Electric Vehicles & Fuel Retailers

1 DEGREE PROJECT IN INDUSTRIAL MANAGEMENT, SECOND CYCLE, 30 CREDITS

STOCKHOLM, SWEDEN 2020

Electric Vehicles & Fuel Retailers

Challenges with the Provision of Fast Charging in

Swedish Fuel Stations

GUSTAV THORSELL

KTH ROYAL INSTITUTE OF TECHNOLOGY

3

Electric Vehicles & Fuel Retailers

Challenges with the Provision of Fast Charging in Swedish

Fuel Stations

by

Gustav Thorsell

Master of Science Thesis TRITA-ITM-EX 2020:324

KTH Industrial Engineering and Management

4

Elektriska fordon & drivmedelsbolag

Utmaningar med tillhandhållandet av snabbladdning på

svenska drivmedelsstationer

av

Gustav Thorsell

Examensarbete TRITA-ITM-EX 2020:324

KTH Industriell teknik och management

5

Master of Science Thesis TRITA-ITM-EX 2020:324

Electric Vehicles & Fuel Retailers

Challenges with the Provision of Fast Charging

in Swedish Fuel Stations

Gustav Thorsell

2020-06-12

Lars Uppvall

Mats Engwall

Swedish Transport Administration

Contact person

Björn Hasselgren

Abstract

Battery-Electric Vehicles (BEVs) are gaining increasing momentum and their adoption is projected to continue at a rapid pace. In contrast to the predominate vehicle type, Internal-Combustion Engine Vehicles (ICEVs), the BEVs rely on electricity rather than conventional fuels for propulsion. For the fuel retailers, the actors managing fuel stations and supply fuel to vehicle users, the shift towards BEVs may impose a demand decline for their core product. One way of mitigating the effects of this transition for fuel retailers, has been attributed to the provision of charging for electric vehicles in fuel stations, something that is offered to a limited extent by a few fuel retailers in Sweden today. The purpose of this study has been to explore the main challenges associated with the provision of fast charging in Swedish fuel stations.

Anchored in a study of the four largest fuel retailers in Sweden, the findings suggest that the provision of fast charging is subject to financial challenges in terms of long investments horizons, with the associated investment risk being negatively influenced by uncertainties of technology development, future price levels for fast charging and future utilization rates of the chargers. Furthermore, with the lack of direct financial incentives of charging sales in the short-term, other motives have been observed for the provision of fast charging, namely goodwill and traffic building. The prevalence of different types of fuel retail formats have been shown to influence the extent of which benefits can be claimed from fast charging provision, and that this is manifested in a bias towards certain formats and locations for the choice of fuel stations to provide fast charging in. Challenges have also been identified in the sense of fulfillment of necessary preconditions to facilitate a business model suitable for fast charging provision. The study concludes that the main challenges revolve around the attainment of a financially sustainable business case for fast chargers.

6

Examensarbete TRITA-ITM-EX 2020:324

Elektriska fordon & drivmedelsbolag

Utmaningar med tillhandhållandet av

snabbladdning på svenska drivmedelsstationer

Gustav Thorsell

2020-06-12

Lars Uppvall

Mats Engwall

Trafikverket

Björn Hasselgren

Sammanfattning

Elfordon blir allt mer populära och utvecklingen förväntas försätta i hög takt. I kontrast med den vanligast förekommande motortekniken för vägfordon, förbränningsmotorn, så använder elfordon just el istället för konventionella bränslen för framdrift. För drivmedelsbolagen, de aktörer som ansvarar för drivmedelstationer och tillhandhåller drivmedel till fordonsanvändare, kan en övergång mot elfordon innebära en minskad efterfrågan på deras kärnprodukt. Ett sätt att hantera konsekvenserna av en minskad efterfrågan på drivmedel, har tillskrivits tillhandahållandet av laddning för elfordon på drivmedelsstationerna, ett erbjudande som idag tillhandahålls av ett fåtal drivmedelsbolag på ett begränsat antal stationer i Sverige. Syftet med den här studien har varit att undersöka de huvudsakliga utmaningarna med tillhandhållandet av snabbladdning på drivmedelsstationer i Sverige.

Baserat på en studie av de fyra största drivmedelsbolagen i Sverige, pekar resultaten på att tillhandahållandet av snabbladdning medför stora finansiella utmaningar. Detta innefattar långa investeringshorisonter, med en investeringsrisk som påverkas negativt av osäkerheter kring teknikutveckling, framtida prisnivåer för snabbladdning och framtida användningsgrad av snabbladdare. Med en avsaknad av direkt kortsiktig finansiell vinning från försäljning av snabbladdning, så har det framgått att det finns andra motiv för tillhandahållandet av snabbladdning, det handlar då om kundflöde och goodwill. Förekomsten av olika stationsformat för drivmedelsförsäljning har visat sig påverka till vilken grad skapat värde från snabbladdningserbjudandet kan tillgodoses, och att det har påverkat val av stationsformat och plats för tillhandahållandet av snabbladdning. Ytterligare utmaningar har identifierats gällande att skapa nödvändiga förutsättningarna för att möjliggöra en affärsmodell anpassad för snabbladdning. En slutsats har dragits att de huvudsakliga utmaningarna med tillhandahållandet av snabbladdning handlar om att uppnå en finansiellt hållbar affär.

7

Table of Content

Abstract ... 5

Sammanfattning ... 6

Table of Content... 7

List of Figures and Tables ... 9

Acknowledgement ... 10 Abbreviations ... 11 1 Introduction ... 12 1.1 Background ... 12 1.2 Purpose ... 13 1.3 Delimitations ... 14 1.4 Contribution ... 14 1.5 Thesis Sponsor ... 14 2 Theory ... 15 2.1 Technological Change ... 15 2.2 Business Models ... 17

2.3 Three Aspects of Retailing ... 20

2.4 Theoretical Framework ... 23 3 Methodology ... 25 3.1 Research Design ... 25 3.2 Pre-Study ... 25 3.3 Literature Review ... 26 3.4 Collection Procedure ... 26 3.5 Data Analysis ... 30 3.6 Research Quality ... 30 3.7 Ethical Considerations ... 32

4 The Industry & Findings ... 33

4.1 A Brief Introduction to the Oil Industry ... 33

8

4.3 Fast Charging Infrastructure ... 38

4.4 Industry Perspectives ... 41

5 Discussion & Analysis ... 47

5.1 Why do fuel retailers offer fast charging? ... 47

5.2 How do fuel retail formats influence the provision of fast charging? ... 50

5.3 How do the inclusion of charging influence fuel retailer business models? ... 55

6 Conclusions ... 59

6.1 What are the main challenges associated with the provision of fast charging in fuel stations, from a fuel retailer perspective? ... 59

