IN THE FIELD OF TECHNOLOGY DEGREE PROJECT
MEDICAL ENGINEERING
AND THE MAIN FIELD OF STUDY TECHNOLOGY AND HEALTH, SECOND CYCLE, 30 CREDITS
,
STOCKHOLM SWEDEN 2017
Feasibility of Life Cycle
Assessment for Complex
Medical Devices
SOFIA SVENSSON
This master thesis project was performed in collaboration with Elekta Instrument AB. Supervisor at Elekta Instrument AB: Mattias Lidström
Feasibility of Life Cycle Assessment for
Complex Medical Devices
Genomförbarhet av livscykelanalys för
komplexa medicintekniska produkter
SOFIA SVENSSON
Abstract
The interest in environmental issues is increasing and for this reason, assessing the potential environmental impacts of a product or system is of interest. A methodology developed for this particular purpose is the life cycle assessment, also known as LCA. It is not purely of interest these aspects are investigated though, as increasing requirements on organizations also matter. The purpose of this thesis was to investigate the feasibility to implement the methodology of LCA in the aspect of complex medical devices. To do this, the framework for the methodology has been reviewed and a case study performed. The case study comprised of conducting an LCA study on the radiosurgery device Leksell Gamma Knife® Icon™.
The outcome of the investigation showed that conducting an LCA study means a wide range of aspects need to be considered and specified to a high degree. A particular issue was the data requirements, as obtaining data meeting several objectives was challenging. The modeling was also identified as a difficulty. Tools such as software and databases with predefined processes were used, though as complex medical devices can use materials not common in other fields, a lack of appropriate predefined processes hinders the feasibility.
The conducted case study was able to attain valuable insights even though the study did not comply with the standards providing the framework, the ISO 14040 series. To conduct a compliant LCA study for complex medical devices, extensive resources would be required as well as the involvement of relevant parties along the supply chain. It is seen improbable to achieve a compliant study the first time a particular type of complex medical device is investigated. However, it is believed the feasibility would increase as studies are repeated, as the data quality is likely to increase. Advancements of the tools, as well as ongoing research on the environmental impacts of more materials, are other factors thought to increase the feasibility of conducting LCA studies on complex medical devices in the future.
Key words: Life cycle assessment, LCA, feasibility, environment, complex medical device,
Sammanfattning
Intresset för miljöfrågor ökar och därav finns det ett intresse att undersöka vad den potentiella miljöpåverkan är för en produkt eller ett system. En metodologi utvecklad för detta specifika syfte är livscykelanalys som även kallas LCA. Det är inte enbart utav intresse som aspekterna utreds, de ökande krav som ställs på olika aktörer spelar också roll. Syftet med detta examensarbete var att undersöka genomförbarheten av livscykelanalyser med avseende på komplexa medicintekniska produkter. Detta gjordes genom att granska regelverken för LCA samt genomförandet av en fallstudie, vilken utgjordes av en livscykelanalys på strålkniven Leksell Gamma Knife® Icon™. Resultaten av undersökningen visade att genomförandet av en livscykelanalys innebär att ett stort spann av aspekter måste beaktas och specificeras i hög grad. Ett särskilt problem var kraven på data då det var utmanande att samla in data som skulle möta flera behov. En annan identifierad svårighet var modelleringen. Verktyg användes i form av mjukvara och databaser med fördefinierade processer men då komplexa medicintekniska produkter kan bestå av material som inte är vanliga inom andra områden, var bristen på passande fördefinierade processer ett hinder för genomförbarheten.
Den genomförda fallstudien gav värdefulla resultat trots att den inte var utförd helt enligt standarderna i ISO 14040 serien. För att en LCA studie för komplexa medicintekniska produkter skall möta dessa krav krävs omfattande resurser och att flera berörda parter längs försörjningskedjan involveras. Det ses därför som osannolikt att en studie som genomförs för första gången på en viss typ av komplex medicinteknisk produkt kan leva upp till regelverket. Dock så förmodas genomförbarheten öka i takt med att studier upprepas, då kvaliteten på data tros öka. Utveckling av verktygen samt pågående forskning om miljöpåverkan från olika material är andra faktorer som anses öka genomförbarheten av livscykelanalyser på komplexa medicintekniska produkter i framtiden.
Nyckelord: Livscykelanalys, LCA, genomförbarhet, miljö, komplex medicinteknisk produkt,
Acknowledgements
This master thesis was conducted during the spring of 2017 and is the final piece for the completion of a civil engineering program of medical technology at the Royal Institute of Technology, KTH. It has been a pleasure performing this project and I want to express my gratitude to a number of people, whose help and guidance have been very valuable.
First, I want to thank my supervisor at Elekta Instrument AB, Mattias Lidström, for giving me the opportunity to conduct this interesting project and the encouragement throughout its course. I also want to express my gratitude towards the employees within the same company who took the time to participate in the project by contributing with their knowledge and support. The contribution from the people who participated in the data collection has also been invaluable and they deserve to be recognized as well.
I also want to express my appreciation towards Maksims Kornevs, my supervisor at KTH, for valuable input and support during this project.
My family also deserves to be recognized for their love and support throughout my whole education at KTH, it has been invaluable. Friends and fellow students have also played important roles for which I will be forever grateful.
Last, but not least, I want to thank the person who has been my greatest support throughout the whole journey of this education, my fiancé Peter Hellberg. I would not have been able to go through all of this without you, so thank you for always being there.
