Life cycle assessment of bridges, model development and case studies
GUANGLI DU
Doctoral Thesis
Stockholm, Sweden 2015
TRITA-BKN. Bulletin 129, 2015 ISSN 1103-4270
ISRN KTH/BKN/B-129-SE
Akademisk avhandling som med tillstånd av Kungliga Tekniska högskola framlägges till offentlig granskning för avläggande av teknologie doktorsexamen i bro- och stålbyggnad måndagen den 30 mars 2015 kl 10:00 i sal Kollegiesalen, Kungliga Tekniska högskola, Brinellvägen 8, Stockholm.
© Guangli Du, March 2015
Tryck: Universitetsservice US-AB
i
Abstract
In recent decades, the environmental issues from the construction sector have attracted increasing attention from both the public and authorities. Notably, the bridge construction is responsible for considerable amount of energy and raw material consumptions. However, the current bridges are still mainly designed from the economic, technical, and safety perspective, while considerations of their environmental performance are rarely integrated into the decision making process. Life Cycle Assessment (LCA) is a comprehensive, standardized and internationally recognized approach for quantifying all emissions, resource consumption and related environmental and health impacts linked to a service, asset or product. LCA has the potential to provide reliable environmental profiles of the bridges, and thus help the decision- makers to select the most environmentally optimal designs. However, due to the complexity of the environmental problems and the diversity of bridge structures, robust environmental evaluation of bridges is far from straightforward. The LCA has rarely been studied on bridges till now.
The overall aim of this research is to implement LCA on bridge, thus eventually integrate it into the decision-making process to mitigate the environmental burden at an early stage. Specific objectives are to: i) provide up-to-date knowledge to practitioners; ii) identify associated obstacles and clarify key operational issues; iii) establish a holistic framework and develop computational tool for bridge LCA; and iv) explore the feasibility of combining LCA with life cycle cost (LCC).
The developed tool (called GreenBridge) enables the simultaneous comparison and analysis of 10 feasible bridges at any detail level, and the framework has been utilized on real cases in Sweden.
The studied bridge types include: railway bridge with ballast or fix-slab track, road bridges of steel box-girder composite bridge, steel I-girder composite bridge, post tensioned concrete box-girder bridge, balanced cantilever concrete box-girder bridge, steel-soil composite bridge and concrete slab-frame bridge. The assessments are detailed from cradle to grave phases, covering thousands of types of substances in the output, diverse mid-point environmental indicators, the Cumulative Energy Demand (CED) and monetary value weighting. Some analyses also investigated the impact from on-site construction scenarios, which have been overlooked in the current state-of- the-art.
The study identifies the major structural and life-cycle scenario contributors to the selected impact categories, and reveals the effects of varying the monetary weighting system, the steel recycling rate and the material types. The result shows that the environmental performance can be highly influenced by the choice of bridge design. The optimal solution is found to be governed by several variables. The analyses also imply that the selected indicators, structural components and life-cycle scenarios must be clearly specified to be applicable in a transparent procurement.
This work may provide important references for evaluating similar bridge cases, and identification of the main sources of environmental burden. The outcome of this research may serve as recommendation for decision-makers to select the most LCA-feasible proposal and minimize environmental burdens.
Keywords: Sustainable construction; Life cycle assessment; LCA; Global warming; Bridge
management; CO
2emissions; Cumulative energy demand
ii
iii
Acknowledgements
This research project was conducted at the Department of Civil and Architectural Engineering, KTH Royal Institute of Technology, with the financial support from the Swedish Transport Administration (Trafikverket), ETSI Project and the Division of Structural Engineering and Bridges at KTH.
I would like to express my deepest gratitude to my supervisor Professor Raid Karoumi and co- supervisor Professor Håkan Sundquist for providing me the opportunity to work on this project.
I highly appreciate their professional guidance and constant support, which made this research productive. Special thanks to Jean-Marc Battini, who helped reviewing this thesis with valuable comments and suggestions. I sincerely acknowledge Adjunct Professor Lars Pettersson for his discussions and detailed comments in paper revision. Special gratitude is sent to Paul Holmgren, Costin Pacoste and Christoffer Svedholm from the consultant company ELU, for their help in bridge information. I highly appreciate the help provided by the company PEAB, Skanska and ViaCon on the data collection from the construction sites. I am grateful for the comments given by Otto During from the Swedish Cement and Concrete Research Institute. I would also like to thank for the encouragement from Visiting Professor Weiwei Guo and Professor Yongming Tu.