6.2 Academic Implications ... 60

6.3 Sustainability Implications ... 60

6.4 Future Research ... 61

7 References ... 62

9

List of Figures and Tables

Figure 1. Strategic responses to technological change (Adner and Snow, 2010). ... 16

Figure 2. Figure based on the Ansoff Matrix (Ansoff, 1957). ... 17

Figure 3. The Business Model Canvas framework (Osterwalder and Pigneur, 2010). ... 18

Figure 4. An overview of oil industry value chain steps. ... 33

Figure 5. Vehicle energy offerings mapping. Figure based on the Ansoff Matrix. ... 47

Table 1. Retail Format Framework. ... 24

Table 2. Table of conducted interviews. ... 28

Table 3. Email correspondence with informants. ... 30

Table 4. Outline of fuel station division by major fuel retailers in Sweden. ... 37

Table 5. Outline of general vertical structures of fuel stations in Sweden. ... 38

Table 6. Retail format mapping. ... 51

Table 7. Full-service stations, highlighted downstream relationship and financial viability. ... 52

Table 8. Full-service stations, highlighted location. ... 54

Table 9. Traditional Role. ... 55

Table 10. Fast charging at fuel stations... 56

10

Acknowledgement

I would like to express my gratitude to several people who have supported, guided and enabled me to write this thesis. A special thanks to my supervisor Björn Hasselgren and Elin Näsström at The Swedish Transport Administration for your guidance, feedback, and the opportunity for writing this thesis on your behalf. I would also like to express my gratitude to my KTH supervisor, Mats Engwall who have guided me through the process of academic writing and contributed with valuable insights that have helped me shape this thesis. Thanks to all that have participated in interviews for your time and effort to help me create the foundation which my thesis builds on. Furthermore, I would like to thank my seminar leader Lars Uppvall and the members of my seminar group for your feedback and help to develop my thesis. I would also like to express my gratitude to my family and my partner for all your support, enabling me to write this thesis.

11

Abbreviations

B2B – Business to Business B2C – Business to Consumer BEV – Battery Electric Vehicle

COCO – Company Owned and Company Operated CODO – Company Owned and Dealer Operated DODO – Dealer Owned and Dealer Operated ERS – Electric Road Systems

GHG – Green House Gas HDV – Heavy Duty Vehicle kW - Kilowatt

kWh – Kilowatt-Hour LDV – Light-Duty Vehicle SEK – Swedish Krona

12

1 Introduction

1.1 Background

From the late 19th _{century, electricity as a mean of energy provision have on repeated occasions} been up against and displaced its incumbent counterparts. This have been seen to take place in the city and home lightning industry, with electric lightning replacing gas lightning, mainly due to its superior lightning effectiveness (Utterback, 1994). The train and railway industry have experienced similar shifts with the transition from coal powered to electric trains, seeing first deployments in Sweden as early as 1905, this time fueled by the motivation of decreasing the dependence on imported coal (Vattenfall, n.d.). The development has continued and electricity has become the predominant energy delivery system for trains in Sweden, with 80 % of the Swedish railway system being electrified today (Swedish Transport Administration, 2019a).

Currently, the question revolves around road-based transportation, an area which today is dominated by Internal-Combustion Engine Vehicles (ICEVs) using gasoline and diesel for propulsion. These vehicles account for 90.6 % of the cars, 97.5 % of the Light-Duty Vehicles (LDVs) and 98.6 % of the Heavy-Duty Vehicles (HDVs) in Sweden (SCB, 2020). However, the use of the vehicles with conventional fuels have been widely acknowledged to negatively impact the environment by large emissions of green-house gases (Ministry of the Environment, 2018). The technology of Battery-Electric Vehicles (BEVs) has matured significantly in the 21st _century, becoming increasingly competitive compared to ICEVs. BEVs have in turn been attributed with the potential of reducing vehicle related emission during use (IEA, 2019). This view is shared by the Swedish Government, who see vehicle electrification as one key path alongside renewable fuels for reaching emission reduction goals. Especially the goal of reducing the emissions of Green-House Gas (GHG) emissions from domestic transportation (excluding aviation) by at least 70 % in 2030, compared to 2010 (Ministry of the Environment, 2018). There are some challenges associated with BEV usage i.e. range anxiety (Noel et al., 2019) and a lack of fast charging stations (locations with one or more charging points) in some areas in Sweden (Swedish Transport Administration, 2018). Nevertheless, the adoption of BEVs is projected to continue at a rapid pace (Power Circle, 2019; Transport Analysis, 2019).

In the case of ICEVs, vehicle energy is generally provided by refueling at fuel stations (BCG, 2019). In Sweden, there are roughly 3000 fuel stations, making up an extensive infrastructure for refueling (SPBI, 2019a). Fuel stations are established in several different types, ranging from self-service locations to full-self-service facilities offering a variety of complementary products and services to customers (SPBI, 2019b).

13 means of energy delivery, but also alters the vehicle user behavior with the majority of electric charging taking place at home (Swedish Transport Administration, 2018).

Technology imposed demand changes have on numerous occasions been addressed to explain challenges to firms in the industry where the technological change originates from (e.g. Christensen, 1997; Tushman and Anderson, 1986; Utterback, 1994). Scholars have however paid little attention to the effect of technological change on firms in other industries with businesses dependent upon the incumbent technology in question, an issue apparent with the development and increasing adoption of BEVs potentially driving a decline in conventional fuel demand. This novel issue is yet to receive notable attention from scholars. Instead, consultants have been some of the first to pick up on the issue (Accenture, 2018; Arthur D. Little, 2019; Oliver Wyman, 2017; REPL, 2019). Their primary contribution lies in the recognition of the issue, saying little of the practical measures necessary for problem mitigation. BCG (2019) have pursued the issue further, pointing out several areas of attention for fuel retailers to mitigate the effects of the issue. One central area is charging for electric vehicles, something that is offered to a limited extent by fuel retailers in Sweden today (Österberg, 2020). Additionally, BCG (2019) discuss the need for conjointly addressing business models and that certain fuel station retail formats, i.e. the outlines of the fuel stations, may see a shift in relevance in the future. These measures have however been described as manifestations of desired future states of fuel retailing business, saying little of how the current way of doing business affect the transition towards the offering of charging for electric vehicles. In light of the scarce research on the topic, a large gap is exposed regarding the fundamental challenges faced by fuel retailers in a transition towards the offering of charging for electric vehicles. Challenges which are potentially complicated and influenced by the concerned motives of fuel retailers, the variety of fuel station types, and the ambiguity surrounding future business models.

1.2 Purpose

The purpose of this study is to explore the main challenges associated with the provision of fast charging in Swedish fuel stations. This calls for a deeper understanding of fuel retailers’ perceptions of the associated benefits, how different fuel retail formats affect the deployment and the business model implications of introducing fast charging.