Stockholm May 21, 2017
Table of Contents
1 Introduction ... 1
1.1 Aim, Research Question and Goals ... 1
1.2 Structure of the Work ... 2
2 Theory ... 3
2.1 Sustainable Development ... 3
2.2 Life Cycle Assessment ... 3
2.2.1 LCA Methodology ... 4
2.3 Tools for LCA ... 5
2.4 Feasibility of LCA and Complex Medical Devices ... 7
2.5 Importance of Radiosurgery ... 8
3 Feasibility of LCA ... 9
3.1 Phase 1: Goal & Scope Definition ... 9
3.1.1 Feasibility of Phase 1 ... 10
3.1.2 Case Study Phase 1 ... 11
3.2 Phase 2: Inventory ... 15
3.2.1 Feasibility of Phase 2 ... 15
3.2.2 Case Study Phase 2 ... 17
3.3 Phase 3 & 4: Impact Assessment & Interpretation ... 19
3.3.1 Feasibility of Phase 3 & 4 ... 20
3.3.2 Case Study Phase 3 ... 22
3.3.3 Case Study Phase 4 ... 23
4 Discussion ... 31
5 Conclusion ... 33
6 References ... 34
Appendix A: Data Collection Form ... 38
Appendix B: LCI Data for Use Phase ... 42
List of Figures
Figure 1. Life cycle assessment methodology according to ISO 14040:2006. ... 5
Figure 2. Schematic overview of LCA tools and their relations. ... 6
Figure 3. Model of Leksell Gamma Knife ® Icon ™. Published with consent from EIAB. ... 12
Figure 4. Boundary between technical system and the environment. ... 13
Figure 5. System overview with technical boundary and indication on the type of data collected for the respective parts. ... 14
Figure 6. Network over the single score impact. ... 22
Figure 7. Impact assessment using the method ILCD 2011 Midpoint+ V1.07. ... 23
Figure 8. Process contribution to the single score. ... 24
Figure 9. Process contribution to the impact category mineral, fossil and renewable resource depletion. ... 25
Figure 10. Process contribution to the impact category human toxicity, cancer effects. ... 26
Figure 11. Contribution of components to the overall score of the potential environmental load from the Leksell Gamma Knife® Icon™. ... 28
List of Tables
Table 1. Relation between common objectives and classification for the respective type of LCA, according to different sources. ... 9Table 2. Functions of the Leksell Gamma Knife ® Icon ™. ... 13
Abbreviations
Bill of Materials BoM
Cone Beam Computed Tomography CBCT
Control system CS
Elekta Instrument AB EIAB
International Reference Life Cycle Data System ILCD International Organization for Standardization ISO
Life Cycle Assessment LCA
Life Cycle Inventory LCI
Life Cycle Impact Assessment LCIA
Patient Positioning System PPS
1
1 Introduction
The environment has undergone and is undergoing significant changes due to human activities. It becomes increasingly important to both acknowledge the issue and to act in accordance, meaning a need to incorporate a more environmentally friendly approach in all sectors. Healthcare has the purpose to help and cure people, ultimately to save lives. Even though medical devices aid the noble cause, the environmental aspects of the equipment need attention and to be cared for. In the healthcare industry, environmental aspects have been on the agenda, but mostly in regard to the hospital buildings, where the energy efficiency has been the focus. Besides the buildings, chemicals is another issue healthcare has had as a priority to reduce the environmental impact. Recently, medical devices are being considered as well. In 2012-2014, the Swedish Environmental Management Council was involved in the development of criteria for sustainable procurement of medical devices, more specifically, of Health Care Electric and Electronic Equipment [1]. Medical devices face high demands in terms of safety and performance. An ethical dilemma may thus exist where the objectives of those high demands and wanting to produce an environmentally friendly product contradict one another. To make educated decisions in this matter, it is necessary to understand how much, and in what ways, the product may affect the environment. Through studies adopting the methodology of life cycle assessment, LCA, this issue is analyzed. A number of standards provide the framework for the methodology, including numerous specifications one ought to follow when conducting such analyses. The revised standard of ISO 14001 requires organizations to incorporate a life cycle perspective to understand their environmental impacts [2]. It does not require a detailed LCA study, though the demands of the standard might be difficult to meet without such an analysis. A number of initiatives exist to promote the use of LCA and the interest for the methodology is increasing [3]. The progression of the environmental work of corporations is not only due to laws and regulations, other influential factors are concerns about the brand, reputation, and position within the field [3].
1.1 Aim, Research Question and Goals
The environmental responsibilities for companies are increasing and understanding the complexity of implementing the LCA methodology is thus of interest. Therefore, the aim of this thesis is to investigate and understand the complexity of implementing LCA, in the aspect of complex medical devices. The research question is formulated to: What is the feasibility of conducting LCA studies on complex medical devices? To investigate this, a case study is performed on a radiosurgery device, more specifically, on the Leksell Gamma Knife® Icon™.
The goals for the thesis are the following:
1. Identify the main challenges for each phase in the LCA case study. 2. Create a model in LCA software of the Leksell Gamma Knife® Icon™.
2
1.2 Structure of the Work
To provide an overview of the thesis, the report has the following structure. In the second chapter, theories regarding sustainable development and aspects of both the framework and tools of the LCA methodology are described. Also, the object of the case study is presented in this section. The focus of the third chapter is on the feasibility of the LCA methodology. The structure of this chapter is outlined to follow the different phases of LCA. The subchapters begin with details on the respective phase according to the framework, which is followed by identified aspects affecting the feasibility, then closing with the conducted case study. The fourth and fifth chapter include the discussion and conclusions of the thesis while the final chapters contain the references and appendices.
3
2 Theory
To follow the reasoning of the project in its entirety theories regarding sustainable development and the methodology of LCA are outlined in this section. A definition of the concept of complex medical devices is also provided along a presentation of the object for the case study.
2.1 Sustainable Development
The awareness of environmental issues is increasing and several initiatives exist to drive a more sustainable development. The environment is one aspect of sustainability, however, the word has a broader meaning, incorporating social and economic aspects as well. In 2015, The United Nations formulated seventeen goals for sustainable development, namely the Sustainable Development Goals [4]. The purpose was to make countries take responsibilities and work actively to achieve the goals by 2030 [5]. The goals demonstrate the wide meaning of the word sustainability by concerning aspects of poverty, of the environment, and of equality. The goal relating the most to the topic of this project might be the twelfth goal, regarding sustainable consumption and production. A part of the work by the European Commission, for that specific goal, includes promoting LCA by providing both a platform on the topic and a comprehensive handbook [6]. Many organizations work with environmental issues, two of them playing key roles are the Society of Environmental Toxicology and Chemistry, and, the United Nations Environment Programme. A collaboration between these organizations led to the International Life Cycle Initiative, promoting the concept of life cycle thinking and the exchange of ideas within the field [7].
The driving forces promoting the development of more environmentally friendly products are often policies and regulations. At the introduction of new regulations, there are often exemptions. An example of this is the European Directive 2002/95/EC which restricts the use of hazardous substances in electrical and electronic equipment. When implemented in 2006, medical devices were exempt but is included since 2014 [8]. Another European directive with environmental considerations, relevant for product development, is the Ecodesign Directive 2009/125/EC, with requirements on the energy performance of the product throughout its lifetime [9]. A more general regulation is the standard ISO 14001, Environmental Management System. The revised version of it, ISO 14001:2015, includes a few significant changes from the previous version. The changes increase organizations responsibilities by putting further demands on incorporating environmental work in the organization, and to use the concept of life cycle thinking [2].
