested to acquire knowledge in order to offer competitive decommissioning services in the future. Finally, it is in the public interest to make offshore energy as affordable as possible without neglecting the environmental and social aspects of decommissioning. Low decommissioning costs may lower electricity prices for the public but also increase the attractiveness for investors to build new offshore wind farms.

It is therefore of interest how offshore wind farm operators deal with decommissioning within an uncertain environment where the task at hand is seen as a necessary evil. These two factors characterize the problem of decommissioning within the sectoral system of innovation (Figure 3).

1_{Some authors employ the concept of innovation ecosystem in order to analyze the firms’ learning and}

(15)

9

2.2. Offshore wind energy as a sectoral system of innovation

Organizational learning usually does not take place in isolation. An innovation system is comprised of many actors. In a decommissioning project, not just the offshore wind farm operators, but also recycling companies, network operators, or construction companies, are engaged. Hence, in order to understand the phenomenon of how offshore wind farm operators learn about decommissioning, the relationship to other actors is relevant. The concept of sectoral systems of innovation and production from Malerba (2002) is useful for such a descriptive analysis. It provides a useful tool to better understand the interaction between the various actors in this industry and the factors that influence the innovation and learning process.

According to the sectoral system of innovation framework of Malerba (2004), innovation is affected by knowledge and technologies, actors and networks, and institutions. These three main ‘building blocks’ form the sectoral offshore wind energy system (Figure 3).

(16)

10 national or regional perspective is not appropriate for the offshore wind energy sector. A technological innovation system is similar to a sectoral innovation system and could also describe the offshore wind energy sector as it is defined as a “network of agents interacting in the economic/industrial area under a particular institutional infrastructure and involved in the generation, diffusion, and utilization of technology” (Carlsson and Stankiewicz, 1991, p. 94). However, a technological system sets the focus on the development of specific technologies which is why Malerba’s sectoral system of innovation concept is best suited for this research work, as it can also be interpreted more broadly without being limited to technology. Generally, the different concepts are not mutually exclusive, and the sectoral system can be a global system at the same time, for example.

The actors in a sectoral system of innovation are firms (e.g., wind farm operators or suppliers) and non-firm organizations (e.g., research universities, government laboratories, or government agencies). Groups of organizations, such as the European wind industry association ‘WindEurope’, are also actors that directly facilitate a network between different agents. Some of these actors, such as recycling firms, can directly profit from the upcoming decommissioning challenge and may actively push the learning so that they are gaining new profits. Others, especially non-firm organizations, may have no economic incentive to learn but want to foster the knowledge creation and exchange among actors so that the decommissioning issue gets clearer. One example is publicly funded projects, such as DecomTools and SeeOff, which have the objective to develop efficient strategies for offshore wind farm decommissioning.

(17)

11 institutions, such as regulations, form the ‘rules of the game’ for the sectoral innovation system, they can reduce uncertainty for involved actors (Woolthuis et al., 2005).

Knowledge and technologies form the basis of innovative activity in the sectoral system. Moreover, the existing knowledge from the oil and gas industry, where decommissioning projects have been carried out for decades and many will follow in the coming years, may help the wind energy sector to find answers to the challenge of decommissioning offshore wind farms. Thus, the decommissioning knowledge is cumulative in that way that the sectoral system can build upon existing processes and technologies and there is no need to learn from scratch. This research work also investigates which learning processes wind farm operators use (Figure 3) to build up their knowledge base.

Figure 2: Decommissioning is embedded within the sectoral system of innovation

(18)

12 There is a broad literature studying innovation systems’ weaknesses (Frishammar et al., 2018) and the described character of decommissioning could make the search for innovative solutions more difficult.

In general, the boundaries of a sectoral system are not fixed (Malerba, 2002) and different problems could be embedded in the center of Figure 3. Moreover, as illustrated in Figure 2 the sectoral system of offshore wind energy has also overlapping boundaries with other sectoral systems (e.g., with the offshore oil and gas sector) and several problems or subsystems can be part of a sectoral system.

2.3. Learning in the sectoral system of innovation

This study posits, as mentioned, that learning is essential to handle the decommissioning challenge. Wind farm operators should try to learn as much as necessary about decommissioning options. Also, the uncertainties can become “more known” as additional information is collected on the project over time (Ramasesh and Browning, 2014). There may be uncertainties that are not even known yet by the operator. The literature uses the term ‘unknown unknowns’ or ‘unk unks’ to describe these uncertainties (Ramasesh and Browning, 2014). The company is not aware of these uncertainties and therefore has difficulty in preparing for them. However, the authors argue that it is helpful if the company learns about the knowable unk unks and thus transfers them into known unknowns in order to prepare for their occurrence. Related to that concept unforeseeable uncertainties are “events that cannot possibly be foreseen at the outset” (Sommer et al., 2009, p. 129).

