Life Cycle Assessment in the Automotive Industry

(1)

Life Cycle Assessment in the Automotive Industry

Considerations for First-Tier Suppliers

ALY IBRAHEM NILS SJÖQVIST

KTH ROYAL INSTITUTE OF TECHNOLOGY

SCHOOL OF INDUSTRIAL ENGINEERING AND MANAGEMENT

(2)

(3)

Automotive Industry

Considerations for First-Tier Suppliers

by

Aly Ibrahem Nils Sjöqvist

Master’s Degree Project KTH Royal Institute of Technology

School of Industrial Engineering and Management

Department of Sustainable Production Development

(4)

Considerations for First-Tier Suppliers

Aly Ibrahem Nils Sjöqvist

Approved 2021-June-17

Examiner KTH Andreas Archenti

Supervisor KTH Seyoum E. Birkie Commissioner

Veoneer Sweden AB

Contact person at company Karin Käck

Abstract

Climate change is increasingly gaining the attention of governments, companies, and the general public. Many original equipment manufacturers (OEMs) have set their own milestones to achieve carbon neutrality, as early as 2039 (Daimler, 2018). This presents a competitive opportunity for first-tier suppliers to aid the OEMs with their targets. Life cycle assessment (LCA) is proposed as an approach that provides visibility into the potential life cycle impacts of a product. LCA methodology as defined and described by ISO 14040:2006 and ISO 14044:2006 is general. Therefore, methodological choices specific to the application need to be considered and defined.

This thesis attempted to address this issue for first-tier suppliers in the automotive industry by compiling the current practice in industry and academia. A literature review was conducted, LCA reports from OEMs were studied, standards on LCA were consulted, and two OEMs and an organisation representing suppliers were interviewed.

The findings from these methods were compiled and discussed. From the discussions, key recommendations were made on how a first-tier supplier might conduct an LCA study, following the methodology from ISO 14040:2006 and ISO 14044:2006.

(5)

Övervägningar för förstahandsleverantörer

Aly Ibrahem Nils Sjöqvist

Godkänt 2021-Juni-17

Examinator KTH Andreas Archenti

Handledare KTH Seyoum E. Birkie Uppdragsgivare

Veoneer Sweden AB

Företagskontakt/handledare Karin Käck

Sammanfattning

Klimatförändring får alltmer uppmärksamhet från regeringar, företag och allmänheten.

Många biltillverkande företag har satt egna milstolpar för att uppnå neutralt koldioxidutsläpp, så tidigt som 2039 (Daimler, 2018). Detta presenterar konkurrensmöjligheter för förstahandsleverantörer till dessa stora företag att hjälpa dem uppnå deras milstolpar. Livscykelanalys (LCA) föreslås som ett tillvägagångssätt som tillhandahåller synlighet i den potentiella livscykelpåverkan av en produkt.

Livscykelanalysens metodik som definieras och beskrivs i ISO 14040:2006 och ISO 14044:2006 är allmän. Därför behövs det definieras metodologiska val som är specifika till tillämpningen.

Denna avhandling har försökt att adressera denna fråga för förstahandsleverantörer inom fordonsindustrin genom att sammanställa nuvarande praxis i industrin och forskning. En litteraturstudie genomfördes, LCA rapporter studerades, standarder för LCA konsulterades, och två biltillverkare samt en organisation som representerar leverantörer intervjuades.

Resultaten från dessa metoder sammanställdes och diskuterades. Från diskussionerna drogs viktiga rekommendationer om hur en förstahandsleverantör skulle kunna utför en LCA studie, genom att följa metodiken från ISO 14040:2006 och ISO 14044:2006.

(6)

(7)

Acknowledgements

This thesis has been a degree project for the master’s programme of Sustainable Production Development at KTH. The project was commissioned by Veoneer Sweden AB and supervised by Veoneer Sweden AB and the HPU department at KTH.

We would like to express our sincere appreciation to our supervisor at Veoneer Sweden AB, Karin Käck, for supporting us every step of the way. The continuous support and guidance are truly appreciated.

We would like to extend our sincere appreciation to our supervisor at KTH, Seyoum E.

Birkie, for always challenging us intellectually to think one step beyond and helping us secure an academically valid thesis.

We would like to thank Cathrine Stjärnekull, Daniel Åhlström, Jennie Viskari, Ola Boström, Pierre Hultstrand and Tobias Aderum at Veoneer Sweden AB for their valuable time and assistance provided to our work. The discussions, insights, and contacts they have provided us with have made all the difference.

We would like to express our gratitude to Gustav Hanberger, Vincent Lingehed, and all other staff at Veoneer Sweden AB that have shown interest in our thesis and provided their support.

We would like to show our gratitude to Erik Postma at CLEPA, Lisa Bolin and Christian Samson at Polestar, and Andrea Egeskog at Volvo for allocating their time to be interviewed.

Without these interviews, the thesis would not be complete.

(8)

(9)

(10)

List of Figures

Figure 1 - Historic Development of the LCA framework (Curran, 2015) ... 5

Figure 2 - LCA Framework according to ISO 14040:2006 ... 7

Figure 3 - Thesis Research Design Diagram ... 11

Figure 4 - Publications per year for papers included in the literature review ... 13

Figure 5 - Example of a Process Flow Diagram by Silva et al. (2018) ... 18

Figure 6 - ReCiPe 2016 Methodology (Huijbregts et al., 2017) ... 22

Figure 7 - Example Diagram of LCA Data Sources (Polestar, 2020) ...28

Abbreviations

EPD Environmental Product Declaration GHG Greenhouse Gas

GWP Global Warming Potential LCA Life Cycle Assessment LCI Life Cycle Inventory

LCIA Life Cycle Impact Assessment OEM Original Equipment Manufacturer PEF Product Environmental Footprint

(13)

(14)

1 Introduction

This chapter presents the scope of this thesis by addressing the background, purpose, objectives, delimitations, and report outline.

1.1 Background

In 2015 the Paris agreement laid the foundation for global collaboration to limit global warming to less than 2 °C. In December of 2019, the European Commission presented the European Green Deal roadmap, which sets targets for how much emissions should be cut down. By 2050 the EU aims to be emissions net neutral (European Commission, n.d.).