2.2 Life Cycle Assessment
The purpose of adopting the methodology of life cycle assessment, LCA, in a study is to quantify the potential environmental impacts of a product or system throughout its entire life cycle, including all steps from commodity extraction to waste management. Established standards, the ISO 14040 family, provide the framework for conducting such analyses. Environmental aspects may include trade-offs, however, LCA has the ability to recognize these, making the approach useful and comprehensive [10]. The methodology has also been identified by the European Commission to be the “best framework for assessing the potential environmental impacts of products” [11].
4
resources, and studies investigating these issues, known as Resource and Environmental Profile Analyses, were conducted [12]. In 1969, an investigation with a different approach was performed by the Midwest Research Institute, studying a variety of beverage containers for the Coca-Cola Company. Similar studies followed shortly and the methodology of LCA was starting to form. [13] The concept of studying the environmental impacts of a product throughout its lifetime spread, as well as the methodologies and terminologies, which during the 1970-1990 led to the need to standardize the method. It was not until 1990 during a workshop, by the Society of Environmental Toxicology and Chemistry, the name LCA was coined [10]. Thus, in the following decade, both the terminology of LCA and the framework for the methodology was formed [13]. The International Organization for Standardization, ISO, played an important role, and still do, by formulating the framework and publishing standards. The first standard of LCA, ISO 14040, was formulated in 1996 but revised in 2006 [10]. The standard provides some insight on what to include, however, the details on how to conduct analyses are not regulated therein, leaving room for interpretation. Through the development of the method, a number of initiatives have taken place over the years, resulting in a number of published scientific reports, as well as several handbooks regarding the methodology of LCA. An example of a comprehensive handbook is the International Reference Life Cycle Data System, ILCD, Handbook – General guide for Life Cycle Assessment, which was developed by the European Commission [6]. The interest in LCA has continued to grow over the years, and it has been applied in areas of both tourism and military systems, showing great diversity in applicable fields [13].
The healthcare industry, and its suppliers of medical devices, have only applied LCA to a certain degree and it mainly seems to be for comparative purposes of single- and multi-use products, two examples are studies which investigated laryngeal mask airways, and, dental burs [14] [15]. LCA studies on advanced medical technologies have not been conducted to a high degree, at least not which has led to publications. The lack of published materials may though not represent the usage of LCA, as it may be used for internal purposes. The actual knowledge of what the potential environmental impacts are of such devices is thus difficult to assess, but, the lack of published material on the matter is believed to indicate the knowledge of these aspects is low. Radiosurgery devices are advanced medical devices and as of autumn 2016, there has not been any official analysis of the environmental impacts of that type of device.
2.2.1 LCA Methodology
5
Figure 1. Life cycle assessment methodology according to ISO 14040:2006.
The first phase includes to plan the study and outline the framing details while the second phase involves collecting all the necessary data and model the system under study. In the third phase, the data is mapped to various impact categories and in the final phase, the results are interpreted and analyzed. According to the mentioned standards, an LCA study is an assessment of a product or system throughout its life cycle, from cradle-to-grave. The life cycle includes several steps ranging from raw material extraction, to manufacturing, to transportation, to use, and, to waste management. For practical reasons, variations of the method are adopted, meaning simplified versions of the regular LCA. These are sometimes referred to as partial LCA studies, including cradle-to-gate analyses, meaning the steps from raw material extraction to a finished product are included, while steps representing the use phase and the waste management are excluded [10]. The cradle-to-gate assessment has been identified as the most common type of LCA applied in practice [18].
2.3 Tools for LCA
6
Figure 2. Schematic overview of LCA tools and their relations.
7
The LCIA methods provide the framework for the features of the impact assessment in an LCA. This means having predefined impact categories, a list of all included substances and rules on which substances effect each impact category by how much. A range of methods exist and the differences between them are their respective rules, forming their frameworks. This means the impact categories differ, both in regard to the number of categories, and, the aspects they cover. Other differences are which substances that are included in the assessment, and the rules for which substances that are mapped to a certain category and also the effect each substance has on the category it is mapped towards.
A widely used method is ReCiPe, developed by Pré Consultants, who also was one of the founders of the software SimaPro. The method was developed using the LCIA method Eco-indicator 99, which earlier was the most commonly used method by LCA practitioners [18]. Another method is the ILCD 2011 Midpoint+ method, developed by the European Commission, and it is based on the same work that led to the ILCD Handbook, by the same organization. Though the method is not as commonly used as ReCiPe, at least not for published studies, it has been applied for varying fields, though with the older version, ILCD 2011 Midpoint [22] [23] [24] [25].
2.4 Feasibility of LCA and Complex Medical Devices
Through the development of LCA, the requirements have increased due to the work for standardizing the methodology. The combination of these requirements, and the increased interest of LCA, makes investigating the feasibility of LCA studies interesting. A study from 2014 which reviewed LCA studies on desalination facilities to investigate the methodology, showed that two areas in need of improvement were the feasibility and reliability of LCA [26]. Thus, investigating the feasibility of LCA methodology for complex medical devices are of great importance.
This thesis will concern complex medical devices and for that reason, the concept will be defined. The meaning of the word complex can vary greatly and be depending on the context. In the article On the complexity of medical devices and systems, the issue of what complexity means is debated [27]. Two perspectives can be seen on the complexity of medical devices, from the user and the designer. In the first case, the perceived complexity depends on how easy or difficult it is to operate the device, while in the latter, the perception refers to the number of components and the technical specifications of the device [27]. Although both perspectives are important and a medical device can experience both types of complexity, this thesis will focus on the complexity from the perspective of a designer.