Furthermore, I owe my sincere appreciation to my friends, and all the colleagues at the Division of Structure Engineering and Bridges, who accompanied me with the enjoyment during the past years.
Finally, I give my heartfelt appreciation to my beloved parents for their invaluable love and
support.
iv
v
This doctoral thesis presents the research that the author has carried out in the past 5 years with 80% dedication. The rest 20% effort is devoted on the teaching work, including 4 master theses supervision, participation as teaching assistant during 8 study terms for 4 master-level courses.
The course covers the subject of advanced structural engineering, structural dynamics, concrete and steel structures, and the bridge design. This research is based on the work presented in the following six publications:
List of publications
Paper I: Guangli Du and Raid Karoumi (2014) "Life cycle assessment framework for railway bridges: literature survey and critical issues." Published by the Journal of Structure and Infrastructure Engineering, 10(3), pp. 277-294.
Paper II: Guangli Du and Raid Karoumi (2013) "Life cycle assessment of a railway bridge:
comparison of two superstructure designs." Published by the Journal of Structure and Infrastructure Engineering, 9(11), pp. 1149-1160.
Paper III: Vincent Thiebault, Guangli Du, Raid Karoumi (2013) "Design of railway bridges considering LCA." Published by the Journal of ICE Bridge Engineering, 166(4), pp. 240- 251.
Paper IV: Guangli Du, Mohammed Safi, Lars Pettersson, Raid Karoumi (2014) "Life cycle assessment as a decision support tool for bridge procurement: environmental impact comparison among five design proposals." Published by the International Journal of Life Cycle Assessment, 19(12), pp. 1948-1964.
Paper V: Mohammed Safi, Guangli Du, Raid Karoumi and Håkan Sundquist (2015) "Holistic approach to sustainable bridge procurement considering LCC, LCA, User-cost and Aesthetics", submitted.
Paper VI: Guangli Du, Lars Pettersson and Raid Karoumi (2015) "Life cycle environmental impact of two commonly used short span bridges in Sweden", manuscript.
The first author was responsible for data processing, model analysis and writing in Paper I, II, IV
and VI. In Paper III, Guangli initiated, planned and revised the paper. In Paper V, Guangli
performed the LCA analysis and wrote the LCA sections. In Paper VI, the first author and the
second author had worked closely on data collection. All the authors had participated in planning
the paper and contributed in the revision.
vi
vii
Besides, the author has also contributed in the following additional publications:
Guangli Du (2012), "Towards sustainable construction: life cycle assessment of railway bridges", Licentiate thesis in the Division of Structural Engineering and Bridges, Department of Civil and Architectural Engineering, KTH Royal Institute of Technology, Stockholm, Sweden.
Guangli Du and Raid Karoumi (2013), "Environmental life cycle assessment comparison between two bridge types: reinforced concrete bridge and steel composite bridge", 3rd International conference on Sustainable Construction materials and Technologies (SCMT3), Japan Concrete Institute, Kyoto, Japan.
Barbara Rossi, Ivan Lukic, Naveed Iqbal, Guangli Du, Diarmuid Cregg, Ruben Paul Borg, Peer Haler (2011), "Life cycle impacts assessment of steel, composite, concrete and wooden columns", COST Action C25: Proceedings of the international conference sustainability of constructions-towards a better built Environment. Innsbruck, Austria.
Guangli Du and Raid Karoumi (2012), "Environmental comparison of two bridge alternative designs", fib symposium Stockholm, Concrete Structure for Sustainable Community, pp.353-356, Stockholm, Sweden.