The following section refers to relevant learning forms and explains them with the corresponding literature, which then leads to the conceptual framework.

(19)

13

2.3.1. Learning by doing

In one of the original works about learning by doing, Arrow (1962) describes this learning process by conducting repetitive work, for example, in a production process. His seminal work led to broad literature about learning curves. Rosenberg (1982), for example, shows that due to learning by doing, engine maintenance costs decrease for a firm with the years the engines are in service. The renewable energy literature also discusses learning by doing and how firms benefit from it. Qiu and Anadon (2012), for example, build up a learning curve model for the wind industry in general and show that the learning processes can lead to a price reduction for the produced wind power. Part of the learning by doing process is also a trial-and-error practice where firms learn from their mistakes. By trial-and-error learning where a firm actively searches for new information, the original “goals and course of action” (Sommer et al., 2009, p. 118) can change. In this way, actual repowering or decommissioning plans can be reversed if new findings occur, the environment changes, or failures become apparent.

However, with a project scope, such as the decommissioning of offshore wind farms, it is questionable if firms can afford to follow only such a practice. For decommissioning projects, the knowledge acquired should be sufficient before the “doing” begins. Nevertheless, employees can build up their experience from learning by doing so that this learning form generates mainly tacit knowledge (Kamp, 2004).

2.3.2. Learning by using

(20)

14

2.3.3. Learning by searching

Learning by searching is linked to the R&D of a firm and closely linked to an explorative learning style. Some authors use the term learning by learning (e.g., Sagar and Van der Zwaan, 2006). Although Malerba (1992) states that learning by searching takes place within the firm, a certain interaction between different agents may take place here as well. For example, the R&D activities of firms often involve universities or research centers. Usually, firms do not innovate alone. Simulation studies about different decommissioning options or specialized engineering teams are also part of this learning form, for example. Firm internal search activities (i.e., exploration) require more resources than using external knowledge spillovers or acquisitions (Cassiman and Veugelers, 2006).

2.3.4. Learning by interaction

Sources of innovation do not come only from the firm itself but are found in the whole innovation system a firm is part of, for example, from universities, suppliers, or customers (Powell, 1990). Powell et al. (1996) found that especially in environments where the expertise is widely dispersed among different sources, learning happens within a network of inter-organizational relationships and not within the firm.

(21)

15 might be even more relevant to support innovative activities when it comes to decommissioning as offshore wind farm operators may have few incentives to acquire knowledge because decommissioning is not their core business.

Wind farm operators that have already carried out decommissioning projects can share their experience with other wind farm operators. However, inexperienced operators can also share their knowledge about decommissioning schedules, the use of vessels, or other resources. With the knowledge available in multiple companies, decommissioning plans can be adapted to the needs of the wind farm operator over time. Learning by interacting with other firms can happen through informal knowledge sharing (von Hippel, 1987) or through official agreements (Hagedoorn and Schakenraad, 1990). A practical example of a formal learning collaboration in the wind energy sector is the wind farm Alpha Ventus. It is operated by a joint venture of three companies and research is conducted with various actors from research institutions and private industry. In general, joint operations between competitors are not unusual in the offshore wind energy sector. However, the questions arise if firms in such a coopetitive relationship learn together or on their own and who is responsible for decommissioning coordination at the end of the asset’s life. This form of horizontal collaboration differs also from vertical collaboration between the operator and suppliers, which is more common in every industry.

However, learning by interaction does not only take place between organizations. Several research works show that cross-individual and cross-group interactions drive learning and can, therefore, increase productivity and planning (e.g., Boh et al., 2007; Narayanan et al., 2009).

2.3.5. Learning by knowledge spillovers

(22)

16 The processes for offshore wind farm decommissioning will not differ significantly between different wind farms. The pioneers with the oldest wind farms must first decommission their assets, especially because these turbines have much smaller capacities which makes repowering with bigger turbines impossible. Therefore, these players must first develop efficient strategies. This could lead to a passive and reactive learning behavior of recent incumbents, as they could hope to benefit from the first experiences of others.

2.4. Conceptual framework

The following subsection outlines the conceptual framework (Figure 3), which gives an overview of the basis of the case studies and helps to present the underlying research questions.

Knowledge is a key resource for innovating and learning a crucial activity to gain knowledge (Lundvall, 2010). Most of the described learning forms of subsection 2.3 are actively decided and implemented. These learning processes differ “from simple automatic learning which is a by-product of doing” (Malerba, 1992, p. 845). Thus, learning by doing does always exists if there is economic activity, i.e. as long as the operator is active in the offshore wind energy sector. However, it is unclear if the firm already collects and uses the knowledge derived from learning by doing for decommissioning considerations. Furthermore, it is of interest which additional learning forms a firm chooses to acquire knowledge about wind farm decommissioning especially before the first decommissioning project takes place.