The research questions are presented below, the Main Research Question (MRQ) is devoted to aligning the study with the purpose, whereas the Sub-Research Questions (SRQs) are addressed to answer the MRQ.

MRQ: What are the main challenges associated with the provision of fast charging in fuel stations, from a fuel retailer perspective?

SRQ1: Why do fuel retailers offer fast charging?

14 To fulfill the purpose and answer the research questions, the current state of the Swedish fuel retailing industry is mapped, and interviews with informants representing industry actors are conducted to gain insights to their respective problem perceptions. A framework is proposed and applied to describe the established fuel retail formats in Sweden.

1.3 Delimitations

Several delimitations are made in this study, which are followingly presented to clarify the study scope. Regarding charging types, the focus is devoted to public fast charging, whereas other types of charging are considered to facilitate the discussion on fast charging. Regarding vehicle types, the primary focus of the study is on cars. However, LDVs and HDVs are also considered. LDVs and HDVs are defined as trucks with a total laden weight up to 3.5 tons respectively over 3.5 tons. Electric busses are not included in the scope, as most charging is expected to take place in depots (Swedish Transport Administration, 2018). Further delimitations are made to only include fuel retailers in Sweden. The motivation is to decrease the differences introduced by considering fuel retailers across several nations with subsequent differences in policy, technology adoption rates and national market particularities. Major actors are prioritized in the selection of fuel retailers to be included.

1.4 Contribution

The study contributes to an increased understanding of the barriers and challenges fuel retailers face in order to provide charging for electric vehicles. Furthermore, it provides a description of the current industry landscape with a focus on major actors, and outlooks on the industry´s contribution to fast charging infrastructure for electric vehicles.

This study also contributes to the literature on retail formats, for instance Fox & Sethuraman (2010), by proposing a framework for describing retail formats. In line with the interpretation of retail formats by Reynolds et al. (2007), this framework emphasize the connection between the retail format and the concerned business model.

1.5 Thesis Sponsor

15

2 Theory

The academic research on fuel retailing have been primarily concerned with market structure influence on fuel prices (e.g. Carranza, Clark & Houde, 2015; Nowakowski & Karasiewicz, 2016; Faber & Janssen, 2019; Sharma, Jayaraj & Jain, 2017; Borenstein & Shepard, 1993; Hastings, 2004; Houde, 2012), influence of assortment and pricing on competition (Kishore and Patel, 2012; Soetevent and Bružikas, 2018; Stavrakakis and Chountalas, 2015) and location choice influence on performance (e.g. Slade, 1992; Chan, Padmanabhan & Seetharaman, 2007; Gagné, Nguimbus & Zaccour, 2004; Netz & Taylor, 2002). These works however have little significance for addressing the issue at hand. Instead, the attention is turned to other areas of literature.

The focus is firstly turned to the literature on technological change to explore why technological change may pose a threat to incumbent firms. Strategic alternatives of firms in the face of technological change are discussed to illuminate different strategic courses of action to address the threat. The business model concept and one division of business model components are followingly introduced. These serve as units of analysis to explore how the inclusion of new offerings can alter the way fuel retailers conduct business. Shifting the focus to retail, the nature of retail innovation is described to illustrate the means of which retailers can adapt and develop their businesses. Retail formats are introduced as a concept for differentiation of retail salespoints and as a unit of analysis for retail innovation. Green product assortment is discussed to show how retail format innovation by altering the product assortment may enhance brand goodwill.

2.1 Technological Change

Technological Change and the Demise of Incumbent Firms

Linking innovation to the demise of incumbent firms, Schumpeter (1942, cited in Lipset, 1993) shed light on the potentially destructive forces of innovation. The concepts of technological paradigms and trajectories (Dosi, 1982), enabled a description of the state in which certain technological problems are pursued and the related pattern of problem solving activities. A shift to another technological paradigm would subsequently redefine the set of technological problems to be addressed. Such a shift was described as a discontinuity in technological change. An important notion is the existence of different types of innovation. Incremental innovation implied continued progress within the current paradigm, whereas radical innovation addressed an emerging new paradigm.

16 responses due to heavy investment into existing technology and infrastructure, attention requirements from main activities (Utterback, 1994), and loss of firm unit flexibility when pursuing economies of scale (Abernathy and Utterback, 1978). New entrants have been ascribed the abilities to displace incumbents in situations of technological change, with e.g. Utterback (1994) emphasizing the advantage of not having to defend or maintain existing technology and/or activities. However, Bergek et al. (2013) argued that the new entrants‘ abilities to displace incumbents often have been overestimated, and that the incumbents’ abilities to adapt have been underestimated.

Technological Change & Strategy

One way for incumbent firms to address technological change is by strategic measures. Howells (2002) suggested the existence of two generic strategic alternatives in the face of technological change. Whether to stay in the market with the incumbent technology, and if not, whether to participate in the market with the substituting technology.

Figure 1. Strategic responses to technological change (Adner and Snow, 2010).

17 Existing Market New Market

Existin

g

Pr

oduc

t

Market Penetration Market Development

N

ew

Pr

oduc

t _{Product Development Diversification}

Figure 2. Figure based on the Ansoff Matrix (Ansoff, 1957).

Broadening the scope of available strategic alternatives of the firm, diversification is a strategic route that can encompass both technology and products (Breschi et al., 2003). Ansoff (1957) primarily saw diversification as an alternative to facilitate company growth alongside the alternatives of market penetration, market development and product development (see Figure 2 for orientation). However, diversification was also described as a course of action when holding expectancies of demand decline for the primary offered product. This is an aspect that have been pursued further in the area of corporate turnaround strategies, an area mainly concerned with strategies and mechanisms for revitalizing declining firms. In this context, diversification can be seen as an offensive repositioning action. This can be contrasted to other repositioning actions of targeting market niches, market penetration, decreasing prices and divesting products, which are mainly defensive in nature. Offensive repositioning as a strategy for firm turnaround may be appropriate in cases when non-diversified firms are in decline due to external causes (Hoffman, 1989).

2.2 Business Models

The Concept of Business Models

18 & Massa (2011) suggested that business models may be seen as a mechanism, the gestalt of a firm, or as a strategy concept depending on the research area in question. There is also an ongoing debate of whether business models is a distinct research area or a part of strategy research (Massa et al., 2017).

Though the prevalence of different opinions on what business models are, the business model concept is generally addressed in order to explain what businesses do and how they do it (Zott et al., 2011). As described by Shafer, Smith & Linder (2005), this has revolved around the analysis of a multitude of business model components. One type of selection and arrangement of these can be made in accordance with the business model canvas (Figure 3), a business model framework by Osterwalder & Pigneur (2010). Going forward, this framework is used for outlining and describing the nine included components.