Guangli Du (2010), "A literature review of life cycle assessment for bridge infrastructure", COST
Action C25: Sustainability of Constructions: An Integrated Approach to Life-time
Structural Engineering, Malta.
viii
ix
Contents
Abstract ... i
Acknowledgements ... iii
List of publications ... v
Contents ... ix
CHAPTER 1 INTRODUCTION ... 1
Background ... 1
1.1 Aim and scope ... 1
1.2 Research contribution ... 1
1.3 Outline of the thesis ... 2
1.4 CHAPTER 2 LIFE CYCLE ASSESSMENT ... 3
Life cycle thinking ... 3
2.1 Brief history of LCA ... 3
2.2 Main steps of LCA ... 4
2.3 2.3.1 Goal and scope definition phase ... 4
2.3.2 Inventory analysis ... 5
2.3.3 Impact assessment ... 6
2.3.4 Interpretation ... 8
CHAPTER 3 BRIDGE LCA MODEL AND THE DEVELOPED TOOL ... 11
The current Swedish bridge stock and BaTMan ... 11
3.1 3.1.1 The feasibility of integrating LCA into BaTMan ... 11
3.1.2 A systematic Bridge LCA model ... 12
3.1.3 Monetary evaluation of environmental impacts ... 12
An LCA based computational tool: GreenBridge ... 13
3.2 3.2.1 How to use GreenBridge ... 15
CHAPTER 4 SUMMARY OF THE APPENDED PAPERS ... 19
Current research status and literature review (Paper I) ... 19
4.1 4.1.1 Current status of Bridge LCA research ... 19
4.1.2 Literature review ... 20
Application of the Bridge LCA model to real cases ... 24
4.2
4.2.1 The Banafjäl Bridge (Paper II and Paper III) ... 24
x
4.2.2 The Karlsnäs bridge (Paper IV) ... 25
4.2.3 Attempt to integrate LCC, LCA, lifespan, user-cost and aesthetics (Paper V) ... 26
4.2.4 Two commonly used short span bridges in Sweden (Paper VI) ... 26
CHAPTER 5 CONCLUSIONS AND FUTURE RESEARCH ... 27
General conclusions ... 27
5.1 Future research ... 28
5.2 REFERENCES ... 29
PAPER I ... 37
PAPER II ... 57
PAPER III ... 71
PAPER IV ... 85
PAPER V ... 105
PAPER VI ... 137
1
CHAPTER 1 INTRODUCTION
Background 1.1
The construction sector is attracting increasing attention due to its high environmental burden. Currently, the Swedish Transport Administration (Trafikverket) owns more than 29000 bridges. With the rapid development of infrastructure, the sustainability and environmental performance of these bridges is raising concerns from the public, stakeholders and authorities. Bridges, as the fundamental structural elements of transportation networks, not only have long life spans, but also consume large amounts of natural resources and energy in their construction and maintenance. For instance, the greenhouse gases generated from producing a cubic meter concrete has estimated equivalent to a person traveling 1000 km by car (Ecoinvent v2.2) or 4000 km by an Airbus A320 (Jardine, 2009). Consequently, in addition to economic and technical aspects, there are strong motivations to increase the focus on environmental sustainability for bridges.
Life Cycle Assessment (LCA) is a comprehensive, standardized and internationally recognized approach for quantifying all emissions, resource consumption, related environmental and health impacts linked to a service, asset or product (Treloar, et al., 2000; ISO 14040, 2006; ILCD, 2010). Although LCA has a broad application in various industries, its implementation on bridges is rare and needs more investigation. Most previous studies have only considered few indicators or structural components, or a specific life stage (Du and Karoumi, 2014). For example, the pioneer study by Widman (1998) confined the study scope on three selected air emissions of CO
2, CO and NO
x; Itoh and Kitagawa (2003), Martin (2004), Collings (2006) and Bouhaya et al. (2009) focused on the energy consumption and CO
2emissions; Itoh and Kitagawa (2003) excluded the end of life phase (EOL), while Bouhaya et al. (2009) omitted the substructure of foundation. Due to the complexity of the environmental problems and bridge structures, robust environmental evaluation of bridges is far from straightforward.
Aim and scope 1.2
The overall aim of this research is to implement LCA on bridge, thus eventually integrate it into the decision-making process. Specific objectives are to: 1) investigate the current state of the art and identify the critical issues; 2) recommend an operational framework to the practitioners; 3) develop a computational tool to facilitate the calculation; 4) apply the developed framework on various types of bridges; and 5) explore the feasibility to combine LCA with LCC.