Not only the EU, but also the general public is increasingly more interested in carbon neutrality and other environmental efforts. The Greenhouse Gas Protocol (2011) states that as impacts from climate change become more frequent and prominent that governments are expected to set new policies and provide additional market-based incentives to drive significant reductions in emissions. These new policies and market drivers are argued to direct economic growth on a low-carbon trajectory. This means, according to Arena et al.

(2013) that carmakers can no longer treat sustainability as a matter of compliance. They argue that instead that carmakers must increasingly look to environmental sustainability as an opportunity to gain competitive advantage.

Following the targets of the EU, many of the largest automotive original equipment manufacturers (OEMs) are committed to achieving carbon neutrality as early as 2039 (Daimler, 2018). To enable this, product manufacturers need to have insight into their products’ life cycle impacts. The Greenhouse Gas Protocol (2011) states that not long-ago, companies have focused their attention on emissions from their own operations. However, they conclude that companies increasingly understand the need for also accounting the greenhouse gas (GHG) emissions along the value chain and product portfolios in order to comprehensively manage GHG emissions related risks and opportunities.

This creates a scenario of opportunity for first-tier suppliers to gain competitiveness by aiding the OEMs in achieving their commitments. This could be done by securing accurate visibility of environmental impacts of the value chain and evaluating the environmental impact of first-tier suppliers’ products.

Thus, steps need to be taken towards incorporating frameworks for applying tools that assess life cycle environmental impacts. Life cycle assessment (LCA) is regarded as a holistic

(15)

approach to investigate the impacts (of e.g., a product) over all stages of a life cycle. ISO provides a well-established generic framework for conducting LCA. The ISO 14040:2006 Environmental management – Life cycle assessment – Principles and framework standard provides principles and a framework for LCA methodology. While the ISO 14044:2006 Environmental management – Life cycle assessment – Requirements and guidelines standard provides requirements and guidelines for conducting LCA. The ISO methodology is general but actual implementation of the methodology needs to be specific to the context of its application.

This implies that different sectors and different links in a value chain need to develop their own customised approach to LCA. Doing this, helps a company to make more informed decisions for decreasing overall life cycle environmental impacts. Many companies have limited knowledge, skills, and resources required to develop a context specific LCA approach, which limits their ability to enhance (product) environmental performance.

Current studies do not address in detail how parts/sub-assembly suppliers to automotive OEMs could implement LCA, in order to contribute to more accurate LCAs for the OEMs’

vehicles. This means that a context specific interpretation of LCA methodology needs to be investigated for first-tier suppliers to automotive OEMs.

1.2 Purpose and Objectives

The purpose of this thesis is to investigate what considerations need to be made for a customised attributional LCA approach for a first-tier supplier to automotive OEMs. Such an approach should allow for scalability and be compatible with established LCA approaches of the automotive OEMs. In order to fulfil this purpose, the objectives of this study are defined as:

1. To explore and investigate prevailing considerations of existing LCA approaches applied to the automotive sector.

2. To discuss the prevailing considerations of LCA approaches in the automotive industry and conclude suitable recommendations to aid future formulation of a customised implementation of LCA for first-tier suppliers.

1.3 Delimitations

Keeping in mind the time limitations and field of expertise of the authors of this thesis, the study defines the following delimitations. Firstly, the thesis is not intended to deeply investigate data science aspects of LCA, as this is out of the field of expertise of the authors

(16)

and the department of which this thesis work is written at. Secondly, the intention of the thesis is not to perform a comprehensive LCA, but rather investigate considerations of LCA application within the proposed context. Thirdly, the thesis is not to investigate approaches and considerations for consequential LCA, only attributional LCA. The difference between these two is briefly explained in section 2.3.

Finally, the thesis bases its findings around the methodology of ISO 14040:2006 and ISO 14044:2006. It should be noted that these standards were, at the time of writing, published 15 years ago. Additionally, the GHG protocol (2011) standard was also explored in this thesis and was, at the time of writing, published 10 years ago. These standards could be regarded as becoming old, this should be kept in mind.

1.4 Report Structure Outline

The rest of this report is structured as follows. Chapter 2 makes up the theoretical framework, which introduces and explains some key concepts that serve as background information for understanding this thesis report. Chapter 3 describes the methodology that was used for obtaining the results. The results are presented in chapter 4, structured by LCA methodology topics as defined in ISO 14040:2006. Chapter 5 discusses the results. Chapter 6 concludes the study by making key recommendations, acknowledging the limitations of this study, and addresses possible future research.

(17)

(18)

2 Theoretical Framework

This chapter consists of concept definitions that make up the existing theories that support this thesis.

2.1 Development of LCA Methodology

Strategic environmental approaches have developed through the development of laws and regulations aiming to reduce pollution. Initial approaches focused on End-Of-Pipe solutions and waste minimisation. These measures do not address the potential pollution caused during all the lifecycle stages or end-of-life for the product system. Therefore, a more holistic approach was needed to cover all the potential produced pollutants along a product's overall life cycle stages.

The term LCA was first coined in 1990 in the US in a workshop held by the Society of Environmental Toxicology and Chemistry (SETAC) (Curran, 2015). The first LCA framework proposed by SETAC consisted of three interrelated components, namely; Inventory, Impact analysis and Improvement analysis. Later, the fourth component, the definition of Goal and Scoping was added. ISO published their first standard in 1997, namely ISO 14040:1997, describing Environmental management - Life cycle assessment - Principles and framework.

The framework took its current shape when it was updated in 2006 (see Figure 1).

Figure 1 - Historic Development of the LCA framework (Curran, 2015)

2.2 Purpose of LCA

LCA can be seen as an approach for assessing any potential impact associated with a product life cycle. This approach requires quantification of all inputs and outputs of material, energy and emissions throughout the product life stages. This inventory of inputs and outputs can then be translated into impacts.

(19)

Typically, only environmental aspects are covered in LCA practise, leaving out economic and social aspects of the product life cycle (ISO14040:2006).

2.3 Attributional versus Consequential LCA

There are two types of LCA approaches recognised in current methodology. Namely, the attributional approach and the consequential approach.

Attributional LCA (ALCA) investigates the relationships between the physical flows from and to the product or process (Curran, 2015), i.e., directly attributable to the product or process. ISO 14040:2006 defines this approach as assigning elementary flows and potential environmental impacts to a specific product system. Sometimes this is referred to as an accounting LCA. A typical example of an attributional LCA could be the study by Silva et al.

(2018), where they account for the environmental impact of an engine valve.