The Business Model Canvas

Figure 3. The Business Model Canvas framework (Osterwalder and Pigneur, 2010).

Customer Segments

19 Value Propositions

The value proposition is what the company offer to its customers in terms of products and services (Demil and Lecocq, 2010). The suitedness of the value proposition design is heavily connected to the customer segment in question. Johnson, Christensen & Kagermann (2008) emphasize the importance of understanding the needs of the customer and to design the offering with the particular problem of the customer in mind.

Channels

Channels concerns how the firm communicates with, and by which means the value proposition is delivered to customers. This encompass five purposes, raising awareness of the value proposition, helping customers to evaluate the benefits of the value proposition, facilitating a purchase, delivery of the value proposition, and after sale support (Osterwalder & Pigneur, 2010). The choice and management of channels should be made with care in order to reach the designated customer segment (Weill & Vitale, 2001 cited in Osterwalder, 2004).

Customer Relationships

There are several reasons to why management of customer relationships is important to firms, for instance to acquire and retain customers, and for upselling. The firm may intend to pursue one or more types of relationships. Personal assistance and dedicated personal assistance represent relationships where close human interaction is desired. With service and automated self-service, the customers are to be given sufficient means to tend themselves or tend themselves with automated help. Utilizing community relationships can be beneficial when the dissemination of knowledge between customers is desired. Another type of relationship is customer co-creation, where customers are involved in value-creating activities (Osterwalder & Pigneur, 2010).

Revenue Streams

Generating revenue streams revolves around pricing and pricing mechanisms for the value propositions. The revenue stream may be of either transaction-based type, with revenues occurring from particular purchases, or in reoccurring form, as ongoing payments either to deliver a value proposition or by customer support. Pricing mechanism may for instance be used in line with bargaining, list, market dependent, auctioning or volume dependent pricing. Revenue streams can stem from for example usage fees, subscription fees, asset sales, renting, lending, leasing and licensing (Osterwalder & Pigneur, 2010).

Key Resources

20 (Osterwalder & Pigneur, 2010). In other words, this may revolve around the technology, products, facilities, equipment and brand necessary to deliver the value proposition to the intended customer segment (Johnson et al., 2008).

Key Activities

Key activities can be described as the activities and processes necessary for leveraging the key resources to enable customer value delivery (Johnson et al., 2008). This may include production, problem solving and/or networking activities (Osterwalder and Pigneur, 2010).

Key Partnerships

There are numerous partnerships a firm can engage in to facilitate a specific business model, for instance strategic alliances with non-competing actors, joint ventures, buyer-supplier partnerships, and coopetition (alliances between competing actors). Alliances and partnerships may be formed for several reasons, for example to enable economies of scale, for risk reduction and/or to gain access to particular resources and activities (Osterwalder & Pigneur, 2010).

Cost Structure

A certain business models may be more or less sensitive to cost, by the means of how value is achieved. The business model may be cost-driven, focusing on low cost as an advantage, whereas value-driven business models are less concerned with cost and rather focusing on enhanced value creation. Cost structures may be of several types such as economies of scale, economies of scope, variable costs or fixed costs (Osterwalder & Pigneur, 2010).

2.3 Three Aspects of Retailing

Retail Innovation

To illuminate the nature of innovation in retail, the attention is followingly turned to explore its defining characteristics. Hristov & Reynolds (2015) describe retail innovation as a concept close to the entrepreneurial meaning of innovation, as to allocate existing resources for novel use. With this in mind, retail innovation can still encompass both process and technological innovation (Reynolds et al., 2013).

21 Innovation in retail is often the result of deliberate actions, retailers do however not necessarily acknowledge these as deliberate innovation initiatives (Reynolds et al., 2007). These may for instance be referred to as experimentation, continuous improvements or refinements (Hristov and Reynolds, 2015). Generally, retail innovation is fairly easy to replicate for competitors. This has two implications, increasing the importance of short time to market (Reynolds et al., 2007), and retail innovation may not be a source of sustainable competitive advantage (Sorescu et al., 2011).

Retail innovation has in summary been used to describe how retailers innovate. This warrants an additional discussion of what to innovate. Innovation efforts is often directed to the change and development of retail formats (Botschen and Wegerer, 2017; Reynolds et al., 2007), a concept which is followingly addressed.

Retail Formats

A retail format can be described as a set of retail salespoints with identical or nearly identical offerings. These generally have comparable promotion strategies, pricing, assortment and location character (Fox and Sethuraman, 2006). There is a strong connection between business models and retail formats, where Reynolds et al. (2007) describe a retail format as the physical manifestation of the business model. The development of a certain retail format may be a way of leveraging core competencies. It can also be used as a tool to target markets with weaker competition, when the dissemination of certain retail formats is uneven (Ahlert et al., 2006). Furthermore, the location choice for salespoints have been suggested to heavily influence the success of retailers (Lovreta et al., 2013; Mendes and Themido, 2004). The factors determining a suitable location choice depend on the retail format in question (Davies and Clarke, 1994). There are some differing opinions on what types of retail salespoints the format concept encompasses. The interpretation of retail formats by Ahlert, Blut & Evanschitzky (2010), includes a rather narrow selection of retail salespoints, mainly focused on different types of stores. Others have used the concept to describe a broader set of salespoints, for instance by the inclusion of fuel stations (Berné-Manero et al., 2009; Davies and Clarke, 1994).

Several distinctions have been made when using the retail format concept to describe fuel stations. Berné-Manero, De la Fuente-Mella & Marzo-Navarro (2009) described convenience stores located in fuel stations as a distinct retail format. They furthermore differentiated common or independent management of conjointly located fuel stations and convenience stores. Borenstein & Bushnell (2005) suggested a refined categorization of stakeholder relationships in fuel stations. This categorization was referred to as the vertical structure, which can be described as the organization of stakeholders along the fuel station value chain. This organization can be differentiated on three main aspects, station branding, station ownership and station management/operation.

22 is the development of the product/service assortment (Botschen and Wegerer, 2017), to which the focus is now turned.

Green Product Assortment

One approach for discussing retail branding is by dividing the topic into five dimensions. These are (1) access, (2) in-store atmosphere, (3) price and promotion, (4) cross-category product/service assortment and (5) within-category brand/item assortment. Following the dimension of cross-category product/service assortment, a broader assortment increases the possibilities to meet the customer needs. It may also increase the number of times the customer recollects and considers the retailer (Ailawadi and Keller, 2004). On the downside, a broader assortment increases costs (Ailawadi and Keller, 2004), and an extension of the category range may be difficult when the brand is being seen as prototypical for a given category (Farquhar & Herr, 1993 cited in Ailawadi & Keller, 2004).