The study covers various types of bridges, including railway bridge with ballast or fix-slab track, road bridges of steel box-girder composite bridge, steel I-girder composite bridge, post tensioned concrete box-girder bridge, balanced cantilever concrete box-girder bridge, steel-soil composite bridge and concrete slab-frame bridge. Three LCA methodologies (CML 2001, Eco-indicator 99’ and ReCiPe) have been applied. A series of environmental impact indicators, ranging from the pollutant level to the mid- point impact indicators, and two monetary evaluation approaches, have also been applied. The data are retrieved from Ecoinvent v2.2, ELCD, World steel and U.S. LCI.
Research contribution 1.3
This research has made several contributions to the application of LCA to bridge structures, as
summarized below:
CHAPTER 1 INTRODUCTION
2
• A thorough literature review related to LCA implementation for bridges has been performed.
The review addressed a series of problems and possible future research needs. It also compared diverse LCA methodologies and databases.
• A comprehensive LCA framework has been developed, to enable practitioners to make decisions in early planning stages that enhance bridges’ environmental sustainability.
• A Matlab-based calculation tool has been developed for assessing both road and railway bridges.
The tool enables simultaneous analysis and comparison of 10 feasible bridge designs at any detail level. It can be applied to 27 types of impact indicators, covering all bridges’ life stages and their entire structural components. Numerical results are automatically presented in Excel files and figures in Word documents. The dominant structural component/scenario in terms of environmental performance can also be identified.
• The feasibility of applying LCA to various types of roadway bridges and railway bridges has been explored, with illustrative real cases at various levels of detail.
• To enhance bridges’ environmental sustainability, an approach for integrating LCA with Life Cycle Cost (LCC) analysis in current decision-making processes has been developed, verified by application to real cases that representing the most common bridge types in Sweden.
• Particular attention has been paid to LCA-related aspects of the construction phase, for these two types of commonly used short-span bridges in Sweden, thus to address a major gap in previous research.
Outline of the thesis 1.4
This thesis is structured in two parts: the extended summary, and the six appended papers. Part one
provides supplementary description for this work, and consists of five Chapters: Chapter 1 introduces
the research background and the scientific contribution. Chapter 2 presents the theoretical methodology
of LCA. Chapter 3 illustrates the recommended LCA framework for bridges, and the developed
calculation tool based on it. Chapter 4 summarizes the appended papers in order, including the current
research status, literature review, and a series of practical cases explored in this work. Finally, Chapter 5
gives the general conclusions and the future research needs.
3
CHAPTER 2 LIFE CYCLE ASSESSMENT
A detailed background theory in terms of LCA framework, life cycle impact assessment (LCIA) methods survey, a number of life cycle inventory (LCI) database and software descriptions are presented in the appended Paper I. Therefore, this Chapter only provides the supplementary knowledge beyond that.
Life cycle thinking 2.1
Sustainability is not easy to measure, but if there is a solution, it will be based on methods derived from life cycle thinking (LCT) (Klöpffer W, 2003). For keeping the sustainability balance among the economy, environment and the society, a variety of LCT based approaches are propagated: life cycle management (LCM), life cycle cost (LCC), and life cycle assessment (LCA). LCT is a systematic approach intends to consider the products from ‘cradle’ to ‘grave’, from the raw material extraction, through products manufacture, use and maintenance until the final disposal. Decisions are needed in each life stage of the product, the application of LCT can thus help connect all environmental and social issues in an integrated holistic system, meanwhile avoid the short term policy resulting in the long term unbalanced development, or transferring the burden from one stage to another, such as saving money/resource/environmental impact in the initial production phase but leading to the increase in another phase. Several common LCT terms can be defined as follows:
i. Life cycle assessment (LCA) is an approach for environmental evaluations, based on LCT.
LCA is a comprehensive and internationally recognized approach for quantifying all emissions, resource consumption and related environmental and health impacts linked to a service, asset or product (Treloar, et al., 2000; ISO, 2006; ILCD Handbook, 2010).
ii. Life cycle cost (LCC) is an economic concept based on LCT, which takes account of all the monetary costs of a product or service from ‘cradle’ to ‘grave’. Besides, the equivalent monetary value of LCA can also be transformed and combined into LCC.
iii. Life cycle management (LCM) is the application of LCT to modern business practice, with the aim to manage the total life cycle of an organization’s product and services toward more sustainable consumption and production (SAIC 2006).