(23)

17 processes if it is uncertain whether the same subsidies or electricity prices will be paid in 20 years.

Figure 3: Conceptual framework

Are there differences in the learning forms used between offshore wind farm operators? Or do all firms have a similar attitude towards decommissioning and use the same learning strategies? The interview-based research work aims to identify the learning activities used in the industry as well as the incentives to address the problem.

The overall research question of this thesis is: How do offshore wind farm operators learn about offshore wind farm decommissioning?

(24)

18

3. Methodology

In order to learn about the wind farm operators’ attitudes, incentives, and behavior towards decommissioning and the learning forms they use, an exploratory multiple case study was conducted because this phenomenon has been poorly explored. In addition, uncertainties and problems of decommissioning are worked out by conducting interviews with different actors of the sectoral innovation system. Therefore, it is beneficial that a case study helps to comprehend the phenomena under exploration “in its natural setting” (Benbasat et al., 1987, p. 370). For an exploratory study, case study research is rational (Yin, 2014) and knowledge of the whole offshore decommissioning context can be gathered, too.

Various measures were taken to ensure the validity and reliability of the study (Table 3). The measures and methods are explained in the following two subsections which introduce the data collection approach and data analysis.

3.1. Data collection

The data for the qualitative research design was mainly gathered by semi-structured expert interviews with various actors in the offshore wind energy sector. It also secured an in-depth understanding of the innovation system itself. The focus, however, was on wind farm operators as responsible actors for decommissioning.

A selection criterion for offshore wind farm operators was that they must already have a commissioned offshore wind farm that is actively managed. Investment management corporations or banks, for example, that are only shareholders of a wind farm were excluded because they do not plan for decommissioning themselves. This supports the validity of the different cases because each operator has the decommissioning liability and faces the same challenges. Moreover, by interviewing multiple firms, possible comparisons can be identified. This can help to identify possible patterns for all cases and not just a finding of one specific case (Eisenhardt, 1991).

(25)

19 working at wind farm operators were contacted by phone before the actual interview took place to ask first questions and to see if the company is suitable to participate in the actual case study. Additionally, general experts of the sectoral innovation system were identified via online research and the method of snowball sampling (Biernacki and Waldorf, 1981) because each interviewee partner was asked to recommend colleagues or other experts of the innovation system at the end of the interview.

Table 1: Interview overview

Actor Role Number of

interviews

OWF (Offshore wind farm) operator Alpha

OWF owner; responsible for decommissioning. 1

OWF operator Beta OWF owner; responsible for decommissioning. 3

OWF operator Gamma

OWF owner; responsible for decommissioning. 1

OWF operator Delta OWF owner; responsible for decommissioning. 1

OWF operator Epsilon OWF owner; responsible for decommissioning. 2

Offshore wind association

Lobbying activities for OWF operators 2

Recycling firm A Analysis of technical methods for disposal and recycling of wind turbine blades

1

Recycling firm B Recycling and disposal of wind turbine blades 1

Sea port Provision of logistics infrastructure and supply chain coordination for all life cycle phases.

(26)

20 Decommissioning

research project A (DRP A)

Researcher for publicly funded research project for OWF decommissioning

2

Decommissioning research project B (DRP B)

Coordinator for publicly funded research project for OWF decommissioning

1

Offshore research foundation

Coordination and financing of research projects on offshore wind energy

1

Offshore wind consultant

Consultancy projects for multiple OWF operators in the area of construction and operations

1

Service provider Offering all maintenance services for OWF 1

Network operator Provision and operation of offshore transformer stations.

1

(27)

21

Table 2: Case companies overview

Case Alpha Beta Gamma Delta Epsilon

Number of employees in the offshore wind energy department >200 <200 >200 <200 >200 Number of offshore wind farms 3 3 4 2 3 3.2. Data analysis

To increase the validity of the research, firm documents, newspaper articles, annual reports, and the website of the wind farm operator were studied. External validity was secured by comparing the results to the offshore wind energy and innovation systems literature. The case study protocol in Appendix B shows the structure and included questions of the semi-structured interview procedure which also ensures the construct reliability of the study. The interview protocol is structured in three sections that ensured all relevant questions were addressed. First, general information about the company and the interviewees’ experience was gathered. Second, the general strategy of the firm regarding decommissioning plans was discussed. Information about involved stakeholders, the planned timing, and the level of knowledge was gathered, too. The third interview section focused on the learning processes and the factors that have an impact on the firm’s behavior. The inductive methodology allows one to modify the specific questions (Appendix B) and to ask new upcoming questions according to the interviewee’s experience and answers. Thus, there was no fixed wording for the questions, which were also adapted according to the company’s background.