Consequential LCA (CLCA) aims at describing the effect of changes within the life cycle, given that changes lead to a series of consequences through chains of cause-effect relationships (Curran, 2015). ISO 14040:2006 defines the consequential LCA approach as studying the environmental consequences of possible (future) changes between alternative product systems. Curran (2015) states that consequential LCAs are more applicable on industrial operations of larger scales on the regional and national levels. An example of a consequential LCA is one presented by Palazzo et al. (2019), they performed a study investigating the environmental effects of replacing steel with aluminium in production of vehicles in the North American industry.

Deciding on an attributional approach versus a consequential approach might result in different conclusions. Searchinger (2008) as referenced by Curran (2015) presented a study where attributional LCA results on corn-based ethanol showed a decrease of 20% in GHG emissions compared to gasoline. The consequential LCA on the other hand showed an increase of 47% in GHG emissions compared to gasoline. These additional emissions can be attributed to the predicted increased land-use to meet the increasing demand for corn for ethanol production.

2.4 LCA Methodology According to ISO 14040:2006 and ISO 14044:2006 ISO 14040:2006 describes the principles and framework for LCA. The main aspects that the standard covers are:

- Goal and scope definition of the LCA

- The life cycle inventory (LCI) analysis phase - The life cycle impact assessment (LCIA) phase

(20)

- The life cycle interpretation phase - Reporting and critical review of the LCA - Limitations of the LCA

- Relationship between LCA phases

- Conditions for use of value choices and operational elements

ISO14040:2006 describes LCA methodology as 4 phases that are applicable for any type of product or service; Goal and scope definition, Inventory analysis, Impact assessment, and interpretation (see Figure 2).

Figure 2 - LCA Framework according to ISO 14040:2006 Goal and Scope Definition of the LCA

The credibility of LCA outcome relies heavily on a clear and unambiguous goal definition, as the goal will determine the direction, depth, and width of all the next steps in the framework.

In other words, the goal definition defines the LCA study scope. The goal definition states the following components:

- The reason for carrying out of the study.

- The intended application of the results.

- The target audience.

The scope definition describes the following components:

- The studied system.

- The functional unit: a quantified description, of the function or services provided by the system, to which the data will be related.

- Chosen impact categories to be studied.

- Assumptions. These can be made to simplify the study or compensate for lack of data.

- Limitations, as a result of the chosen scope and way of carrying out the study.

(21)

- Data requirements: specifying temporal and special attributes.

- Allocation procedures.

- Type of critical review.

Life Cycle Inventory (LCI) Analysis

LCI involves mapping and quantifying all relevant inputs and outputs of energy, materials flows and emissions. This process requires collecting a considerable amount of data, validation of the data and relating it to the processes. Therefore, simplifications and assumptions can be made to ease the LCA implementation. Documentation of all assumptions is needed to reserve the LCA study transparency and credibility.

Life Cycle Impact Assessment (LCIA)

The mapped inputs and outputs of materials and emissions then can be associated with specific impact categories and translated into a list of potential impacts. The purpose of LCIA is to evaluate the significance of these potential impacts from the LCI results list. One example of an LCIA framework is the ReCiPe method (explained in 4.2.5).

Life Cycle Interpretation Phase

According to ISO 14044:2006 the interpretation of the LCA study aims to identify the significant issues from the LCI and LCIA phases such as energy consumption, emissions, waste, etc. Furthermore, it thoroughly inspects the completeness, sensitivity, and consistency of the results from these phases in relation to the goal and scope definition of the LCA study. A full review of the functional unit, system boundaries and the limitations introduced by data quality should be considered at this stage.

On the topic of a sensitivity analysis. The LCA results can be affected by many sources of uncertainty, such as methodological choices, chosen system boundaries, and assumptions.

To examine the robustness of these results, LCA practitioners can use sensitivity analysis.

According to Pichery (2014), sensitivity analysis is the process of evaluating the effect of one or more input variables on the output variables of a numerical model. In other words, the sensitivity analysis highlights the values of an input variable beyond which, the output would change significantly. In practice the practitioners change the input parameters of the model to their extremes, in order to evaluate the significance of change on the overall results.

Reporting and Critical Review of LCA

Reporting LCA study results should articulate the different life cycle analysis phases, as iteration might be needed to further refine the collected data and scope of study. ISO

(22)

14040:2006 standards recommend that in case the LCA study extended to LCIA or reported to a third party the following aspects should be described in the report:

- A description of data quality.

- Characterization models.

- The chosen impact categories (e.g., GWP, Land use).

- Endpoints to be protected (e.g., human health, ecosystems).

- The indicators result profile.

- The factors and environmental mechanisms.

- The relationship with the LCI results.

2.5 Sustainability and LCA

Sustainability as a concept could be defined as the creation of value without compromising the needs of future generations. The concept of sustainability is composed of three “pillars”.

These are the environmental, social, and economical pillars (sometimes referred to as people, planet, profit).

Being environmentally sustainable could be defines as creating value without destabilising or depleting natural resources and systems. Being socially sustainable could be defined as proactively contributing to the improvement of society; locally, along the supply chain, and for the customer. Being economically sustainable could be defined as creating long-term economic growth without negatively impacting environmental and social aspects.

In order to know if sustainability is compromised on, visibility into impact performance is needed. LCA provides this visibility and gives insight into impact performance. LCA studies, traditionally, have been applied to the environmental pillar of sustainability only (ISO 14040:2006).

(23)

(24)

3 Research Methodology

The thesis research design is visualised in Figure 3. Before carrying out of the main task of the thesis, a short pre-study was performed. The research design of the thesis is divided into three steps: methodological design, findings from carrying out of those methodologies and conclusions on those findings.

The first step, methodological design, covers three different sources of information:

academia through a literature review, findings from industry through interviews and public reports, and findings from standards. The findings from these three sources are triangulated and discussed.

Figure 3 - Thesis Research Design Diagram

3.1 Academic Literature Review

For the first step of the research, a literature review was conducted. The aim of the review is to determine the state of the art of LCA in the automotive industry from the perspective of academia.

In order to find review papers in a systematic manner, the literature review consisted of the following steps: (1) Formulation of review question. (2) Formulation of inclusion and exclusion criteria. (3) Search for literature. (4) Synthesising of the findings in the literature.

Formulation of Review Question

The following review question was defined to guide the literature review: “What LCA tools and approaches are used in, the automotive industry?”. The question has been formulated to the specific context of this thesis, in order to filter any content that is not applicable to the case.