Brief history of LCA 2.2
Even though today’s LCA has been involved in diverse industrial sectors, with various tools and methodologies formulated, the application of LCA is still new in history. In the 70s, the initiation of LCA concept mainly comes from the oil crisis and energy shortage. According to Hunt and Franklin (1996), the first LCA study was performed by Harry Teasley at Coca-Cola Company in 1969, for the purpose of quantifying the energy, material and environmental burdens from the beverage packages.
Since then, the LCA studies were steadily conducted under various names, such as resource and environmental profile analysis (REPA), integral environmental analysis, environmental profiles.
According to Baumann and Tillman (2004), LCAs published between 1969 and 1972 were all limited to packaging and waste management issues, and solid waste was the main concern rather than energy consumption and emissions.
In the early 1970s, the first LCA computer program was developed in the USA. Meanwhile, the
American LCAs projects inspired the identical ideas in Britain and Germany, which was later followed
CHAPTER 2 LIFE CYCLE ASSESSMENT
4
by Sweden. Between 1975 and 1988 there were still very few public documents about LCA. In 1990, the term ‘LCA’ was first coined during the workshop organized by the Society of Environmental Toxicology and Chemistry (SETAC). Soon the first complete LCA methodology appeared in a peer reviewed scientific journal in 1992 (Hunt and Franklin, 1992). In addition to SETAC, from the 1990s onwards, the International Organization for Standardization (ISO) has developed global standards (ISO14040 to ISO14044) in efforts to streamline and harmonize LCA guidelines (Fava, 2011). Various studies, software tools and methodologies were also developed after the 1990s. Table 2.1 presents the main LCA developments between the 1970s and 2006.
Main steps of LCA 2.3
As already mentioned, LCA is a systematic method for quantifying the potential environmental impacts of a product, asset or service throughout its whole life cycle, from raw material acquisition, through manufacture, use and maintenance until the end of its life (Baumann and Tillman, 2004). The potential environmental categories include resource depletion, human health, and ecological health (ISO14040, 2006; ILCD 2010). The LCA process can be used to determine the potential environmental impacts of any product, process, or service (ILCD, 2010). The application of LCA can provide scientifically based results for decision-makers, thus helping efforts to set new criteria for environmentally friendly design, compare and choose environmentally competitive products, and identify environmental deficiencies of a product for further optimization. As stated in the ISO LCA standards, the general LCA framework consists of the following four phases:
1) Goal and Scope definition defines the aims, product system, and expected result of the study;
2) Inventory analysis quantifies all the emissions related to the product system based on the functional unit of the product.
3) Impact assessment transforms the inventory result into the environmental impact categories.
4) Interpretation explains the results with the goal of the study through the whole analysis procedure.
2.3.1 Goal and scope definition phase
An LCA starts with the goal and scope definition phase, in which the object and scope of the study, the
purpose and expected results, the functional unit and relevant assumptions are all defined. This phase is
the most important and mandatory part for every LCA study, not only because the stated definitions
will affect the course of the entire study, but also because it is essential for clear dissemination of results
and conclusions following completion of the study (Guinée, 2001). In order to obtain valid results, the
relevant parameters and adopted perspectives need to be clarified here. The example of the goals might
be given as: ‘identification of the product component with the most severe environmental impact,
characterization of effects of changing the design of an element on overall environmental performance,
and/or determination of means to optimize a product’s environmental performance.’ Once appropriate
goals have been identified, it is important to determine the types of information required to answer the
questions (SAIC, 2006; ILCD, 2010). For definition of scope, a well-designed flowchart is generally
helpful to ensure that all of the activities that may affect the whole system are covered.