(28)

22 consisted of three major steps. First, the data were analyzed for a within-case analysis to understand each of the five firms’ decommissioning approaches in depth. Next, in the following cross-case analysis the different statements were compared to find differences or common patterns across the five cases. Last, the interviews of the other actors of the sectoral system were analyzed and incorporated into the cross-case analysis and results.

Table 3: Reliability and validity in the five case studies

Reliability/Validity criterion

Implementation

Reliability Use of a case study protocol (Appendix B); interviews with offshore wind farm operators were transcribed and coded in NVivo 12 Pro.

Internal validity Description of the general research objective and use of a conceptual research framework; offshore wind energy experts were interviewed.

Construct validity Use of multiple sources of information such as semi-structured interviews, reports, and firm documents; anonymity was assured to interviewees and companies; data analysis parallel to the interview phase to incorporate new results and consult other actors of the innovation system on these new results.

(29)

23

4. Within-case analysis

This section discusses the results and describes the learning activities and incentives to innovate for each of the five offshore wind farm operators. A number placed after the firm pseudonym, such as Alpha1, refers to a particular interviewed expert of that firm.

4.1. Alpha

Alpha is a large power producer with a focus on thermal and wind energy projects. Alpha’s first offshore wind farm was connected to grid in 2017. Supposing a lifetime of 20-25 years, the decommissioning for Alpha is still far in the future. However, in terms of power production, offshore wind energy is already the most important energy source for the company.

4.1.1. Learning activities

For Alpha, learning about decommissioning is in second place. Alpha would like to first make various repowering or life-cycle extension considerations before creating more detailed decommissioning concepts:

“For us, repowering or continued operation after 20-25 years is the first consideration. If these options are excluded, then we can consider decommissioning. […] if you look at other infrastructure projects or power plants in particular, usually they all have a longer operation time than originally planned. In terms of lifetime, they all run longer at the end of the day than the original plan.” (Alpha1).

Alpha1 explains that he believes that “a large part of the infrastructure that is being created out there right now will not have reached the end of its service life after 20 years or even 25 years. And insofar as it is not necessarily economical to dismantle at that time. It’s not a question of fundamental dismantling, but rather the time of dismantling”.

(30)

24 need to be created for regulatory authorities because the operator needs to prove safety standards to get an extension of the operating permit. Thus, the company is first trying to acquire knowledge about the exact end-of-life timing and how to extend the lifetime. To make decisions and to know which parts may need to be renewed at the end-of-life, Alpha is already collecting data of the current operation (learning by using):

“[In order to extend the lifetime] I have to do a lot in advance, I have to provide a lot of evidence to be able to guarantee stability and operational safety. […] One of the keywords here is timely data acquisition via condition monitoring systems in the various areas of the farm.” (Alpha1).

With such a database Alpha hopes to learn about the expected lifetimes of the assets. Condition monitoring data can provide evidence for the regulatory authorities to show that the wind farm is still safe to operate in case lifetime extensions or repowering options are pursued.

Alpha does not have a large R&D department and direct learning activities about decommissioning do not exist yet. This became clear both from company documents and from statements in the preliminary telephone conversation:

“[We are not the company] where different concepts already exist or where the engineering side is already intensively taking care of [decommissioning] […] What is Alpha doing? We are following this [the innovation system’s learning activities about decommissioning] with interest, we are of course also interested in the studies on renewed cost estimates because this also defines the amount of collateral at the end of the day and must also be included somewhere in the profitability analysis of the entire case, namely, the lower I can assume the costs, the more I can show comprehensible efficiencies, the lower the collateral that must be deposited. So yes, we are looking into that, but we don’t see ourselves as a driver of innovation for decommissioning concepts.” (Alpha1).

(31)

25

“[It is essential] to make the actual decommissioning process as efficient as possible. I think decommissioning concepts have been very rudimentary in the past, that’s no criticism of the people who created them, but it was planned as precisely as you could do 20 or 25 years in advance, but I think there is still a lot of potential in making this efficient, especially for the orders of magnitude that are to come.” (Alpha1).

Alpha is following product and process innovation for decommissioning learning by knowledge spillovers. This also includes the knowledge spillovers from competitors that need to execute decommissioning projects earlier than Alpha:

“It is indeed the case that we learn from the experience of others” (Alpha1).

In brief, next to the hope to benefit from knowledge spillovers of the innovation system, learning by using plays an important role for Alpha as they would like to know, as precisely as possible, when decommissioning will take place and what the weak points in the offshore wind farm are.

4.1.2. Incentives to innovate

Alpha1 stresses the importance of innovation in decommissioning processes and techniques with the main goal of reducing costs. Although learning activities happen mostly outside their company, they can profit from an efficient decommissioning plan. Alpha’s focus so far has been on the development of new offshore wind farms. Hence, the main incentive to innovate is in the construction and operations phase in order to win further tenders to build new offshore wind farms:

“We will develop more farms and hope to build more farms. […] There will certainly be more to come” (Alpha1).