(25)

Formulation of Inclusion and Exclusion Criteria

Both inclusion and exclusion criteria were be based on the title, abstract, and the full text of the article. Interesting citations from the identified literature were included based on the same criteria. The inclusion criteria were:

- The study should mainly concern LCA in any shape or form.

- The publication should be a journal article.

- The LCA study should address the automotive industry.

The exclusion criteria were:

- Any publication that does not fulfil the inclusion criteria will be excluded.

- Books and Conference publications will be excluded.

- publications that are older than 2007 are excluded (ISO 14040:2006 was published in December of 2006).

- Any publication in a language other than English is excluded.

Search for Literature

Based on the inclusion and exclusion criteria, search parameters can be set in the form of a single query. The search was conducted on WebOfScience, from the core collection. The timespan of publications was set to 2007-2021. The search query was set to search for key words in the abstracts. Only English publications were selected. Only articles were selected, any other publications, such as books or conference papers, were excluded.

The query logic was built to search for articles that included three components:

1. Any way of spelling LCA or substitutes of the word (LCA, Life Cycle Management, Cradle to grave, etc.).

2. Automotive or any related word starting with “Automo”, such as Automobile.

3. Any indication of a case or original work, such as Case, Application, Tool, Framework, etc.

The exact original search query was defined as follows: (AB=(("Life Cycle Assessment" OR

"LCA" OR "Life Cycle Management" OR "Cradle to grave" OR "Cradle to gate")AND(Automo*)AND("Case" OR "application" OR "develop*" OR "guideline*" OR

"implement*" OR "tool*" OR "framework" OR "paradigm"))) AND LANGUAGE: (English) AND DOCUMENT TYPES: (Article)

(26)

The search resulted in 145 papers applicable to the search parameters. From this any paper that fit the inclusion and exclusion criteria (from the title) was selected for reading. 92 papers were selected. The next exclusion step was reading the abstract of these papers, through which additional papers were excluded. After this, the remaining papers’ full text were read. Resulting in an additional exclusion of 17 papers. Access could not be gotten to the full text of some of the papers. Additionally, upon reading the full text of each paper, additional papers were omitted. The remaining number of included papers in the study are 41. The full list of papers can be found in Appendix 1. A visualisation of number of publications per year is found in Figure 4. There seems to be an increasing trend in publications on the topic, as defined by the search query. This could indicate an increasing interest in LCA within the automotive industry.

Figure 4 - Publications per year for papers included in the literature review

Separately from this slightly formal literature review, an initial exploratory literature was also performed. Student thesis publications and some journal articles were read to form an understanding of recent academic thesis work, to help scope this thesis and receive an indication of the state-of-the-art on the field of LCA.

Synthesis from Literature

Relevant findings from the literature were synthesised by explicit themes in a spreadsheet.

That spreadsheet served as a basis for the results of the literature review. The themes were product system, system boundaries, functional unit, assumptions, limitations, type and format of the report, impact categories, allocations, allocation procedures, data quality, data collection, data calculations, critical review, and optional steps of LCIA.

(27)

3.2 Industry Review

The industry review consists of interviews with representatives from industry and findings from public reports from automotive OEMs.

Industry Interviews

Three interviews were conducted within the automotive industry. Two of the interviews were with experts within the field of LCA at automotive OEMs Volvo and Polestar. The Polestar interview was represented by Lisa Bolin and Christian Samson, which lasted 60 minutes.

The Volvo interview was represented by Andrea Egeskog, which lasted 30 minutes.

An interview with CLEPA was conducted and was represented by Erik Postma, who is involved with the LCA taskforce of CLEPA. The interview lasted 60 minutes. CLEPA is an organisation that represents suppliers in the automotive industry.

All three interviews were conducted as semi-structured interviews, according to the question list in Appendix 2.

Industry Reports

In addition to the interviews, public reports from automotive OEMs were also consulted.

LCA reports from Volvo (2020), Polestar (2020), and Daimler (2018) were analysed. Annual reports and/or sustainability reports from the respective companies were also consulted for relevant information.

3.3 Standards Review

The third source of information of the research of this thesis comes from ISO 14040:2006, ISO 14044:2006 and additionally from the GHG Protocol (2011). The ISO standards were chosen since they are regarded as common practise LCA methodology. The GHG Protocol’s product life cycle accounting and reporting standard was chosen since both Volvo (2020) and Polestar (2020) mentioned using the standard for methodological considerations.

(28)

4 Results

This chapter is a compilation of the results from the literature review, conducted interviews, industry reports, and consulted standards. The chapter is structured by the different methodological elements of LCA according to ISO 14040:2006.

4.1 Goal

The goal is the very first stage of LCA methodology, together with the scope (chapter 4.2).

According to ISO 14040:2006, the goal definition should include the following aspects: the intended application of the study, the reason(s) for carrying out the study, the intended audience of the study, and whether the results are intended to be used in comparative assertions intended to be disclosed to the public.

It is imperative to define these topics, as they all have implications for how the study will/should be carried out.

Intended Application of LCA Studies

The ISO 14040:2006 lists a couple of examples of applications of LCA methodology: product development and improvement, strategic planning, public policy making, marketing, etc.

But what is LCA being used for in practise?

From the literature review, it seems that academics is focussed on one of the following two applications. The first being environmental performance evaluation of a component or a material. E.g., Ribeiro et al. (2007) who intended to improve the environmental performance of a multi-material car component. Or e.g., Kemp et al. (2020) performed an analysis of energy and emission impacts of a cooperative connected autonomous vehicle.

Another application being a comparative study between different materials or designs to evaluate which one is performing better environmentally. E.g., Das (2014) examined and compared the potential life cycle impacts of two material designs.

One example of an LCA application is hotspot analysis. A hotspot analysis, sometimes referred to as a screening LCA, is an LCA study that focuses on finding the most impactful aspects of a product system. Such a study could look at which product life cycle stage could be the most impactful, or which process within a life cycle stage could be the most impactful.

As the focus is on which aspects are the most impactful, rather than how impactful they are, this allows for using values from databases based on global/industry averages.

(29)

The stated intended applications for the Volvo (2020) and Polestar (2020) LCA reports were to evaluate the carbon footprint of their respective car models. Daimler (2018) defined the intended application to evaluate and compare different vehicles, components, and technologies from an environmental performance point of view.