2.3 MAIN STEPS OF LCA
5
Table 2.1 Brief history of LCA development
1969 Original LCA study conducted by Coca-Cola Company. (Hunt & Franklin, 1996) 1970 Pioneering studies of REPA (early name of LCA). (Hunt & Franklin, 1996)
1972 Dr. Boustead wrote teaching text on silica glass production in the UK (Baumann and Tillman, 2004)
1973 The first LCA computer program was funded. (Hunt & Franklin, 1996)
1974 First complete LCA report regarding data and methodology published by the US Environmental Protection Agency (EPA). (Hunt & Franklin, 1974)
1979 SETAC (Society for Environmental Toxicology and Chemistry) founded.
1980 Public domain of a comprehensive peer reviewed LCA database provided by Solar Energy Research Institute (Bider et al., 1980; Hunt & Franklin, 1996)
1975-1988 Low public interest in LCA.
1990 International forum of REPA in U.S. (Hunt & Franklin, 1996) 1990 The term of ‘LCA’ was coined in a SETAC workshop.
1992 Publication of the first complete LCA methodology in a peer-reviewed scientific journal (Hunt & Franklin, 1992).
1993-1996 ISO14040 standard for LCA-principles and framework was accomplished.
2002 Launch of the International Life Cycle Partnership by UNEP and SETAC to promote LCT worldwide, rather than regionally.
2006 ISO 14040 (1997) extended with several additional standards, to ISO 14040 (2006).
Afterwards LCA was exploited in use in various extended fields.
The functional unit is another important definition. It provides an equivalent basis that all the material and energy flows would refer to. It must represent the function of the compared options in a reasonably fair way (Baumann and Tillman, 2004), as each of the calculated material and energy flows must be consistently based upon it. An appropriate functional unit of a steel girder bridge might be ‘during a 120 years life span with an annual traffic volume of 20312 tkm train freight.’ The functional unit is most important when multiple products are compared, for instance we cannot directly compare a 20 m long steel bridge with a 50 m long concrete bridge directly.
2.3.2 Inventory analysis
The result of LCA is heavily dependent on the input data (Du and Karoumi, 2014). A number of data
can be obtained from the life cycle inventory (LCI) database, which takes account of the inputs and
outputs inventories related with the product. It requires the collection of numerous data both regionally
and globally. LCI is used to assist the calculation for the quantities of resources, emissions and waste
CHAPTER 2 LIFE CYCLE ASSESSMENT
6
generated per functional unit (Rebitzer et al., 2004). This process considers the energy and raw material as input into the model, and the products, environmental releases as output in the form of pollutants (CO
2, NH
3, CO, P, NO
2) into air, water and soil. The inventory data of the energy, transportation, material consumption and waste treatment covers a wide range of sources, including manufacturing companies, government bodies and scientific journals. However, LCI data are largely affected the regional conditions, the process technologies applied and variations in data from different sources. For instance, the CO
2emission of normal concrete from Ecoinvent v2.2 is 2.5 ×10
5g/m
3while the concrete for road construction in Stripple (2001) is 3.3 ×10
5g/m
3. This is mainly due to differences in technological processes and concrete properties. In addition, a number of current available LCI databases are listed in Table 1 in Paper I.
SAIC (2006) provides the following framework for performing a life cycle inventory analysis:
1. Based on the goal and scope definition, develop a flow chart of the processes that under evaluation. The flow chart diagram is established by covering all of the activities and steps considered in the whole system, with the allocation of the interlinking relations between activities. The more complex and detailed the flowchart is, the more accurate the results that can be obtained.
2. Create data collection plan based on the goal definition, data sources availability, and data quality indicators. The level of the accuracy and potential users of the data can largely affect this process. Diverse sources can be used, which should overcome problems related to geographical, timeframe or technological differences. Site- specific data are always preferable to the average data, but difficult to obtain.
3. Collect the data. Data compilation is one of the most work intensive steps in LCA.
The information and data collected during this stage assures the data quality and final result, thus it is important to keep the procedure transparent and clear.
4. Evaluate and report results. The result of LCI is usually a long list contains a numerous emissions in terms of water, air and solid. The final environmental profile of the product highly related with the quality of the LCI data. Thus, the report should clearly document the data in terms of the methodology, system boundary and assumptions.
2.3.3 Impact assessment
The life cycle impact assessment (LCIA) is the evaluation process of the potential human health and environmental impacts, from the environmental resources and releases identified during the LCI (SAIC 2006; ILCD 2010). The magnitude and significance of environmental or social costs associated with specific life cycle activities are identified during this phase (Pennington et al. 2004; Pelletier et al. 2007).