(32)

26 However, these decommissioning programs are often very crude and therefore there is not yet much pressure for Alpha to fund R&D activities, for example:

“We still have time and will use this time and want to use this time of course. But why should we put a lot of resources into an activity now, when we assume today that it will happen maybe in 20 years, maybe even later?” (Alpha1).

4.2. Beta

Beta is a subsidiary of a large utility company and is responsible for all its offshore wind energy activities. From a legal perspective, the parent company is the asset owner and has the decommissioning liability. Nevertheless, Beta, as a 100% subsidiary, acts as if they are the offshore wind farm owner. Beta’s first offshore wind turbines started their operation in 2010 and in the meantime, three offshore wind farms are in operation. Beta does not plan to build any further offshore wind farms for themselves because the financial commitment is too high. However, what is special about Beta is that they plan the building of offshore wind farms for external customers on an international basis. Operations and maintenance services are also part of their service portfolio. In the future, it is also planned to offer decommissioning services for the sector. Since their first offshore wind farm was built for an operating time of 20 years, half of the expected lifetime has already passed.

After 10 years of operations Beta starts to think about the basic three end-of-life options (Figure 1):

“We are beginning to approach these questions. Do my assumptions of ten years ago still apply today? If not, what has changed? What influence does this have on costs? […] and that is what we’re actually doing at the moment, at least for our wind farm, which has now reached about half of its nominal operating horizon.” (Beta1).

(33)

27 created in the planning phase of the offshore wind farm will probably not occur in that way. Even the planned lifetime of 20 years is uncertain. It might make sense to repower the assets a few years earlier if this increases the return on investment or “because older assets have now reached a level of maintenance costs that is hardly sustainable.” (Beta1). Thus, for Beta, it is important to learn about possibilities for the end-of-life phase several years in advance.

Learning by using is an important knowledge source for Beta. As a service provider, Beta also gets direct insights into different offshore wind assets. Moreover, learning by searching plays a role, too, as Beta already conducts project studies to plan and simulate end-of-life options. However, these are only project studies and there is no decommissioning team searching for innovation on a broad scale.

Finally, Beta1 emphasizes that learning by doing will play an important factor:

“You can get better at this if you have done it before. With such a mega-project you can learn a lot just by doing it, you can't plan everything perfectly in advance.” (Beta1).

Beta has an incentive to innovate because they want to offer decommissioning services not only inhouse but also for other offshore wind farm operators. They are not only learning to handle their own decommissioning projects but also, as a service provider, to offer decommissioning services to third parties:

“We are a technical service provider, our competencies, we say, cover the entire life cycle of a wind farm from planning, construction, commissioning to decommissioning. In this respect, we are very interested when decommissioning projects take shape, in supporting them with our technical know-how and in developing them further as part of our service portfolio.” (Beta1).

(34)

28

“…these ships, which can quickly cost a hundred thousand Euros or more a day just to be there. This shows that such projects, such as decommissioning of the wind farm, naturally have a considerable financial impact.” (Beta1).

In short, Beta is interested in decommissioning innovation, but its learning process has just started. Learning by using the current assets is the main source of knowledge acquisition at the moment. Some projects already deal with the decommissioning topic on paper because practical projects are still to come. Beta is optimistic that there will be innovation soon:

“The topic will inevitably come up on our table, and we will devote ourselves to it, we will devote ourselves to it much more than we do today. [...] In five years at the latest, many things will be much clearer, and many more players will be there who are really thinking about how to do it.” (Beta1).

In short, learning by doing is a firmly planned component of Beta’s preparation. The drive of Beta’s learning processes comes mainly from the fact that their oldest wind farm is already eleven years old and that in the future money should be earned with decommissioning.

4.3. Gamma

With the goal to become a climate-neutral company, Gamma, a large energy supply company, shows that renewable energy projects will play a major role in their business portfolio in the coming years. The company has already invested heavily in offshore wind farm projects and operates four offshore wind farms with more to come.

(35)

29 wind farms in deeper waters. Decommissioning, however, does not play a role in the R&D project or in the design phase of these floating wind turbines.

The commercial operation date of Gamma’s oldest wind farm is 2011 and the company already wants to gain knowledge about decommissioning options. Gamma is a member in a publicly funded decommissioning project and shares information with the research institutions. Within the company, offshore wind farm decommissioning does not play a role yet. Hence, Gamma also does not share any information with other actors of the innovation system:

“No, we do not exchange knowledge with other companies. We exchange knowledge with the coordinating universities [of the decommissioning research project]. But there is no official exchange among each other except, if you will, there is information exchange at conferences and the like.” (Gamma1).

In brief, there are no active learning activities yet, although Gamma is generally open for learning by interaction. According to Gamma1, it is still too early to deal with the topic in the company.