From the interview with Volvo, it became clear that they will require/push their suppliers to carry out hotspot analysis studies, on environmental performance along the supply chain, followed by sensitivity analysis studies. They are asking for this so that their suppliers will become aware of their environmental impacts and hotspots and start working on reducing their impacts.

Reason(s) for Carrying out LCA Studies

The reason(s) for carrying out an LCA study are at a higher strategic level than the intended application. There can be many reasons for carrying out LCA studies. The Greenhouse Gas Protocol (2011) list a few business goals that can serve as reasons for a company to carry out an LCA study: climate change management, supplier and customer stewardship, etc. These business goals could be made more specific to serve as a reason for carrying out an LCA study, for example: Measure and report GHG performance over time, partner with suppliers or customers to achieve GHG reductions, achieve competitive advantage by pursuing GHG reduction opportunities and cost savings to create a low-emitting product.

From the literature review it could be concluded that academic LCA studies fall into two purpose groups. The first being product development and furthering scientific knowledge in the field. E.g., Gebler et al. (2020), who aimed to provide a base for planning and decisi0n- making regarding decarbonisation. Or e.g., Zhang et al., (2020) who aimed to propose an optimisation method for product design based on LCA and Life Cycle Cost (LCC). The second main reason for carrying out an LCA study was described as contributing to regulations and policy making. E.g., Lehmann et al. (2018) and Palazzo et al. (2019) who presented recommendations for policies and regulations.

For the Volvo (2020) and Polestar (2020) LCA reports, the purpose was to develop a methodology that can be used to produce carbon footprints of their car models. A second motive they mentioned was to be able to use the complete vehicle carbon footprints to examine the effects of changes in e.g.: material composition, efficiency of the vehicle or Polestar manufacturing, or changes in the energy systems. The Daimler report (2018) mentions that their reason is to evaluate the environmental performance of the car by integrating the “design for environment” approach.

(30)

The Intended Audience of LCA Studies

ISO 14040:2006 states that it shall be clearly and unambiguously stated to whom the results of the LCA study are intended to be communicated. From the literature review it can be concluded that academic LCA studies mainly address case companies, the automotive industry, and the scientific community. E.g., Ferreira et al. (2019) and Chen et al. (2020) who are performing case studies in business. Or e.g., Simboli et al. (2015) who present an academic conceptual model on eco-innovations.

For Volvo (2020), Polestar (2020), and Daimler (2018) their LCA reports are intended for both the public and internal use.

Whether the Results are Intended to be Used in Comparative Assertions Intended to be Disclosed to the Public

Regarding whether the results are intended to be used in comparative assertions intended to be disclosed to the public, ISO 14044:2006 states that LCI studies should not be used for comparative assertions that are intended to be disclosed to the public. And it requires the practitioners to conduct an LCIA in case the results will be used in comparative assertions that are meant to be disclosed to the public. The GHG protocol (2011) does not support such a comparison between products' environmental performance.

From the literature review it seems that none of the reviewed articles has committed to the ISO standards in this regard. However, it is observed that OEMs like Volvo tend more toward publicly disclosing LCA reports that compare the carbon footprint of its models (Volvo, 2020). Similarly, Daimler conducted LCA to compare the environmental performance improvements of a new model with its predecessor (Daimler, 2018).

4.2 Scope

The scope is the second part of the first stage of LCA methodology according to ISO 14040:2006. The standard states that the scope should be sufficiently well defined to ensure that the breadth, depth, and detail of the study are compatible and sufficient to address the stated goal. The standard lists the following items as part of the scope description: the product system, the functions of the product system, the functional unit, the system boundaries, allocation procedures, impact categories selected and methodology of impact assessment and subsequent interpretation, data requirements, assumptions, limitations, initial data quality requirements, type of critical review (if any), type and format of the report for the study.

(31)

The Product System

ISO 14040:2006 defines the product system as; a system of consecutive and interlinked unit processes performing one or more defined functions, which models the product life cycle.

Similarly, the GHG protocol (2011) standard defines the life cycle as interlinked stages from raw materials to the end-of-life profile. In the literature, the product system and its boundaries were mostly described in text. Some articles presented the product system as unit processes flow diagrams, an example can be seen in Figure 5. The product systems studied in the Volvo (2020), Daimler (2018) and Polestar (2020) LCA reports comprise the different consecutive process units from raw materials extraction and refining to the end-of- life scenarios.

Figure 5 - Example of a Process Flow Diagram by Silva et al. (2018) The Functional Unit and Functions of the Product System

ISO 14040:2006 defines the functional unit as: quantified service of a product system for use as a reference unit for the LCA results and it is called the Unit of analysis in the GHG protocol standard. The GHG protocol (2011) states that a well-defined functional unit shall present the magnitude of the function or service, the duration or service life of that function or service, and the expected level of quality. Similarly, the ISO 14044:2006 states that the

(32)

functional unit should be clearly defined and consistent with the goal and scope of the study.

The reference flow shall be defined. If additional functions of any of the systems are not considered in the comparison of functional units, then these omissions shall be explained and documented.

The papers in the literature review did not seem to strictly follow these guidelines and requirements set by ISO (2006) or the GHG protocol (2011). However, the functional unit in the literature review was commonly defined as the function of the product being studied driving for a certain distance in a specific vehicle. E.g., Das (2014) described the functional unit as the transportation service over 250.000 km in North America. Koffler et al. (2014) described it as the function of the component over a vehicle lifetime of 150.000 miles.

Poulikidou et al. (2015) described it as the function of the component over a life cycle distance of a truck of 1.000.000 km. In the case when the product is a component of a car, the functional unit is defined as per the distance the respective car is estimated to drive over its lifetime. Similarly, the distance driven by a specific car is used as functional unit in the LCA reports of Polestar (2020), Volvo (2020), and Daimler (2018).

The System Boundaries

The system boundaries determine which processes and flows of the product system will be included in the LCA study. ISO 14044:2006 states that it is important to set the system boundaries in line with the goal of the study. The standard mentions the following requirements and guidelines: Decisions shall be made regarding which unit processes, inputs, and outputs to include in the study and the level of detail to which these unit processes shall be studied. The criteria that were used to set the system boundaries should be explained. The deletion of life cycle stages, processes, inputs, or outputs is only permitted if it does not significantly change the overall conclusions of the study. Any decisions to omit life cycle stages, processes, inputs, or outputs shall be clearly stated, and the reasons and implication for their omission shall be explained.