There are however additional aspects to how expanding the product/service assortment may influence brand goodwill. The introduction of environmentally sustainable products to the assortment is one such aspect (Danciu, 2015). These products may be referred to as green, suggesting environmentally friendly attributes (Sarkar, 2012). There is however no stringent definition to what products qualify as green (Hartmann and Iba, 2006). The term have furthermore been used in a relative manner, when a product has environmental benefits compared to the conventional alternative (Stephen & Kane, 1996 cited in Sarkar, 2012).

One keynote on brand goodwill, is the influence of customer perception. Thus, how the customer perceives the offered products are central when discussing brand goodwill. This also applies in the opposite direction, where the brand perception influences the offering attractiveness to customers. In this context, its desirable to have a brand that is seen as green by customers. Customers are also inclined to choose products marketed under green brands even in the cases where the product in question is not more environmentally friendly than the alternatives (Sarkar, 2012).

However, attempting to create green associations with a brand does not come without risk. Companies engaging in green marketing of products and services that are not environmentally friendly enough in the eye of the customer can be accused of greenwashing. This can have a detrimental impact on brand goodwill. Another case when the same effect may occur even when the attempts are legitimately green and fruitful, are when companies in industries with a heritage of high pollution engage in such activities. These companies may for example reside in chemical, petroleum or mining industries. Paradoxically, they receive criticism of either not engaging enough and when engaging, they risk being accused of greenwashing (Ottman, 2011).

23

2.4 Theoretical Framework

Framework Outline

As described in the Introduction, the purpose of this study is to explore the main challenges associated with the provision of fast charging in Swedish fuel stations. To fulfill the purpose, three theoretical perspectives are deemed necessary for the sensemaking of the empirics. These correspond to the theory presented in 2.1 Technological Change, 2.2 Business Models and 2.3 Three Aspects of Retailing. This multifaceted analysis approach reflects on the abductive approach of the study, with several perspectives chosen to account for the disparate nature of the empirical findings deemed relevant to answer the MRQ. In addition to the reference use of the presented literature, two distinct frameworks are applied to analyze and visualize chosen parts of the empirics.

Proposed Retail Format Framework

24

Retail Format Definition (Reynolds et al., 2007)

Format Type

Vertical Structure

(Borenstein and Bushnell, 2005)

Downstream

Relationship

Business Model Canvas

(Osterwalder and Pigneur, 2010)

Value Proposition Key Resources Customer Segments Financial Viability

Table 1. Retail Format Framework.

The framework is used to describe the established fuel retail formats in Sweden in section 5.2. The attained format descriptions are used to demonstrate the influence of fuel retail formats on fast charging provision.

Value Proposition, Value Creation & Value Capture

Understanding how the role of fuel stations may be altered when fuel retailers embrace electric vehicle charging calls for a perspective on how the business changes. Drawing from the perception of the business model as an unit of analysis for capturing business opportunities (Amit and Zott, 2001), this approach holds potential for capturing the role of the fuel station in a changing business environment.

Inspired by an aggregated business model approach used by e.g. Tongur & Engwall (2014) and Chesbrough & Rosenbloom (2002), the proposed framework is based around the analysis of the value proposition, value creation and value capture. In line with Lindblad & Nygårds (2018), the aspects of value creation and value capture are expressed as an aggregated representation of the business model canvas components (Osterwalder and Pigneur, 2010). This is done with the purpose of linking the presented literature to the framework. The framework can be described as follows.

- Value proposition, the set of products and services offered to customers.

- Value creation, how value is created, as an aggregated description of key partners, key processes, key resources, and cost structure.

- Value capture, how the company retains value, as an aggregated description of customer relationships, customer segments, channels, and revenue streams.

25

3 Methodology

3.1 Research Design

The research design constitutes the logical sequence linking the empirical data to the research questions and in extension, to the conclusions (Yin, 1994). The phenomenon under attention is present in a setting of complex interdependencies and interactions between many actors, influenced by e.g. policy, technology, and market preferences, suggesting the importance of context. Furthermore, it is highly relevant to describe the contemporary setting as the interplay of factors such as technology, policy and market structure evolve and shift over time. According to Yin (1994), research questions formulated with why or how, focusing on contemporary issues where behavioral events can´t be controlled suggest that a case study may be an appropriate research strategy. The exploratory purpose underlines the quest of discovering new perspectives, aspects, and dimensions of the problem at hand.

A case study design enables a more detailed descriptions of complex settings and the capture of new dimensions of an issue, making it suitable for exploratory purposes (Blomkvist & Hallin, 2014). Following the given motivations, an exploratory case study was chosen. The research design relies on the use of qualitative data, used in accordance with the distinction from qualitative research by Eisenhardt & Graebner (2007). In the events of unexpected empirical findings and new theoretical insights during the course of the study, the theoretical framework was adjusted accordingly, which aligns with an abductive approach (Dubois and Gadde, 2002). The purpose and research questions were revised on several occasions to align the study with the discoveries made in the gathered empirical data and in the literature review.

According to Yin (1994), the case study approach can be applied for industry analysis. Following the purpose and the delimitation, the focus is devoted to fuel retailers in Sweden. However, a one-sided focus on the initiatives and actions of fuel retailers would likely only enable a partial view of relevant contextual developments. To mitigate this risk, the organization categories of government agencies, charging operators, electric utility companies and industry interest organizations were included. These were chosen based on the fulfillment of either one of the following criteria, specialist knowledge in industry policy, specialist knowledge in industry structure and products, or being either partners or competitors in the vehicle electricity market. As the fuel retailer actor role depicted in this study represents an industry of multiple firms, industry interest organizations of fuel retailers were included to facilitate an industry perspective not directly influenced by particularities of individual firms.

3.2 Pre-Study

26 familiar with the problem context (Collis and Hussey, 2014). Subsequently, allowing refinements of the purpose and research questions. The pre-study included one meeting with two investigators at The Swedish Transportation Administration, one unstructured phone interview with one investigator of stationary charging infrastructure at the Swedish Transport Administration and one additional unstructured phone interview with an energy market analyst at the consulting firm Sweco, author of a Master´s thesis on charging infrastructure. Additionally, two events were attended by the author, one industry seminar on charging infrastructure organized by the Swedish Transport Administration and the electric car fair eCarExpo Stockholm 2020.

3.3 Literature Review

The literature review has been conducted with two primary purposes, to identify and gain an increased understanding of relevant research fields and for facilitating the development of a theoretical framework. The literature review was conducted in line with an abductive approach (Dubois and Gadde, 2002), as described in section 3.1. This approach allowed the discovery of new relevant areas in literature by unexpected empirical findings, but also in the opposite manner by gaining new understanding of the findings with the theoretical lenses enabled by literature. The literature review was not limited to a specific time frame but was conducted in parallel with other activities in this study.