Although Gamma is an active member of an offshore wind farm decommissioning research project, the issue is not relevant yet within the company. Gamma prepares a rough decommissioning concept for the approval procedure. Actively they do not acquire knowledge inhouse:

(36)

30 Gamma sees the task for innovation as belonging to the other actors in the sectoral system of innovation. Since large parts of the decommissioning execution are outsourced to third parties, it would be important for other actors to find innovative and cost-effective concepts:

“The whole issue of decommissioning is always done under the aspect of costs, i.e. whoever uses the most cost-effective and perhaps most environmentally friendly process gets the contract.” (Gamma1).

4.4. Delta

Delta operates two offshore wind farms and has no other energy generating assets. Although Delta is majority-owned by a large energy group the company operates independently and has planned and operated its two offshore wind farms from the very beginning. Therefore, the head of the R&D department emphasizes that the Delta team would have a strong identification with their offshore wind farms. They are particularly interested in the further operation of the wind turbines, as these are Delta’s only power generating assets. The offshore wind farms were financed by the owners of Delta. The decommissioning liability lies with Delta itself. The company has far less than 200 employees.

According to Delta1, the decommissioning timing is still uncertain. The decommissioning after 20-25 years of operations is not a fixed date for Delta and, if possible, the company wants to extend the lifetime of the two wind farms:

“In the end, this [life-time extension] is only a hope. We hope that we can or may continue to operate the farm – as long as this makes economic sense.” (Delta1).

(37)

31

“decommissioning is still in the far future […] for us and for many other operators, who are still in the construction phase at the moment, this is simply not an issue.” (Delta1).

Moreover, Delta will only conduct the decommissioning planning. The decommissioning itself will be completely outsourced to third parties. Thus, Delta sees no need to concern itself with technical or practical issues of dismantling.

However, Delta is also a member of an offshore wind association and according to Delta1, a certain exchange of information on regulations or the state of the art takes place there. This association is the most important platform for Delta to gather information and to conduct learning by interaction.

According to Delta1 learning about end-of-life options is important. However, there is the uncertainty that even if life-time extension options are found, they will not be approved:

“But at the end of the day it is a date set by the authorities. So theoretically and practically the authority can demand that we must dismantle the assets.” (Delta1).

According to Delta1, the many regulative uncertainties prevent the company from actively exploring new end-of-life opportunities. Delta does not want to invest in learning activities if they do not pay off in the end:

“For example, there are currently areas that may be designated as nature conservation zones, which has already happened in some cases. Or, for example, new fishing zones. If our farm is located in such a newly designated fishing zone, or a nature reserve, then you will certainly not get a permit for repowering.” (Delta1).

Apart from the fact that Delta’s offshore wind farms are only a few years old and decommissioning is therefore still a long way off, the issue of decommissioning will not be addressed at Delta until there is better planning certainty.

(38)

32 Delta is a passive observer and waits to see what will be done with competitors’ older wind farms. When asked whether Delta waits and builds on the knowledge that is then available in the sector after the first offshore wind farms have been decommissioned or repowered, the answer was:

“Yes, that’s exactly our strategy.” (Delta1).

4.5. Epsilon

Epsilon is a large electric utility company with a strong focus on renewable energy projects. Epsilon is already operating three offshore wind farms, and more are already under construction or planned. Their first offshore project was commissioned in 2014.

Similar to the other offshore wind farm operators, Epsilon is roughly thinking about the issue in each offshore wind farm design phase due to the legal requirements. Nevertheless, acquiring further knowledge about decommissioning is already a topic at Epsilon.

Epsilon will not be the first actor that needs to conduct a decommissioning project. Therefore, they hope to profit from knowledge spillovers:

(39)

33 Another finding is that Epsilon is very active in industry associations and tries to acquire existing knowledge and to learn about possibilities for the end-of-life phase. Epsilon actively tries to learn about decommissioning processes to influence the current regulatory debate. Epsion2, for example, reflects:

“if you sit back, of course, and don't actively gather knowledge now, then you have little say in the circles that are now deciding on decommissioning guidelines. And we do not want that.” (Epsilon2).

The focus of this interactive process of knowledge gathering is to learn more about current and coming regulations. Detailed plans as to what will happen to their offshore wind farms at the end of their lifetime are not yet being made. Consequently, Epsilon hopes to profit from knowledge spillovers, but learning by interaction especially at meetings with other actors of the sectoral innovation system is important for the company.

A positive sustainable image is important for Epsilon and is strongly emphasized in the external presentation of the company. This is also supported by interview statements, such as stressed by Epsilon1:

“At the end of the day, it’s all about our reputation” (Epsilon1).