There is some overlap between the GHG Protocol (2011) and the ISO (2006) standard requirements mentioned above. The GHG protocol standard defines some additional requirements: The boundary shall include all attributable processes. Companies shall report the life cycle stage definitions and descriptions. Companies shall report attributable processes in the form of a process map. Companies shall report any non-attributable processes included in the boundary. The boundary for final products shall include the complete life cycle, from cradle-to-grave. The boundary of a cradle-to-gate partial life cycle inventory shall not include product use or end-of-life processes in the inventory results.

(33)

Companies shall report the time period of the inventory. Companies shall report the method used to calculate land-use change impacts, when applicable.

From the literature review it is estimated that approximately 45% of the papers describe a cradle-to-grave boundary of their LCA studies (e.g., Palazzo et al., 2019; Zhang et al., 2020;

Ji et al., 2020). For cradle-to-gate, this is also approximately 45% of which the end-of-life phase and in some cases the use phase is excluded (e.g., Gebler et al., 2020; Ferreira et al., 2019; Cecchel et al., 2018). The remaining 10% describe other boundaries, such as, gate-to- gate or cradle-to-cradle.

The Volvo (2020), Polestar (2020), and Daimler (2018) LCA reports all considered cradle- to-grave system boundaries. A cradle-to-grave will give a holistic picture of the environmental performance of a product. Both Volvo and Polestar have confirmed in their interviews that they are interested in receiving LCAs from their suppliers. The interviews with Volvo, Polestar and CLEPA unanimously conclude that the LCA studies they would potentially ask from their suppliers would be cradle-to-gate. However, all three interviews also acknowledged the fact that including the use phase and end-of-life phase identifies areas of interest about product performance for the use and end-of-life phase. Polestar stated in the interview that indirect emissions from activities (such as e.g., R&D, business travel, etc.) is something that is likely to be excluded in LCA reports.

ISO 14044:2006 states that it is helpful to describe the system and the system boundaries using a process flow diagram showing the unit processes and their inter-relationships. As mentioned above, the GHG Protocol (2011) demands it.

Allocation Procedures

According to the ISO 14044:2006 standard, the inputs and outputs for unit process shall be allocated to the different products according to clearly stated procedures. Step one of those procedures is to, whenever possible, avoid allocations. This can be achieved by dividing the unit process into two or more sub-processes and then collect data about these processes, or by expanding the product system to include the additional functions related to the co- products. Step two of the procedures describes that in the case that the allocation cannot be avoided, the inputs and outputs of the system should be partitioned between its different products and functions in a way that reflects the underlying physical relationships between them. The third step of the procedures describes that when the physical relationships alone cannot be established, other means of partitioning can be used e.g., the economic value of the products. GHG protocol (2011) suggests using allocation procedures similar to what ISO

(34)

14040:2006 mentions. Additionally, it suggests a way of performing the allocation by redefining the functional unit of analysis to include the co-products (additional functions) in the functional unit.

The papers in the literature review, most of the articles have not disclosed the procedures they used for allocations. However, it is mentioned that the allocation of flows and releases is done. Some papers have mentioned some procedures. E.g., Alonso et al. (2007) mention that they have used an incremental approach for fuel consumption and Ribeiro et al. (2007) who have used mass-related allocation approaches for low-mass components.

Daimler (2018) did not clearly mention the used allocations procedures. However, it is mentioned that the LCA report used the default allocation methods in the GaBi software.

Both Volvo (2020) and Polestar (2020) used an allocation method that follows the “polluters pay principle”. Which means that if there are several product systems sharing the same material, the product causing the waste shall carry the environmental burden. The reports state that this approach is recommended by the EPD system as well.

Impact Categories Selected and Methodology of Impact Assessment and Subsequent Interpretation

ISO 14044:2006 stated that the selection of impact categories, category indicators and characterisation models shall be justified. GHG protocol (2011) focuses solely on climate change as an impact category. In the literature review, mainly the midpoint impact categories were considered. In the literature review GWP was the dominant impact category which was used in all articles but one. This specific article, by Baumann et al. (2013), looked at the damage to human health as a solo endpoint impact category (disability adjusted life years; DALY). The LCA reports from Volvo (2020), Daimler (2018) and Polestar (2020) considered only GWP as an impact category. The interviews with OEMs reveal that there is no preferable framework for conducting the impact assessments as the date of the interview, however Polestar is considering the inclusion of ReCiPe 2016 when expanding the LCA’s impact categories scope.

The ReCiPe method is made up of indicators that fall into two levels; 18 midpoint indicators and three endpoint indicators (Huijbregts et al., 2017). The Midpoints categories represent the pressure (emissions and resources extractions) caused by human activities and the Endpoints represent the damage to the human health, ecosystems and resources availability caused by this pressure (see Figure 6). This pressure can be translated into environmental impact scores using characterisation factors. The characterization factors show the

(35)

environmental impact per unit of pressure e.g., Damage occurred per kg of material extracted or emitted (Huijbregts et al., 2017).

Figure 6 - ReCiPe 2016 Methodology (Huijbregts et al., 2017) Data Requirements

From the system boundaries, it becomes clear on which parts of the system data is required.

However, data can be collected from many places in many ways. Therefore, it needs to be decided how data will be collected on each respective part of the studied system. ISO 14044:2006 mentions that data can be collected from production sites or obtained and/or calculated from other sources. The standard argues that in practice, all data may include a mixture of measured, calculated or estimated data. From the literature review, this argument seems to hold up. Since the studied literature uses either measured, calculated, or estimated data, but often a mixture. The Volvo (2020) and Polestar (2020) LCA reports have also used a mixture of measured (themselves or by suppliers), calculated (from LCA databases) and estimated data.

(36)

The GHG protocol (2011) lists a set of requirements for data in a life cycle inventory:

Companies shall collect data for all processes included in the inventory boundary.

Companies shall collect primary data for all processes under their ownership or control.

During the data collection process, companies shall assess the data quality of activity data, emission factor, and/or direct emissions data by using the data quality indicators. For significant processes, companies shall report a descriptive statement on the data sources, the data quality and any efforts taken to improve data quality.

Data quality requirements are mentioned in 4.2.9.