8
2.5 Importance of Radiosurgery
Healthcare is facing many challenges in the near future. The population is growing and aging, meaning more people need care. One challenge is the increasing number of cancer patients. Cancer is a disease which has become more prevalent over the last decades and the World Health Organization predicts a 70% increase in the number of new cases in the following two decades [30]. Statistics provided by the Cancer Research UK show that the incidence rate of brain, other central nervous system, and intracranial tumors increased by 12% in the last decade, and worldwide, it was estimated that in 2012, more than 256 000 such diagnoses were made [31]. Also, statistics from the American Cancer Society show that these type of cancers represent the leading cause of cancer deaths among people younger than 40 years old [32]. There is thus a clear need for treatments of these types of cancers. The treatments vary depending on the type and size of the cancer. Some general principles for treating cancers are surgery, radiation therapy, and chemotherapy. One form of radiation therapy, suitable for some types of brain tumors, is radiosurgery. Besides some types of brain tumors, the method can be used to treat other disorders such as arteriovenous malformations and trigeminal neuralgia. A radiosurgery device is a complex medical device, facing high demands in terms of quality and safety. The device mostly used, and also well-known, utilizes gamma radiation to irradiate the intended target. Other types have been developed, using either protons or are linear accelerator-based solutions. The first radiosurgery device was called a gamma unit but became more known as the Gamma Knife and was developed by Lars Leksell, a Swedish neurosurgeon, and the very first device was put in place at a hospital in 1968 in Stockholm [33]. In 1972, Leksell started the company Elekta AB and as the company grew, structural changes were made. Elekta Instrument AB, EIAB, is the part of the company which developed the most recent version of the Gamma Knife, the Leksell Gamma Knife® Icon™ which will be the object of the case study. The Gamma Knife has evolved over the years though the method is still based on the same principles, using several radioactive sources to focus the radiation to a focal spot, to irradiate the intended target. The evolution lies primarily within radiation protection, the precision, and possibilities to plan the treatment. Today, a Gamma Knife is used to perform radiosurgery on around 70 000 patients each year [34].
9
3 Feasibility of LCA
In the following subsections, each phase of LCA is presented. The first and the second phase of LCA are presented separately, each beginning with the theoretical aspects set by ISO 14044:2006, followed by aspects on the feasibility. The sections end with the conducted case study. The third and the fourth phase of LCA are presented together though the layout is the same as for the other subsections, in regard to the presentation of the theory, feasibility, and, the case study.
3.1 Phase 1: Goal & Scope Definition
The first phase when conducting an LCA is the goal and scope definition, and it can be considered to be the most critical part of an LCA [35]. Other sources also point to the importance of the definition, which forms the foundation for the whole LCA and influences decisions throughout the process [10] [17] [19].
When formulating the goal definition, there are four main areas that need to be considered: (1) intended applications, (2) reasons of the study, (3) target audience and (4) if the study is comparative and with the purpose of disclosing results to the public [16]. As part of the goal definition, the type of LCA that is to be conducted should also be declared. One could argue that identifying the type of LCA is one way to summarize the aspects that need to be considered, as the type of LCA is dependent on the objective. There is no common agreement on what the possible types are, or what to call them. In Table 1, the variation in classification is shown by indicating common objectives and what an LCA with each objective would be referred to, according to different sources. The purpose of identifying which type of LCA it is is the influence the type has on the methodology. If the deliverables are to be disclosed to the public, or to be compared to other products, higher demands are set on the methodology.
Table 1. Relation between common objectives and classification for the respective type of LCA, according to different sources.
Common objectives Learn more about the
system under study
Compare systems
Evaluate impact of decision
Classification according to
Bauman & Tillman [19] Stand-alone Accounting Change-oriented Classification according to
ILCD Handbook [17] Situation A Situation B Situation C Classification according to
other sources [36] [37] Attributional Consequential Consequential
In the scope definition, many aspects of the LCA should be outlined. In fact, 14 aspects need to be addressed, some are in regard to the object of the study while others are methodological choices for the LCA. Describing all of these would mean repeating the ISO 14044:2006 standard and therefore only a few are described.
10
For stand-alone LCA studies though, the definition of the functional unit is not critical [19]. Based on the definition, a reference flow can be determined to indicate how much of a product that is needed to deliver the function and is thereby a measure of the performance of the product [38]. The method to be used in the impact assessment phase should also be stated in the scope, along with the type of impact categories. The requirements on data, in regard to the types and sources as well as of the quality, shall also be addressed. The latter should be specified according to as many as ten different aspects. The intended analyses for the interpretation phase is an additional part required to mention. Another important aspect to state is the boundaries of the system.
Three types of boundaries can be identified, between the technical system and the environment, between the technical system to be studied and other technical systems, and, the boundary between significant and insignificant processes in the system under study [39]. Potential additional boundaries concern geography and time [37]. The boundary of significant processes is sometimes referred to as the cut-off. There has been a debate on what the cut-off should be based upon, which has revealed that it often is subjectively chosen and not scientifically based [40]. Another issue to address regards allocation, how to deal with the problem that several products share an environmental load. The standard ISO 14044:2006 describes possible procedures to deal with allocation and in what order they ought to be applied. The primary procedure is thereby to avoid it, and when impossible, to partition the load between the involved parts based on physical or economic relations [16].
Aspects not described above but required to be included in the scope definition are the assumptions, value choices and the limitations of the study, as well as how the study will be reported and whether a critical review will be conducted.
3.1.1 Feasibility of Phase 1
The first phase can be considered to consist of two parts as there are two definitions to set, the goal and the scope. The goal is easier to define as it is comprised of more general aspects, often known before the study is conducted, such as the purpose for it. A part which may require some efforts though is the specification on the usage of the results, and the documentation of it required for transparency. Determining the type of LCA is important, but, as long as the objective is clear, the name for the type does not affect the study although it would make comparisons between studies easier if there was a common consensus on what to call the different types.
The scope covers all aspects of the study, into detail. Therefore, this part of the first phase is challenging. Depending on who the practitioner is, the knowledge of these aspects will vary. The main issue is how, and when, all 14 aspects should be specified. Conducting an LCA study is an iterative process, and due to the high level of specificity for the requirements, defining all aspects from the beginning will likely mean a need to revise them along the way. This is especially the case when the study is conducted for the first time on a type of device for which no similar study previously has been conducted. If previous LCA studies of similar systems exist, they may give an indication on how to specify the requirements. That information would only be indications though, as the accessibility to information and the details of it may vary between different organizations and be difficult to know before the study begins.
11
The details regarding the data quality are an aspect considered to be extremely difficult to set at the beginning of the study. Again, the methodology of LCA is iterative and allows for alterations to be made throughout the study as long as they are documented. The need to make these decisions early on in the study can be debatable as they are likely to be revised later. Also, when conducting LCA studies for the first time on a particular type of device, it seems unlikely for the practitioner to have an idea, close to the reality, on what data will be accessible, and, the quality of it. The requirement to define the quality of the data in terms of the ten specific aspects is considered to be difficult to set, and, it is also considered to be unlikely that the quality of all the collected data will be consistent.
The points in time when decisions are made in a study is also an issue. The choice of impact assessment method could be argued to be very difficult to set from the start before the data collection even has begun. Making a choice is not difficult, but motivating its appropriateness can be. If the study will be compared to previous studies of the same type of system, sufficient knowledge may exist to make a well-founded choice, but otherwise, it is likely the final choice needs to be made after the inventory, when the included substances are known.