The analysis has shown that the sustainable goal is the strongest incentive for Epsilon to deal with the decommissioning issue already today. Epsilon2 describes it in the following words:

(40)

34 Moreover, Epsilon wants to increase their offshore activities, and hence, they think it is useful to acquire knowledge about decommissioning:

“If we want to continue to get any projects worldwide, then we have to actively show here that we care [about decommissioning].” (Epsion2).

Technical aspects or detailed logistic processes, on the contrary, are not yet explored by Epsilon. Epsilon trusts their supply chain partners who will have dealt with the detailed questions up to that point:

“There will be companies on the market that take care of detailed technical questions, they will work out the logistics questions and offer their services […] I don’t think that we need to develop something in detail.” (Epsilon1).

Similarly, Epsilon2 confirms:

“The know-how comes mainly from the suppliers.” (Epsilon2).

(41)

35

5. Cross-case analysis

The cross-case analysis compares the results of section 4 using the conceptual framework (Figure 3) as a guideline. As stated in the research questions it is of interest if there are any differences in the offshore wind farm operators’ learning activities and how firms address the different end-of-life options (Figure 1). Table 4 briefly summarizes the results of section 4 and section 5 for each company

The five case companies are not equally interested in the decommissioning problem. However, this is only partly because the task is a necessary evil. What is more relevant is which life cycle phase the offshore assets are in and what other incentives there are to deal with the issue (e.g., company reputation). All companies are aware of their responsibility when it comes to dismantling. There is no fundamentally negative attitude towards the topic, although Beta is the only actor that financially profits from cost savings and also plans to generate revenue from decommissioning competitors’ wind farms. Especially Alpha, Gamma, and Delta indicated that most of the decommissioning tasks will be outsourced to suppliers. At least for Gamma and Delta, this could partly explain the relatively low incentive to take action towards learning about decommissioning.

Thus, another finding is that innovation occurs mainly with other actors of the sectoral system of innovation. As indicated by Alpha1 and Beta3 there are, for example, “construction companies and some innovative engineering offices” that start with the development of innovative processes. According to them the offshore wind farm operators trust in the suppliers’ knowledge base and will hire them for most of the decommissioning planning and execution tasks. In that regard, Alpha1 adds:

“I believe that at the end of the day this will be something where the construction companies will be the information carriers because there will certainly be a lot of cost pressure when it comes to decommissioning” (Alpha1).

(42)

36 operators Alpha, Delta, and Epsilon have relatively new turbines and do not consider concrete decommissioning options yet. Delta1, for example, asks:

“Why should we think about it [decommissioning] now? It makes more sense to do so when it becomes relevant for us.” (Delta1).

This attitude was found also with Alpha and Gamma, for example:

“We are still in the phase where most operators actually focus on the commissioning and operation phase. We certainly have an interest in knowing how to dismantle the whole thing cost-effectively. But that’s still a very long way off.” (Alpha1).

Beta and Gamma’s wind farms are already older than ten years. Consequently, these two actors need to make end-of-life decisions earlier. Although Gamma is a member of a publicly funded research project, the firm thinks that it is still too early to discuss the topic in depth. Beta is the only case company that starts to prepare decommissioning concepts internally. Thus, Beta is the only actor that is already conducting some learning by searching activities with a hope to acquire knowledge from learning by doing. Other firms of the innovation system, such as recycling firms, push the topic much more because they see the decommissioning task as a market for their services in the future. One interviewee describes that:

“The construction companies or those who hope to make a profit in the future from decommissioning projects, they are rightly starting quite early […] [The operator then says] everything you do is quite interesting, but for us, it has no concrete benefit today.” (Beta2).

(43)

37 more globalized and more actors and a bigger knowledge base make up the system. Although the exchange of knowledge is still positively highlighted by many actors, some statements describe a relatively lower level of interaction today:

“So, at the very beginning [when the offshore wind energy expansion began] there was still a community spirit. You could see that when you go to industry events, for example, the whole industry was much smaller and the relative degree of networking in my view went even further. […] or what others are doing. An exchange of knowledge was a little easier in the early days of the industry.” (Beta1).

(44)

38

Table 4: Cross-case comparison

Case Alpha Beta Gamma Delta Epsilon

First offshore wind farm (COD, Commercial operation date)

2017 2010 2011 2014 2014

Specific characteristic Focus on life-cycle

extension and repowering options

Subsidiary of a large utility company; aim to offer decommissioning services Member of a publicly funded research project Strong identification with their offshore wind farms

Strong focus on environmentally sound

decommissioning

Further offshore wind farms planned

Yes No Yes Not yet/further

plans open

Yes

Learning activities Learning by

using; hope to profit from learning by knowledge spillovers Learning by using; partly learning by searching; hoping for innovation from knowledge spillovers and learning by doing

Learning by interaction; however, no direct learning activities yet Learning by interaction; however, no direct learning activities yet Learning by interaction; hope to profit from learning by knowledge spillovers

Incentives to innovate Medium:

Still too early; Main goal to reduce costs

Medium:

Costs and future business model

Low:

Still too early; Responsibility for finding innovative concepts lies with other actors

Low:

(45)

39

6. Discussion

This thesis examines how and why firms learn and acquire knowledge about decommissioning. As already discussed, the issue comes along with collective environmental uncertainty that is faced by the whole sector (e.g., expected lifetime). Most of the interviewed firms have recognized the importance of finding efficient and environmentally sound solutions for decommissioning, but various uncertainties and problems have been mentioned that present obstacles along the way. This section discusses the main uncertainties and challenges mentioned in the interviews with the offshore wind farm operators and adds findings from other interviewed actors of the sectoral innovation system (Table 1). The issues are linked to the research question and methods to foster the learning processes of the sectoral innovation system are identified. In this way, the findings of Section 4 and Section 5 are incorporated into the discussion. Subsequently, the section provides theoretical and practical implications.

6.1. Main uncertainties and challenges

The most crucial uncertainties, mentioned by the interviewed experts (Table 1), are outlined in this subsection in connection with existing literature. Table 5 summarizes the four main aspects which are then discussed in detail.

Table 5: Main uncertainties and challenges for offshore wind farm decommissioning

Uncertainty/ Challenge

Specification Example quote

End-of-life decision

Life-cycle extension, Repowering, or Decommissioning

(46)

40 Timing The expected lifetime of 20-25

years, which is repeatedly mentioned in the literature, is uncertain.

“I would even dare to make a prognosis that especially with the older wind farms that already exist, there will be quite a number of people [firms] who plan to dismantle parts of the turbines earlier than originally thought” (Beta1).

“A large part of the infrastructure that is being created out there right now will not have reached the end of its service life after 20 years or even 25 years.” (Alpha1).

Regulatory aspects

It is uncertain which regulatory conditions apply for

decommissioning (e.g., sound insulation for marine life, disposal of the blades).

“Which conditions do we get from the approval authorities for decommissioning? What do we have to dismantle and how?” (Gamma1).

Environment al aspects

It is not yet clear what effect offshore wind farms have on marine flora and fauna and how environmentally sound decommissioning looks like.

“I really think that the disposal, what you do with the garbage, will be the biggest problem.” (Epsilon2)

6.1.1. End-of-life decision

(47)

41 analysis of the interviews, it became clear that there is disagreement about the end-of-life options. While Alpha intends to extend the operational life of its wind farms, others believe that the components of offshore wind turbines physically do not last longer than 20-25 years. Beta even indicates that the technological and economic life of early offshore wind farms could be less than 20 years.

Large-scale repowering, i.e. extending the lifetime with replacements of large components, such as the whole nacelle, is not feasible. For example, the foundations are not adapted to more powerful turbines and the cable structures are also not designed for repowering with larger turbines. Multiple interviewees confirm the impossibility of repowering with new turbines and an offshore service provider summarizes the repowering options as follows:

“The foundation structures are designed to withstand the dynamic loads for a lifetime of 20 years. Anything beyond that will probably not be approved or certified so easily. [...] In addition, in 20-25 years there will be completely different turbines, which will then no longer fit the design at all. [...] So there will be no possibility to use the infrastructure for bigger and new turbines. Yes, you can replace individual components. But in principle, a repowering project is like a decommissioning project with subsequent reconstruction.” (Offshore service provider1).

Thus, learning about end-of-life options is essential but practical experience is needed because the optimal solution cannot only be found through R&D or learning by searching in advance.

6.1.2. Timing

(48)

42

“Many components in the windmills and there are many operators which are also naive, perhaps because they do not have the experience. But many components won't last that long, so a gearbox won't last 25 years. [...] Well, I assume that after 20 years we will have many parks where 20 of the 80 turbines have already been shut down and are no longer running. [...] This means that an extension of the operating time is theoretically possible, but I believe that in many cases it does not make economic sense.” (Offshore service provider1).

The right timing is also related to the end-of-life options and dependent on regulatory permission. It shows that it is not only the actors and knowledge that are decisive but also institutions as the third building block (Figure 3). Interaction and knowledge exchange between offshore wind farm operators and public authorities are essential to reduce uncertainty.

6.1.3. Regulatory and environmental uncertainty

Although there are national differences in the decommissioning regulation, almost all actors mention regulatory uncertainty as the biggest obstacle to planning for decommissioning. As environmental aspects are often directly related to regulations, the two uncertainties are described together in this subsection:

“I would say [the greatest uncertainties are] the legal aspects, whereby environmental law is a subset of the legal aspects. From what I know, it is not quite possible to foresee what environmental regulations, for example, will be when a decommissioning project is actually executed, let’s say in ten years from now.” (Beta1).

The challenge of decommissioning offshore wind farms: how operators learn within the innovation system Dual Master’s Thesis