Assumptions

ISO 14044:2006: stated that the cut-off criteria for the initial inclusion of inputs and outputs and assumptions on which the cut-off criteria are established shall be clearly described. GHG protocol (2011) requires the practitioners to make assumptions about the specific attributable processes involved in creating, distributing, and selling the studied product as they develop their processes flow map. When estimating the emissions for a process, an upper limit should be used and then benchmarked against a threshold to determine its significance. These assumptions should be transparently disclosed in the report. In the literature review, most of the articles described their assumptions. Some of those have dedicated a whole chapter to explain their assumptions.

The Daimler (2018) LCA report has not elaborated on any assumptions made in the LCA study. Both Volvo (2020) and Polestar (2020) stated that general assumptions have been made in a conservative fashion following the precautionary principle, in order to not underestimate the impact from unknown data. Additional processes have been added to the model when needed to represent actual emissions more accurately. The Polestar (2020) and Volvo (2020) studies did not include indirect emissions, e.g., the impact from the charging or fuel infrastructure, which means that only the cars themselves were assessed. Das (2014) made assumptions such as driving patterns, end-of-life recycling rate, and recycling yield.

Limitations

Assumptions, allocation methods, selection of impact categories, data accuracy and chosen system boundaries may affect the accuracy and credibility of the LCA results. Therefore, ISO 14044:2006 requires the disclosure and assessment of the potential effect the cut-off criteria have on the LCA outcome. Similarly, the results from GHG-only inventory should not be communicated as an overall environmental performance indicator for a product (GHG Protocol, 2011). For instance, potential impacts such as ecosystem degradation, resource

(37)

depletion and negative human health impacts are not covered by the GHG protocol (2011).

In the literature review, most of the reviewed articles did not discuss the limitations of the conducted studies. On the other hand, the articles that discussed the limitations, did it from a data availability and system boundaries perspective.

OEMs like Volvo and Polestar disclosed their LCA study limitations caused by the cut-off criteria applied on the product system (Volvo, 2020; Polestar, 2020). The carbon footprint in the LCA report is calculated based on data and assumptions on high levels of the product system, therefore, the results should not be broken down to the component level without reassuring that an acceptable level of detail is also reached on the studied sub-system. The Daimler (2018) LCA report did not discuss the limitations of LCA results in their report. The OEMs’ studies did not take rebound effects into consideration.

Initial Data Quality Requirements

Data quality requirements shall be specified, according to iso 14044:2006, to enable the goal and scope of the study to be met. The quality requirements should cover the following attributes:

• Time-related coverage: age and the minimum length of the data.

• Geographical coverage: the area from which the data originates.

• Technology coverage: specific technology or technology mix.

• Precision; allowed variability of the data values.

• Completeness: percentage of data that is available of a unit process.

• Representation: qualitative assessment of the degree to which the data set reflects the true population of interest.

• Consistency: qualitative assessment of whether the study methodology is applied uniformly to the various components of the analysis.

• Reproducibility: qualitative assessment of the extent to which information about the methodology and data values would allow an independent practitioner to reproduce the results reported in the study.

• Source of data.

• Uncertainty of information (e.g., data models and assumptions).

• How missing data will be treated.

Even though the ISO mentions this step clearly, in the scope in ISO 14040:2006 and explained as above in ISO 14044:2006, the LCA reports from Volvo (2020), Polestar (2020) and Daimler (2018) did not report any initial data quality requirements. Similarly, in the

(38)

literature review only Gebler et al. (2020) talk about classification of data in terms of accuracy (High, Medium, and Low data quality based on uncertainty, assumptions, calculations, etc.). None of the other papers discussed (initial) data quality requirements.

Type of Critical Review

The purpose of the critical review is to ensure the credibility of the LCA results. ISO 14040:2006 requires that the scope and type of the critical review to be defined in the scope phase of an LCA. It divides the possible critical review into two types based on the practitioners of the review process, by internal/external expert or by a panel of interested parties. In case of a review by an expert it is important that the expert is independent of the study. In the case of appointing a panel for the critical review, an external independent expert should be selected as chairperson of a review panel of at least three members.

ISO 14040:2006 requires that review statements, comments, and the response to them shall be documented in the LCA report and the review shall cover the following aspects: That the methods used are consistent with the standard, That the methods used are scientifically and technically valid, That the data used are appropriate and reasonable in relation to the goal of the study, That the interpretation reflects limitations identified in the goal of the study, That the report is transparent and consistent.

In the literature review some of the articles have undergone a critical review. (Koffler et al., 2014) was critically reviewed by an external independent panel. (Alonso et al., 2007) was reviewed by an ISO14040 expert. The GHG protocol standard is in line with ISO standards regarding the importance of the critical review for LCA credibility assurance. Daimler (2018) utilised a third party (TÜV) to carry out the critical review of its LCA report. Volvo (2020) and Polestar (2020) did not mention any type of critical review being carried out for their LCA results.

Type and Format of a Report for the LCA Study

ISO 14044:2006 states that the results of an LCA study shall be completely and accurately reported without bias to the intended audience. The results, data, methods, assumptions, and limitations shall be transparent and presented in sufficient detail to allow the reader to comprehend the complexities and trade-offs inherent in the LCA. The ISO standard 14044:2006 lists an elaborate list on topics and steps of the LCA that should be disclosed in an LCA report, specific to the applicable audience (internal, public or third party, etc.). These topics and steps cover all the stages of LCA methodology as defined in ISO 14040:2006.

However, this list does not demand to disclose all steps and considerations needed to

(39)

perform an LCA. For example, “initial data quality requirements” is not included in this list.

Similarly, the GHG protocol (2011) standard requires companies to clearly disclose general information regarding the studied product, the practitioners who conducted the study, the unit of analysis, the scope of the study, the boundary settings, allocations, data collection and quality, and an uncertainty assessment.

From the literature review, nearly half of the articles mention the use of ISO 14040:2006 and/or ISO 14044:2006 as framework for conducting the LCA study (e.g., Simboli et al., 2015; Delogu et al., 2017; Silva et al., 2018; Chen et al., 2020). The rest of the articles do not mention which framework is followed for conducting the LCA.

The Volvo (2020), Polestar (2020) and Daimler (2018) LCA reports, all report according to the ISO methodology. Volvo and Polestar additionally mention that they have used the Greenhouse Gas Protocol Product Life Cycle Accounting and Reporting standard (GHG Protocol, 2011) for additional methodological considerations. However, these reports seem to be missing some aspects, e.g., description of the data quality, the end-point categories to be protected, the characterisation models used.