The literature search was primarily based on the use of the publications made available by the libraries and search tools of KTHB and Google Scholar. Books found through the specified search engines, not available online but in physical form in the KTH library, were retrieved there. The main search areas were fuel retailing, retail innovation, technological change, business models, strategic reactions to substitutes, charging infrastructure and electric vehicle adoption. The search findings were in a first stage evaluated based on the titles, findings with relevant titles were evaluated based on their abstracts. Findings deemed relevant based on the abstracts were reviewed in full length. In publications of interest, the reference list was consulted for further readings. No limitations where made based on publication year, with the exception of publications specifically concerning recent developments in vehicle electrification, only reviewing those published in 2016 or later.

3.4 Collection Procedure

Interviews

27 number of key informants were deemed to scarce to be suitable for data gathering by questionnaires, and the access to key informants was assessed to be higher with the use of interviews. This subsequently motivated the choice of primarily relying on interviews.

The interviews were based on a semi-structured format, utilizing open-ended questions to facilitate more elaborate responses. The choice of the format was also motivated by the possibility of adding additional question (Collis and Hussey, 2014), and allowing freedom to pursue unanticipated topic aspects. Asking additional questions, also called probing, is one way of adjusting data collection instruments, which allow the embracement of special opportunities (Eisenhardt, 1989).

The interview guide (see Appendix 1 for the basic template) was not standardized but tailored to the organization type in question and continuously adapted throughout the interview process. This mirrors the exploratory approach of the study and the multiple revisions of the research questions. According to Eisenhardt (1989), alterations to the data collection procedure during the study is legitimate to increase the depth of understanding, which can be facilitated by newly discovered opportunities and subsequently require alterations to the collection procedure. The interview guide was not presented prior to the interviews, except for Interview (6) and Interview (7), on the direct request of the respondents (see Table 2 for interviews). The interviews were conducted in person, by telephone or Skype, by the preference of the informant. All interviews were sound recorded, with the author taking supporting notes. All interviews were conducted with only the author and the informant present, except for Interview (7), where two informants were interviewed together.

The purpose of the interviews was to identify and gain an understanding of the benefits, challenges, and outlooks on providing charging for electric vehicles. Efforts were also directed to gather information to facilitate a mapping of the Swedish fuel retailing industry. The study is conducted from an industry outsider perspective, which limits the access to internal, company sensitive information. This issue could to some extent be mitigated by cross-referencing data between actors. However, the outsider perspective comes with the risk of missing out on possibly important aspects of the issue due to reluctance in information disclosure by informants.

28 Organization

Category Number Organization Position

Duration

(min) Type Date Fuel Retailer 1 OKQ8 Product Manager 53 In person 200227 Fuel Retailer 2 St1 Retail Manager 99 In person 200302 Fuel Retailer 3 Circle K Product Manager 55 In person 200304 Fuel Retailer 4 Circle K Senior Manager 41 Skype 200320 Fuel Retailer 5 Preem Senior Manager 21 Telephone 200331 Electric Utility Company 6 Vattenfall Manager Charging Infrastructure 27 In person 200225 Electric Utility Company 7 E.ON Manager Charging Infrastructure & Business Developer 45 Skype 200225 Charging

Operator 8 Bee Market Manager 42 Telephone 200316

Industry Interest

Org. 9

Svensk

Bensinhandel Senior Manager 49 In person 200221 Industry Interest

Org. 10 SPBI Product Specialist 39 In person 200311

Industry Interest

Org. 11 Power Circle eMobility Specialist 48 Telephone 200313

Authority 12 The Swedish Energy Agency Charging Infrastructure Specialist 73 Telephone 200214 Authority 13 Svenska Kraftnät Manager Grid Development 21 Telephone 200318

Table 2. Table of conducted interviews.

29 Vattenfall and E.ON. One additional major actor was deemed to fit the criteria, but access could not be granted for an interview. Bee was chosen for the perspective of a charging operator due to their partnership for fast charging provision to fuel stations.

Three industry interest organizations were chosen, Svenska Petroleum och Biodrivmedel Institutet (SPBI), with the motivation of industry expert knowledge and for their representation of all four chosen fuel retailers. Svensk Bensinhandel was included for their specialist knowledge in fuel retailing industry structure and for the representation of individual and franchise fuel stations operators. The inclusion of Power Circle was deemed relevant to understand the role of fuel retailing from an electric vehicle interest perspective.

The majority of informant choices were made by snowball sampling (Collis and Hussey, 2014), by asking key informants for references to additional informants fulfilling certain criteria. The sequence was initiated through the insights and contacts provided by The Swedish Transport Administration. This posed several benefits by gaining access to contact networks which may not have been accessible by only relying on an outside judgmental approach.

In the case of fuel retailers, the selection criterion was based on the degree of insight and responsibility of either vehicle energy sales (including both fuel and electricity) or retail operations, prioritizing central management positions. This included informants with the roles of product manager, retail manager and two senior managers with direct responsibilities for vehicle energy sales. The informant from The Swedish Energy Agency was chosen based on expertise in the area of vehicle electrification and parallel insights to the policy developments regarding vehicle fuel.

For electric utility companies, managers with responsibilities for vehicle charging infrastructure were targeted. In the case of the charging operator, key insights to market developments and cooperation activities were deemed relevant, why a market manager was chosen. In the cases of the interest organizations, the informant choices were based on the perceived relevance of their respective titles, prioritizing seniority in the case of Svensk Bensinhandel, product knowledge in the case of SPBI and eMobility relevance in the case of Power Circle. For Svenska Kraftnät, expertise knowledge in power distribution was desired, where a grid development manager was identified by snowball sampling.

Documents

30

Emails

On two accounts, information has been gathered by email conversation with informants participating in interviews. This served as a complement to the previously gathered information during the respective interviews, see Table 3 below.

Number Informant (see Table 2) Date 1 Corresponding to Interview 3 200430 2 Corresponding to Interview 12 200504

Table 3. Email correspondence with informants.

3.5 Data Analysis

A thematic approach to data analysis was followed when analyzing interviews. According to Blomkvist & Hallin (2014), this is an appropriate approach in cases when the acquired data is suitable for division into categories. All interviews were transcribed, and the responses were categorized for subsequent grouping into themes, as presented in section 4.4. Data from the interviews relevant for describing industry aspects, was also used for that matter. In line with Marshall & Rossman (2011), a theoretical framework is applied for data analysis (see chapter five), in order to connect the research questions to larger theoretical constructs and the empirical findings to larger issues in the research field.

3.6 Research Quality

In the following section, a discussion on research quality is presented. The discussion is based on an approach commonly used for case study applications, by assessment of the quality criteria of construct validity, internal validity, external validity and reliability (Gibbert et al., 2008). The aspect of internal validity is according to Yin (1994) not relevant in exploratory studies, why it is not included in the discussion.