Describing the system and its functions are likely to be an easier part of defining the scope as this is information often described in documents or knowledge of the practitioner. Describing this in terms of a functional unit is more difficult as it needs to be a quantified unit. The impact of the definition though is debatable. If the deliverables are to be compared to another product, the purpose this definition serves is clear. It is not as obvious though for purely attributional LCA studies. For those, it is still important to understand what the results are referring to, but, as this information is already described, the functional unit can be seen as redundant.
Determining the system boundaries and describing them is another concern. Including parts for which one might lack control over or have very limited information of may not be the most appropriate way. Including as much as possible would, of course, give plenty of information, though the additional resources and what the outcomes would lead to, might not be worth it. Therefore, the benefits of expanding the boundaries and the required resources to obtain enough information to reach these benefits, need to be balanced. It may therefore also be that the system boundaries restrict or widen the objectives of the study, and if so, the modification needs to be documented. Documenting all alterations is required by ISO 14044:2006 for the purpose of transparency, however, this might be an issue for practical reasons, both to actually document it and presenting it in the report of the study.
3.1.2 Case Study Phase 1
12
device. Based on the goal of the study, it is classified as an attributional LCA study. As the definition of the scope covers many aspects, these are described in the following sections.
A general radiosurgery device was briefly described in section 2.5, however, as the system for this LCA study is the Leksell Gamma Knife ® Icon ™, a model of it is presented in Figure 3, and, an overview is described in the following.
Figure 3. Model of Leksell Gamma Knife ® Icon ™. Published with consent from EIAB.
The RU is the part in which the radioactive sources would be placed in and thereby consists of parts allowing the sources to be positioned correctly. The functions could thereby be described as providing radiation, and, to focus and shape the radiation beam. Another function of the RU is the provision of the radiation protection. The part positioning the patient in the correct position is the PPS, composed by a patient couch and the associated features to perform its tasks. Another essential part of the system, though not depicted in the figure above, is the electronics supporting the functions of the Gamma Knife which is referred to as the control system, CS, and additional functions of this part are those associated with the patient surveillance. The covers are, as the name suggests, the covering for the complete system for protection purposes and contribute to the aesthetics of the device. The parts mentioned have all been common for a number of versions of the Gamma Knife, while the Gantry is a new feature with the purpose of providing possibilities to take medical images, by cone beam computed tomography, CBCT.
13
Table 2. Functions of the Leksell Gamma Knife ® Icon ™. Functions of Leksell
Gamma Knife ® Icon ™
Functions of system in relation to LCA
Provide radiation Excluded
Focus, shape radiation Included Provide radiation protection Included Positioning of patient Included Imaging technique (CBCT) Included Patient surveillance Included
Since the lifetime of the device is set to 10 years by the manufacturer, the functional unit for this LCA study is defined as providing the settings for radiation of intended targets in the brain while providing appropriate protection for 10 years. Due to the type of product, the reference flow was the same as the functional unit, since to fulfill the purpose of the device, only one product is necessary.
The first dimension of the system boundaries, between the technical system and the environment, is modeled in Figure 4 and indicates indirectly the exclusion of the waste management. The reason for excluding this part, and thereby not conducting a full LCA, is the limited resources of this project to develop a case specific waste scenario, which would be required as no predefined processes for a waste scenario suitable for this type of device exist.
Figure 4. Boundary between technical system and the environment.
14
Figure 5. System overview with technical boundary and indication on the type of data collected for the respective parts.
The data requirements for this study are not specified in accordance with ISO 14044:2006. Instead, by considering the goal of this LCA study, it is assessed to be sufficient to gather information of two types, primary and secondary data. The first means the information is obtained from both EIAB, and their suppliers while the latter means less effort is made acquiring the data, and, larger approximations are made. In Figure 5, the components for which the two types of information are collected are indicated by their respective boxes. The data used for the modeling in the second phase originates from the database ecoinvent, and thereby the problem of allocation is addressed in the same way the database addresses the issue, mainly by mass or economic allocation. The impact assessment method used in this LCA is the ILCD Midpoint + method, version V1.07. The primary choice was though the ReCiPe method but the change was made due the discovery of it being unsuitable for this particular case. This choice is further discussed in section 3.3.2.
15
they are not for this LCA study. One reason for this is the lack of identified value choices. Another reason is the range of assumptions made throughout the study, outlining every aspect would mean to describe the approximations made for the modeling of each component of the object for the study. As this is not considered to be feasible, the assumptions are discussed in regard to their respective area in the section of the inventory, 3.2.2.
3.2 Phase 2: Inventory
The second phase of an LCA is the inventory. It includes the planning and execution of the data collection, thereby requiring the most resources in the LCA study. It also includes constructing a flow model and make calculations [19]. An important feature of collecting data is to ensure it is of the required quality, which is an essential aspect when conducting an LCA [35]. Often, the data in the LCI are not of consistent quality [10].
3.2.1 Feasibility of Phase 2
As the second phase of an LCA also can be seen to be composed of two major parts, the collection of data and the creation of a model in the software, which enables the calculations, the feasibility is discussed in accordance with those two parts.
For the data collection, the conditions to be met are set by the goal and scope definition. The first concern is how to do the collection. A data collection form helps collect data in similar ways for all parts included in the study. Designing the form can be challenging though. Several objectives exist which may contradict one another. Two such objectives are the need of feedback, and, the need of detailed information. The more details that are requested, the higher is the risk of not obtaining answers. The detailed information is necessary though both for the purpose of creating a model with high external validity, and, to meet the data quality requirements set in the scope. It is also difficult to know what kind of information, and the details of it, the person filling out the form has, which also contributes to the challenging task of designing the form. Promoting people to answer is another difficulty, which may be related to the way the form is distributed.
As complex medical devices are composed of several parts, the chain of suppliers involved in the production of the object for the LCA study may be longer than the suppliers closest to the manufacturer. Obtaining information from all involved parties would require a great amount of resources and is likely to not be feasible for complex devices. The manufacturer, the company initiating the study, is also a provider of information. The type of information the manufacturer and its suppliers have is likely to vary, but for some parts, the same information might be gained from both parties. For those parts, it can be an idea to collect the data from both parties to check whether they match, which is a way to validate the data or to get indications on its uncertainties.
16
The feasibility for attaining information internally within the company is determined by the structure of the data. For LCA studies conducted for the first time, the accessible data may not be structured in the most appropriate way for this purpose. Companies may have lists of the materials for all parts, known as Bill of Materials, BoMs. Information on materials and their respective weights may thus be found in these lists. It becomes difficult when the lists are not accurate. Parts not designed within the company may be referred to as components and thereby not be specified on the same level of detail as other parts.