The interviews with CLEPA, Volvo, and Polestar unanimously stated that the ISO methodology for LCA seems to be standard practise in the automotive industry. This is confirmed by the literature review in which nearly all papers stated ISO 14040:2006 as LCA methodology, for the papers that did disclose their LCA methodology standard. Occasionally ISO 14044:2006 was mentioned in addition.

In addition to LCA study reports that directly follow ISO 14040:2006 and ISO 14044:2006, other forms of reporting life cycle impacts exist, such as an Environmental Product Declaration (EPD). An EPD is a third-party verified document that assesses the LCA study, and therefore provides more credibility to the results. An EPD assessment bases the audit on ISO 14040:2006 and ISO 14044:2006. PEF is a methodology proposed by the European commission that serves as an alternative to ISO 14040:2006 and ISO 14044:2006. It can be characterised as aiming to increase comparability between products, as it decreases some of the flexibility that ISO facilitates. Even though PEF is a different LCA methodology than ISO 14040:2006, it is partly based on ISO standards. PEF, similarly to EPD, requires an independent audit.

4.3 Inventory Analysis

The inventory analysis is the second phase of LCA methodology according to ISO 14040:2006. The standard defines this stage in three steps: data collection, data calculation,

(40)

and allocation of flows and releases. The methodology of inventory analysis, as defined by ISO 14040:2006, is iterative. As more data is collected, more will be known about the system. This may initiate a revision of the data collection procedures so that the goal and scope of the study are achieved. Issues may arise that require a revision of the goal and/or scope definitions.

Data Collection

ISO 14040:2006 requires that for each unit process within the systems boundary data should be mapped under major headings, these data should cover the energy inputs, raw material inputs, ancillary inputs, and other physical inputs. It should also cover products, co-products, and waste. Additionally, it should cover emissions to air and discharges to water and soil. Other applicable environmental aspects should also be covered.

The standard mentions that the data collection step can be a resource-intensive process. It also states that it is important to consider practical constraints on data collection in the scoping stage of the study. The methodology of inventory analysis, as defined by ISO 14040:2006, is iterative. The standard states the following: “As data are collected and more is learned about the system, new data requirements or limitations may be identified that require a change in the data collection procedures so that the goals of the study will still be met. Sometimes, issues may be identified that require revisions to the goal or scope of the study.”

The GHG protocol (2011) has similar requirements to what ISO stated and it additionally requires companies to collect primary data for all the processes under the companies’

ownership or control. In the literature review roughly half of the reviewed articles do not mention where the data is collected from or how it has been done. The other half of the articles mention that the data comes from database averages, assumptions, or empirical data. By far, most of these articles have used a mix of database averages, assumptions, and/or empirical data.

The main data that are needed for the environmental validation are the bill of materials, geometries, masses, and the manufacturing process map (Delogu et al., 2018). The accounting LCA inventories typically reflect global or national averages of the involved unit processes (Ekvall and Andrae, 2006 as cited in Palazzo and Geyer, 2019). The lack of information and high degree of uncertainty hinder the use of traditional sustainability evaluation tools such as LCA during the early phases of product development (Shöggl et al., 2017).

(41)

Daimler’s (2018) LCA data is mainly collected from an in-house database about materials type, weight and other location and process specific data e.g., energy consumption during the production. The International Material Data System (IMDS) and Gabi databases are used as a source of data as well. Data regarding the use phase is retrieved through WLTP (Worldwide Harmonized Light Vehicles Test Procedure). End-of-life phase data is collected by Daimler (2018) itself in accordance with ISO 22628 Road vehicles — Recyclability and recoverability — Calculation method. In the Volvo (2020) and Polestar (2020) LCA reports mention several sources of data (see Figure 7). Primary data from operations run by Volvo or Polestar, such as factories and logistics. Data provided by the supplier for the battery modules, with guidance and support from Volvo and Polestar. Data constructed from LCA databases EcoInvent 3.6 and GaBi. Data constructed from IMDS databases, containing specifications on material compositions.

Figure 7 - Example Diagram of LCA Data Sources (Polestar, 2020)

From the interviews with Polestar, Volvo, and CLEPA, the suggested approach is that a first- tier supplier can start by performing a hotspot analysis by using the available generic data.

The assessed hotspots in that model could then replace the generic data with collected specific data. Over time, more and more data would be collected resulting in a model with as many collected data sets as possible along the whole supply chain.

(42)

Data Calculation

ISO 14040:2006 defines the following procedures for data calculation: validation of data, relating of data to unit processes, and relating of data to the reference flow of the functional unit. These procedures apply for all the unit processes included in the system boundaries, thereby, inventory results are generated for these processes. The GHG protocol (2011) states that companies should apply a GWP-100 factor to the emissions and removals data to calculate the inventory results represented by CO2eq. GWP-100 means the global warming impacts from the emissions over a time period of 100 years. The source of data of the GWP factors shall be reported. Weighting factors for delayed emissions, offsets and avoided emissions shall not be included when quantifying inventory results.

None of the articles in the literature review explained their calculation procedures, save for one. The LCA reports from Volvo (2020), Polestar (2020) and Daimler (2018) did not report how data is calculated.

Allocation of Flows and Releases

The need for allocation arises when a process within the product system has multiple input and/or outputs. Methodology of allocation is explained in 4.2.4. For instance, in the Daimler (2018) LCA report the practitioners have allocated cables and batteries used in the electronics according to their materials composition and considered the electronics components as only circuit boards. In the LCA report from Polestar (2020), 100% of the total emissions from scrap were allocated to the vehicles. That means the emissions from aluminium and steel production are included in the calculations from both scrap and actual mass used in the final product.

4.4 Impact Assessment

The impact assessment is the third phase of LCA methodology according to ISO 14040:2006.

The purpose is stated to evaluate the significance of potential environmental impacts using the LCI results. In general, this process assigns inventory data to specific environmental impact categories and category indicators, thereby providing better understanding for these impacts. The Life Cycle Impact Assessment (LCIA) phase also provides information for the life cycle interpretation phase. Issues such as choice, modelling and evaluation of impact categories can introduce subjectivity into the LCIA phase. Therefore, transparency is critical to the impact assessment to ensure that assumptions are clearly described and reported. The impact assessment as prescribed by GHG protocol (2011) only accounts for GHG emissions.

Life Cycle Assessment in the Automotive Industry