Construct Validity

31 organizations. Additionally, chapter four has been proofread by the informant corresponding to Interview (3).

The issue is however complicated by the differences in market interests by fuel retailers, which raised the importance of mapping out their general differences for subsequent consideration in the analysis. Another concern is the number of informants interviewed in a particular organization, the relative scarcity of informants interviewed in a single organization increases the risk of response bias. However, recognizing this potential weakness must be considered in relation to the limited number of available informants fitting the selection criteria, which provide some justification for even relying on just a single informant in a given organization.

The inclusion of informants from additional organizations (see section 3.4 for motivation), facilitates a broader set of perspectives, subsequently enhancing the construct validity. The discussed choices have been considered in relation to the key points of interview data collection by Eisenhardt & Graebner (2007), which favors the choice of numerous, highly knowledgeable informants with different perspectives on the phenomenon in question, facilitated by origin in different organizations and outside observer positions. Interviews were chosen by higher possibilities of gaining detailed insights and facilitating new problem dimension discoveries. This choice has been made with considerations of interviews being subject to several potential weaknesses such as question bias, response bias, the aspect of respondent recollection, and respondent reflexivity. However, according to Yin (1994), alternative sources of data collection may not rule each other out, but rather fill complementary purposes, subsequently increasing the research quality. Though, this study primarily relies on interviews and documents due to time constraints.

External Validity

32

Reliability

The reliability criteria address the role of sufficient transparency of how the study was conducted, which should allow for replication of the study with equivalent findings. The purpose is to minimize biases and errors of the conducted procedure (Yin, 1994). The area of qualitative research is generally lacking in terms of sufficient transparency to allow replication (Aguinis and Solarino, 2019), where the possibilities for empirical replication are generally poor. The transparency can be enhanced by keeping detailed documentation of the study procedure and a database with the collected material used during the study (Yin, 1994).

Following the transparency criterion framework by Aguinis & Solarino (2019), some aspects are particularly relevant to discuss in the context of this study. The relative importance of informants is to some extent addressed in chapter four, with the traceability of statements to a given interview. Access to interview recordings is not granted to external parties, in agreement with informants to not disclose the recordings, thus lacking transparency in that area. Neither are the names and official titles of the informants disclosed, this decision was made with the consideration of requests of various degrees of anonymity by informants. Furthermore, involving multiple investigators in a study can according to Eisenhardt (1989) both enhance the chances of gaining novel insights during data analysis and increase the reliability of the findings by convergence of multiple investigator opinions. These benefits could not be claimed with a single author conducting this study, thus increasing the risk of bias.

3.7 Ethical Considerations

Ethical considerations of the study have been made in line with the four guiding principles of the Swedish Research Council (Swedish Research Council, 2002). These are, (1) the responsibility to inform the study participants of the study purpose, (2) the given consent of participation, (3) considerations of the confidentiality of study participants and gathered data, and (4) restriction of the information use to the stated purpose.

33

4 The Industry & Findings

In the Introduction, the need for outlining the Swedish fuel retailing industry and to obtain the opinions of informants representing industry actors, were recognized. These aspects are addressed in the following chapter. Firstly, the scene is set with a short description of the oil industry and orientation of fuel retailers along the oil value chain. The attention is then turned to the Swedish fuel retailing industry, providing a general industry description and followingly focusing in on four aspects. Contemporary fast charging infrastructure in Sweden is described to provide the contextual conditions for fast charging provision. Lastly, seven themes identified from interviews with informants are presented.

4.1 A Brief Introduction to the Oil Industry

The Crude Oil Value Chain

Oil industry operations include a wide range of activities. Starting the value chain by obtaining concession rights in order to explore and retrieve hydrocarbons from findings. Concession rights can be granted by states, which is generally done in line with one of three alternative strategies: (1) Granting exploration and subsequently production rights to either public or private companies. (2) Nationalizing the production by giving the exploration and production rights to a state-controlled company or (3) a private company are given the rights to explore and produce on the behalf of the state. The following value chain step is the location and evaluation of possible sources. This is generally done in three steps, seismic surveys and analysis, exploratory drilling, and commercial evaluation. In the event of satisfactory findings, preparations for production are made by equipment installation and establishment of logistics channels (Roslyng Olesen, 2015).

Figure 4. An overview of oil industry value chain steps.

34 product, location and destination. Following the path of crude oil, processing is the next step, involving refining into market fuels, e.g. gasoline, kerosene, gasoil, diesel and fuel oil, and petrochemical products e.g. lubricants, waxes and solvents.

Distribution and sales can be conducted through several channels, where direct sales and retail are common alternatives. Direct sales is done directly to business customers without intermediaries, whereas retail utilize several different types of intermediaries depending on the product, where vehicle fuel is typically sold through fuel stations (Álvarez et al., 2018).

An Overview of Company Types in the Oil Industry

In the classification of oil companies, we have four major categories: (1) National Oil Companies (NOCs), which are state controlled companies with the majority of their investment base in upstream activities in their home countries. These may also to some extent engage in downstream activities. Two examples are Saudi Aramco and Qatar Petroleum. (2) National oil companies with large upstream investments outside of the domestic market (INOCs), e.g. Equinor, the China National Petroleum Company and Gazprom. (3) International Oil Companies (IOCs), which are large integrated and stock market listed companies, for instance BP, ExxonMobil and Royal Dutch Shell. (4) Independents, companies which are either fully integrated but smaller in size than IOCs, or independent upstream actors. Lukoil and Repsol are examples of independents (IEA, 2020). In addition to the four major categories, we have another three types: (1) Service companies, which engage in supporting activities such as engineering services for drilling, construction of infrastructure and reservoir management. (2) Trading companies, which engage in physical trading of oil. These may in some cases also invest in refining, distribution, transport and storage, but then with the main motivation of strengthening their market position. (3) Pure downstream companies, which operate refineries and retail networks (IEA, 2020). In this downstream category, there are Swedish based companies such as Nynas, which refines crude oil to produce naphthenic specialty products and bitumen, which can be used to make e.g. adhesives, rubber, paint and for paving applications. (Nynas, n.d.). Two additional examples are Preem, with refining operations, direct sales and retail sales though fuel stations (Preem, 2019), and Circle K, with retail sales though fuel stations (Circle K, 2019).

4.2 Fuel Retailing in Sweden

35 Fuel stations can be of different types, generally classified based on the offerings at a given station (SPBI, 2019b). The fuels offered can be of conventional types with varying degrees of renewable admixture, or of renewable types (Interview 10). There are multiple fuel station brands in Sweden and one company can have stations under more than one brand. Stations may also have different vertical structures, meaning that stakeholders can have different roles in one station compared to another. For example, the fuel retailer can both own and operate a station in one case. In another case, the station may be owned and operated by a dealer, even under the same brand in the same network (Interview 9).