Another challenge is to obtain the desired information from external suppliers. The reasons for this may be many and a few will be discussed here. The first is the market structure for manufacturers of complex medical devices. The high demands of these devices may limit the number of possible suppliers. This means the relations between a company and its suppliers need to be cherished. Thereby, a dilemma exists on taking environmental responsibilities and setting demands on the suppliers, while not putting too much stress on them as the relationships are very valuable. As suppliers are likely to not be the initiators of the study, their willingness to participate in the study may vary depending on how far they have come in their own environmental work, and, their current workload at the time which the study is conducted. Another factor likely to influence is their accessibility to the data asked for, which might vary between different suppliers. If reaching out to a supplier at an inconvenient time for them, and, the details of the data asked for would require them to spend plenty of resources, the chances for them to participate are likely to be low. This and the previously discussed aspects regarding the data collection will ultimately affect the data quality in the study.
This leads to the second part of the inventory phase of the LCA, creating a model of the system in the software. There is a relation between what the software includes in terms of databases and how processes are structured, and, what type of data that needs to be collected. This affects the external validity of the model as there will be differences between the collected data and the data used in the model. For the specificity of materials, this can be shown by the following example regarding stainless steel. From the manufacturer, it might be possible to obtain information on the level of the type of stainless steel, while from the suppliers, it might only be feasible to obtain information on the general level, using only the term stainless steel. If the database in the software only includes one process representing stainless steel, it would be unnecessary to collect data on the level of details on which type of the material is used, as this would not be represented in the model anyway. This also shows the iterative approach required, and, how determining the data quality is connected to both the collection and modeling parts of the inventory.
17
The database ecoinvent provides market processes which makes the modeling easier for the practitioner when the particular type of material is not known. These processes reflect the average consumption mix of the material, as they are made up of a mix of the possible different types. Another scenario which can make the modeling difficult is when there is no predefined process matching the data collected. Features in the software allow the practitioner to create own processes. But, doing this for a material for which the database has no information is difficult as to model it accurately, detailed information such as data on emissions to air, water, and soil is needed. For complex medical devices, the high demands might require usage of materials uncommon for other fields, and, thereby this issue is highly relevant when considering the feasibility.
An additional aspect of the same scenario, when no predefined process matches the collected data, is the use of similar processes. For other applications, substituting the real process with a similar one might work as they may fulfill the same purpose, however, this approach is not necessarily appropriate in an LCA study. The reason for this is that even though their functions are similar, their environmental impacts may be very different. This means an understanding is required on what the environmental impacts are of the real process, to be able to judge whether the similar process is suitable to use as a substitute in the study.
It is common for complex medical devices to be featured with standard components such as computers, screens and smaller parts such as printed circuit boards and cables. Collecting data and model these components accurately would be challenging, especially as they are likely to be bought as ready-made components. The manufacturer may thereby lack information in regard to what they are composed of and the processes included to make them. To facilitate the modeling of such parts, predefined processes representing this type of electronics exist in databases. However, if the waste treatment is to be included in the LCA study, as it should be in a complete LCA study, these items would not be included as they are not measured in a mass unit. Issues regarding the units could be avoided by altering the units of the existing processes to mass units. The feasibility of conducting such alterations varies on the type of process. Some processes are described in detail and can reveal what one item of the process would weigh, this is not always the case though and an aspect the practitioner needs to be aware of. The ability to use standard components allows for LCA studies to be conducted on complex systems, though the span of the existing parts needs to be widened with an increased number of specified parts to facilitate the modeling further.
3.2.2 Case Study Phase 2
To be able to conduct the inventory, a data collection form is developed. An existing form from a handbook [10] and knowledge on important aspects to cover, gained from a literature review performed at the start of this project, serves as the foundation for the design of it. In the designing process, an evaluation is also made between the gain of obtaining detailed information, and, the ease to fill them out. The final design of the data collection form is found in Appendix A: Data Collection Form.
18
in detail in regard to its material and weight. To obtain an estimation of these parts, discussions are held with experts on each component within EIAB. The information gained from BoMs and the described discussions are used to partially fill out data collection forms for each respective component. These forms are then distributed to the suppliers through e-mails along with an attached cover letter. Also, telephone discussions are held with the suppliers to explain the form and how the results will be used.
To model the components by the inventoried materials and weights, conversions are made to suit existing processes in the database ecoinvent. An example of this is that the BoMs used have specifications on the type of steel, while in the database, all types of stainless steels are modeled using the market process of chromium steel, a process including a mix of different types of stainless steel. When no matching materials are found to the inventoried data, scientific articles, the ecoinvent support forum, and the LCA Discussion List are advised. The modeling of the material tungsten is conducted this way and uses data from the supporting information of the article Life Cycle Assessment of Metals: A Scientific Synthesis [41]. For parts of certain components for which secondary data was collected, approximations are made in the modeling. This means that the parts are modeled using processes in the database representing defined components such as computers, transformers, and cables. To make these approximations, discussions are held with experts within EIAB.
To model the manufacturing, average processes are used as the obtained information regarding the processing of materials in the production is inconsistent. The level of detail differs substantially between suppliers and in the effort to be consistent in the modeling, new processes are developed based on predefined ones. Specific processes for metal working exist for the materials steel, chromium steel, and aluminum. The processes contain links to the same processes, except for the added input of the specific material. The addition of materials exists due to estimated losses during the manufacturing. This means the predefined processes have the same inputs and outputs except for the input of their respective metal. Thereby, the processes created in this project was produced by copying the existing processes and changing the added metal to match the metal collected in the inventory. Another alteration is made on each process representing the metal working, even those predefined, in regard to the amount of added materials. In the original processes, the estimations of the losses are the same for all materials, 22.7%. Information from data collection forms and discussions with suppliers indicates the actual losses for their production is closer to 10%. Therefore, this is used as input for all metal working processes used in modeling.
19
3.3 Phase 3 & 4: Impact Assessment & Interpretation
The collected data from the LCI results in a comprehensive list of all included substances, and as this would be difficult to interpret into what the potential environmental impacts are, the purpose of the LCIA is to make the data easier to comprehend [10]. According to ISO 14044:2006, this phase includes the mandatory steps of selection of impact categories, classification, and, characterization, while the optional steps are normalization, grouping, and weighting [16]. By choosing an existing LCIA method, the framework for the steps is already defined. What is accomplished in each step is described in the following sections.
The selection of impact categories means to determine which environmental issues the study will consider. Classification refers to the mapping of substances from the LCI to their respective impact category. The specific categories used, vary between different methods, though two distinct types exist, midpoint and endpoint impact categories. The first are closer to the inventory data, resulting in less uncertainty while the latter may be easier to understand since data has been further aggregated, but at the expense of increased uncertainty [21]. Characterization is a process which compensates for the fact that substances mapped to the same category may be more or less harmful [21]. This is done by using characterization factors, which are multiplied to the substances. The factors is another aspect in which different methods vary [21].
Normalization is a procedure in which the results in each impact category is calculated in relation to the functional unit, or the reference flow, to ease the understanding of the impact categories in relation to one another [16]. The normalization also means it is possible to evaluate the result of each impact category to one another since this step means all impact categories use the same dimensionless unit [10]. Grouping is another way to facilitate the interpretation of the impact assessment results, by sorting impact categories into groups [16]. Weighting means to assign categories weights, reflecting their relative importance, but, assigning weight introduces subjectivity which can be controversial [12]. After weighting, it is possible to summarize the scores of each impact category to obtain a single score. To obtain a single score, without introducing the subjectivity, equal weights can be used.
20
Monte Carlo techniques can be applied to evaluate some of the mentioned factors and in SimaPro, the feature of parameters can be useful since they can be used as switches to facilitate the varying of factors to observe their influence on the overall results [21].
Checking for completeness is an essential part of the interpretation and shows the iterative approach required in an LCA study. Therefore, it is important the procedure of the interpretation is made systematically, reported thoroughly and in line with the goal and scope. The main purpose of this phase is to present conclusions while making it clear which assumptions that were made and the uncertainty of the results. [10]
Specifications on how to conduct the evaluations are not described in the ISO standards. Two different approaches exist though, procedural and numerical. The first means to perform evaluations with the help from other sources than the data used in the study, such as reports and expert judgment. The second approach means using algorithms on the same data used in the study. Analyzing the influence of the definitions of the system boundaries and the functional unit is also an important aspect. [10]
3.3.1 Feasibility of Phase 3 & 4
When evaluating the feasibility, two aspects can be investigated, the feasibility of reaching the purpose of the phases, and, the feasibility of conducting all parts according to the specifications in the ISO standards. These aspects ought to be the same, though there might exist more ways to fulfill the purposes besides the listed steps in the standard. The purpose of the LCIA is to evaluate what the potential environmental impacts are and their respective significance, while the interpretation should deliver results in line with the goal and scope, and provide a clear understanding of the limitations of the study and reach conclusions [42].
For the LCIA, the included steps are performed almost automatically when using a software, as the tasks are performed according to the framework set by the included LCIA method. A true benefit of the ease of conducting the LCIA, is the opportunities to early on in the study get an idea on what the issues are. The main concern of this phase is, as described in section 3.1.2, the choice of which LCIA method to use. The challenge of this phase is thereby not to complete it but to do so in an appropriate way.
21
specific ranges of uncertainties for the particular study, procedural approaches to evaluate the sensitivity can be considered to be more appropriate. Debatable is whether such approaches would suffice, but again, that is depending on the purpose of the LCA study.
Providing an evaluation on the reliability of the results can be done in different ways and which aspects to consider may vary depending on what kind of LCA is done. Often though, the primary aspect to evaluate is the sensitivity of the identified key issues [16]. However, when an LCA is conducted for the first time on a particular type of system, other aspects such as the influence of the chosen LCIA method might also play a major role. In comparison, studies for which similar systems have been investigated and published, other aspects such as assessing the influence of the system boundaries may be of higher importance. Thus, the context the current study is made within can affect the aspects needed to consider when evaluating the reliability. The importance is though that the issue of reliability is discussed and evaluated.
When evaluating the feasibility of the completeness check, the question is how it is possible to judge whether the data collected is sufficient. To perform the completeness check, investigating the data collected is necessary to identify possible data gaps, and while observing this, to check whether the including parts have been modeled in the same way or if there are inconsistencies. This observation is useful when evaluating the consistency. If data gaps are found, evaluating the potential influence of these missing parts is important. This could be done by comparing how the overall impact is distributed over the various parts. The comparison can provide an indication of how the missing parts would affect the overall results, if it is known what the missing parts are. A crucial element of the evaluation is considered to be understanding what the missing parts are, and how they would influence the overall impact. The amount of material is one aspect to consider, the type of material is another, and the processes and transportation associated are also aspects which might have influence. Going over the data collected is very feasible as the practitioner who collected it is familiar with how the data is organized. To increase the level of objectivity though, it might be a good idea for another person with relevant knowledge to be part of the assessment. Involving more people, perhaps even external parties, would however require more resources which is an aspect to take into account when considering using this additional review.
For the complete evaluation of the consistency, not only how the included parts were modeled matter. Considerations should also be made in regard to the data quality, in terms of the origin of the data used and its accuracy to give a few examples. Evaluating the consistency is seen as feasible as long as the documentation has been thorough throughout the study.
22 3.3.2 Case Study Phase 3
The impact assessment method chosen for this LCA is the ILCD Midpoint+ method, version V1.07. Through SimaPro, a number of methods are available, however, only a few include tungsten in the characterization. The choice of method is therefore based on the following two criteria, inclusion of tungsten in the characterization, and, on the highest number of included substances in the characterization. This is determined by using a function in the software, the Checks tab, listing all excluded substances.
Although results of an LCA can include a few of types of results, as each of the four phases can be considered to give results with varying content, the more interesting results are what the collected information mean, in respect of what the potential environmental impacts actually are. This is what is presented in the following sections. The units used in the charts presenting the results are the units used by the ILCD 2011 Midpoint+ method, which may not be the units one would expect. The potential environmental impact can be modeled as a single score, giving an overall environmental load, as well as on the level of the individual impact categories. The flow of the environmental load as a single score of Leksell Gamma Knife® Icon™ is shown as a network tree in Figure 6. Only top processes are shown and a cutoff of 1.5% was applied, meaning processes contributing with less than that are not displayed. The network shows that the major causes to the environmental load are the processes for the materials cast iron, tungsten, and lead. It is not possible to understand in what ways the system effects the environment from the network below, purely where the causes originate from.
23
The results of the impact assessment show the effect on each impact category and are displayed in Figure 7. The two categories with the largest scores are mineral, fossil, and renewable resource depletion, and, human toxicity, cancer effects. It also shows that the electricity, modeling the use phase, is negligible in comparison to the resources required to produce the Leksell Gamma Knife® Icon™.