In the following sections the focus is turned to four aspects of the industry, namely vertical integration, customer segments, fuel stations, and vertical structures. When applicable, these aspects are described regarding the four major actors Circle K, OKQ8, St1 and Preem. These aspects are presented as means of differentiation between the companies and are used as a basis for the subsequent discussion in chapter five, primarily for the retail format mapping.

Vertical Integration

The four companies under attention are to different extent vertically integrated. This description serves to outline the differences between the companies and the subsequent implications are further addressed in chapter five.

All four companies have fuel station networks for retail sales, in line with the prerequisites to be included in the mapping. However, St1 and Preem are to a higher extent vertically integrated by their refinery operations than Circle K and OKQ8. Thus, St1 and Preem are also fuel producers (Interview 2). The number of sales channels used also differ between companies. The domestic sales channels can be divided into direct sales and retail. Both channels are used by St1 (St1, 2019), Preem (Preem, 2020a) and OKQ8 (OKQ8, 2019), whereas Circle K only have retail (Circle K, 2019). Thus, the range of operations and sales channels varies between the companies.

Customer Segments

To define and distinguish the needs of different types of customers, the attention is turned to customer segments. One general division of customer segments can be made into B2C Cars & LDVs, B2B Cars & LDVs and B2B HDVs (Interview 4). Comparing the first two segments, there may be some differing characteristics in how one chooses a fuel station for a stop. Generally, the consumer has a greater freedom of choice and can choose a station to the personal liking. Whereas for B2B customers, the company agreement with a fuel retailer will generally regulate which station brand the vehicle user will utilize. Customers agreements in the B2B HDV segment will often more strictly regulate where refueling is made (Interview 2).

36 fuel choice and evaluation of vehicle fleet emissions and influence on vehicle Total Cost of Ownership (TCO). This stands in contrast to the consumer relationship, which largely follows the characteristics of mass marketing, with a sales approach inclined towards discount incentives and members clubs to enhance loyalty and sales (Interview 4).

Fuel Stations

Fuel stations come in several different types, which includes variations in the range of offerings and the intended customer segments to serve. Two different views on the categorization of fuel station types have been identified.

Firstly, SPBI proposed a categorization, dividing stations into four types. Namely, (1) full-service station, a lager establishment offering vehicle fuel (with several fuel dispensers), convenience store and vehicle related services. (2) Filling station, offering vehicle fuel (with several fuel dispensers) and a convenience store establishment. (3) Self-service station, an unmanned sales point utilizing electronic or vending machine payment, not located in direct connection to a convenience store, a garage or similar. (4) Single facility, single fuel dispensers located in connection to other establishments, such as a convenience store or a garage (SPBI, 2019b). The second categorization builds on the division of station types into three categories, full-service stations, self-service stations and truck stations. Here, full-service and filling stations are conjointly referred to as full-service stations, and truck stations are self-service stations tailored to HDVs (Interview 3; Interview 9). With currently no single facilities present in Sweden, this type is excluded from the categorization (SPBI, 2019b). This second categorization is used as a basis for subsequent discussions on fuel station types.

37

Company Circle K St1 OKQ8 Preem*

Full-Service Stations 292 (C ir cle K, 202 0) 104 (Int er vie w 2) 299 (OK Q8, 2020) 99 (Pr ee m, 2020 b) Fast Charging 47 0 36 21 Self-Service Stations 291 294 (St1, 202 0) 205 - Fast Charging 1 (Em ail 1) 0 (Int_er vie w 2) 0 - Truck Stations 168 (C ir cle K, 2020) 89 35 -

Table 4. Outline of fuel station division by major fuel retailers in Sweden. Table figures correspond to the number of stations with the specified offering. *Preem has a total of 570 fuel stations (Preem, 2020a), the shares of self-service and truck stations could not be distinguished. See Table 3 for emails.

Vertical Structures

The management/operation, ownership and branding of fuel stations can be made in line with a number of different sourcing strategies, which can be referred to as the vertical structure of a station. As introduced in section 2.3, the concept of vertical structures can be described as the organization of stakeholders along the fuel station value chain, which can be differentiated on the points of fuel station management/operation, ownership, and branding (Borenstein & Bushnell, 2005; Interview 9). To distinguish the involved stakeholders, the following definitions are introduced:

- Fuel Retailer, company selling and/or distributing fuel through fuel stations with license rights to at least one brand.

- Dealer – Person or company operating one or more fuel stations under either a franchise or a distribution agreement with a fuel retailer.

In line with the description provided in Interview (9), the general vertical structures present in the Swedish market are presented in Table 5.

38 Operator Ownership Branding

Corporate Owned and Corporate

Operated (COCO) Fuel Retailer Fuel Retailer Fuel Retailer Corporate Owned and Dealer

Operated (CODO) Dealer Fuel Retailer Fuel Retailer Dealer Owned and Dealer

Operated (DODO) Dealer Dealer Fuel Retailer

Table 5. Outline of general vertical structures of fuel stations in Sweden.

The stakeholder relationships are normally defined by contractual agreements. The CODO and DODO structures are commonly regulated by either a franchise or a distribution agreement. There are a variety of special cases regarding vertical structures (Interview 9). However, in this study, subsequent discussions are based around the COCO, CODO and DODO structures.

In general, it can be said that choosing an appropriate vertical structure may pose benefits of increasing operational and transaction efficiency (Borenstein and Bushnell, 2005). There are several aspects influencing the appropriateness of a certain structure in a given case. For the COCO structure, there is no need for negotiations with dealers of how changes to the fuel station are to be managed. This can be beneficial when engaging in concept testing, for instance when making changes to the offered product range at a station. The COCO format is however capital intensive (Interview 9).

Choosing other structures is largely dependent on the issue of agency. To what extent the effort of managers can be observed influence the choice, with higher observability favoring higher control as enabled by the CODO structure. Lower observability on the other hand, may favor the choice of the DODO, giving the dealer larger management freedom. The role of cost and investment also influence the choice, as the establishment of DODO stations may be a relatively inexpensive way to increase the network size. This may in turn be beneficial in the sense of raising barriers to entry (Serebrisky, 2003). In high traffic areas, which are often urban areas, COCO stations are generally prioritized. Followingly, CODO stations are next in line which may be chosen in areas where COCO stations cannot be directly justified, whereas DODO stations are more common in remote locations (Interview 9).

The majority of the fuel stations included in the station networks of Circle K, OKQ8, St1 and Preem, can be classified according to COCO, CODO and DODO. The truck stations and self-service stations are commonly established in line with the COCO structure. Full-service stations, however, generally vary between COCO, CODO and DODO (see Appendix 3 for further descriptions).

4.3 Fast Charging Infrastructure