Figure 7. Impact assessment using the method ILCD 2011 Midpoint+ V1.07. 3.3.3 Case Study Phase 4
The key issues are identified by contribution analyses while the completeness and sensitivity are assessed by a procedural approach. A literature review is conducted as a part of this approach. Examples of included parts in the review are manuals provided by SimaPro regarding ecoinvent
24
background information and the Worldsteel LCA Methodology Report [43]. Also, the Life Cycle Assessment Specialist for the World Steel Association was consulted. The evaluation of the sensitivity is not conducted in the common approach with Monte Carlo simulations nor the use of parameters as switches, as these features are not included in the educational license used for the software. The lack of these types of evaluations is estimated to not have a great impact though, as the uncertainties of the data collected are considered to be greater than that of the data used from the database. For this reason, the evaluation of the sensitivity focuses on investigating the external validity of the processes used which contributes the most to the environmental load. The consistency is also assessed by a procedural approach by evaluating the consistency in the sources used to obtain data, as well as the methodological choices throughout the project. The latter includes evaluating whether the assumptions made in the LCA are consistent throughout the project and in line with the goal and scope. From the impact assessment, it is clear the categories with the largest impacts are the mineral, fossil, and renewable resources depletion, and, the human toxicity, cancer effects. Through contribution analyses, the underlying processes responsible for the scores in each category, as well as the overall potential environmental impact, are identified. Following the presentation of the key issues, are the evaluations of the sensitivity, completeness, and consistency. The influence of the choice of LCIA method and the appropriateness of used definitions is also discussed, leading to the conclusions of this LCA study.
3.3.3.1 Key issues
The chart in Figure 8 presents the individual processes contributing with more than 1% to the overall potential environmental impact of the Leksell Gamma Knife® Icon™. Processes contributing with less are presented together in the last bar, representing the remaining processes. The chart shows that the three major contributing processes are the production of tungsten, the landfilled slag from the electric arc furnace process and the production of lead concentrate.
25
The process contribution to the impact category mineral, fossil, and renewable resources depletion, is shown in the chart in Figure 9. It shows that the primary production of tungsten and lead concentrate are the two most influential processes. Processes contributing with less 2% are not specified but included in the last bar representing the remaining processes.
Figure 9. Process contribution to the impact category mineral, fossil and renewable resource depletion.
The process contribution to the impact category human toxicity, cancer effects, is shown in the chart in Figure 10. Processes contributing to less than 1% are not specified but included in the last bar, representing the remaining processes. The chart shows that the process having the greatest impact is the process representing the landfilled slag from the electric arc furnace, EAF, and will hereon forward be referred to as the landfilled EAF slag.
26
Figure 10. Process contribution to the impact category human toxicity, cancer effects.
3.3.3.2 Sensitivity
Many choices are made in the LCA. The choices contributing the most to the results are probably the matching of the collected data to the data used from the database. The effect is a discrepancy between the processes used in the model and those in reality. For this reason, processes contributing the most to the impacts are investigated in respect of how well they match the real processes. The first process to be investigated is the landfilled EAF slag. The process is included in the inventory as it is linked to the process of cast iron, an abundant material in the product. The main question is whether the process is representative for the production of the object of this study. From the literature review, it seems unlikely this process is part of the production of the object of this study. According to the World Steel Association and the organization Euroslag, slag from the EAF process should be considered as a by-product rather than scrap, since the slag is used as input in other processes [43] [44]. The slag is suggested to be used in road construction as inputs for road surface materials [43] [44] [45] [46]. The slag should thereby not be modeled as a landfilled waste. Instead, the World Steel Association recommends using system expansion, an allocation method, with gravel production as the avoided production [43]. However, in ecoinvent, the gravel production also includes the same treatment of slag from the same process, meaning there would
27
not yield a different result. Other alternative processes for modeling the slag as road surface materials also includes the process of landfilled EAF slag, and would thereby also give similar impacts. Emission of chromium (VI) to water is identified as the reason the process of landfilled EAF slag provides a high score to the impact category of human toxicity, cancer effects.
From the literature, it is evident that road construction materials using EAF slag as input, also give off leachates, meaning release of chemicals due to water passing through the material. A study investigating the leaching from road surface materials, which incorporated EAF slag, suggests that the leaching may be acceptable, except for chromium, and that the leaching generally is lower than if the EAF slag had been landfilled [45]. It is unclear though what the differences in leaching would be, and the significance of it, if the slag is landfilled or used as input for road construction, though both processes seem to give off chromium as a leaching substance.
The second process to be investigated is the one representing the tungsten production. Since the database ecoinvent did not incorporate any such process, it was created in this project. The LCA Discussion List was advised for the modeling, which resulted in a recommended article already obtained in the literature review, as well as the supporting information to the article, which was very useful. Other processes predefined in ecoinvent has links to associated processes which provides transparency. This was not possible to obtain for the modeling of tungsten, as the limited information obtained was aggregated. The impact this process thus has comes with greater uncertainties than those predefined in ecoinvent. An additional factor influencing the uncertainties is the likely difference in the applied allocation procedures. The impact the tungsten production thereby contributes with is uncertain though through the literature review additional information was gained which is used in the interpretation. To begin with, tungsten has been identified to be a less toxic material than other metals, though there is still a lack of knowledge of its potential environmental effects [47] [48]. In April 2017, an LCA study of tungsten carbide powder production was published demonstrating that the three impact categories with the highest scores were carcinogens, freshwater ecotoxicity and fossil depletion [49]. The contribution of the study to the building of a life cycle inventory database for the tungsten industry shows the novelty in learning what the environmental impacts are regarding this material, and, that further investigations are required to be certain of what the actual environmental impacts are of the material. The article also refers to tungsten as a rare material [49], which suggests the amounts used in the object of this study, would have a great impact on the resource depletion.
Another factor to consider is that the process represents the primary production of tungsten, and does thereby not consider the recycling of the material. Recycling rates are increasing and the global rate was estimated to be around 30% in 2010 [47]. Even though the waste management is not included in this study, the input of recycled materials should be considered. Overall, the processes used from ecoinvent has been market activities to a great extent, meaning many materials have been modeled using predefined ratios of primary and secondary materials as inputs. If the recycling rate of tungsten would have been included, it would have contributed to a beneficial effect on the resource depletion, meaning a slightly lower score on the potential impact. However, a product weighing almost 19 ton will use a considerable amount of material, and thereby likely to have a great impact on the resource depletion.