The LCIA process follows the guideline of ISO standards, transforms the emission flows from the life
cycle inventory level into the intuitive impact categories, by focusing at either the problem-oriented level
(midpoint) or the damage-oriented level (endpoint). The midpoint level aimed to interpret the complex
emission list into the easier and more common accepted group of impact categories (Global warming,
Acidification, Abiotic depletion). Endpoint level focuses on broader overall effects, such as
consequences of a process or manufacture of a given product for human health, ecosystem quality or
resource depletion.
2.3 MAIN STEPS OF LCA
7
LCIA is deemed the most time-consuming stage of LCA. Paper I listed numerous LCIA methods which have been developed by different research institutes and organizations, such as ReCiPe, CML 2007, Eco-indicator 99, EDIP 97, EDIP 2003, EPS 2000, Impact 2002+, JEPIX, LIME, TRACI and IPCC (for global warming). Specific instructions for implementing each LCIA methodology are provided in various publications, such as CML (2007), Goedkoop et al., (2009), IPCC (2013), Frischknecht et al., (2007) and Bare & Gloria (2006); SBRI (2013). They pointed that each method is actually a collection of impact assessment methods for the individual impact categories (e.g., stratospheric ozone protection, human health, etc.), see Table 2.2.
The lack of a standard LCIA methodology is still one of the obstacles encountered by LCA practitioners.
Although those LCIA methods are developed by following the same principles and frameworks from ISO standards, they still differ from the aspects of category groups, orientation levels (midpoint or endpoint), elementary LCI emissions and the covered LCIA steps (normalization and weighting). It has been discussed in several literatures that different LCIA methods may lead to different results. For instance, Amy and Thomas (2008) compared several LCIA methods regarding the bio-fuels, they pointed out that there is no exactly ‘one right LCIA method’, since each of them has been peer reviewed and evaluated for accuracy and internal consistency. The selection of LCIA method is largely depending on the individual study goal and scope.
Guinée (2001), SAIC (2006) and ILCD (2010) have provided general instructions for performing LCIA process, which consist of the following steps:
Select the impact categories Impact category selection is a key element of any impact assessment. In every LCA it is necessary to consider which environmental impacts to take into account. The ISO standard only gives general headlines for impact categories of resource use, ecological consequences and human health (ISO 14040 1997). However, they must be interpreted in terms of more operational impact categories such as global warming, acidification and resource depletion, etc. (Baumann and Tillman 2004). The most appropriate set of impact categories to include is strongly affected by their relevance to the goals and scope of the LCA, the LCIA methodology selected, and the availability of suitable inventory data. In some cases, the compiled information and inventory data are very limited and incomplete.
Classification sorts and assigns the LCI result parameters into a various commonly acknowledged impact categories (Baumann and Tillman 2004). According to SAIC (2006) there are generally two ways of assigning LCI results into the multiple impact categories: partition them into the impact categories to avoid the ‘double counting’ when they affect each other; or assign them to all impact categories when the effect are independent of each other. For example: NO
xcan be fully assigned into the category of Acidification as well as another category of Eutrophication (Baumann and Tillman 2004).
Characterization refers to the quantification of a number of chemicals in terms of an equivalence scale to determine their contributions to the overall impact of the focal product or process in a given category.
It involves summing effects of all of the relevant substances, using appropriate characterization factors, e.g. effects of ammonia, hydrogen chloride and nitrogen oxide emissions on acidification. A series of characterization factors and methods have been presented by Guinée et al. (2001) and ReCiPe (2008).
Sleeswijk et al. (2008) noted that only a relatively small proportion of the total number of interventions
is responsible for a large proportion of potential environmental impacts. Although there were 858
environmental interventions collected and investigated, 46 of them account for over 75% of the impact
scores. An example of commonly used impact category indicators (ReCiPe, 2008) is presented in Table
2.3. Impact indicators can be characterized by following the equations below (Baumann and Tillman,
2004; Hammervold J. et al., 2009):
CHAPTER 2 LIFE CYCLE ASSESSMENT
8
𝑒 ij = 𝑥 i ∙ 𝑓 ij
𝑑 k = � (𝑒 ij ∙ 𝑐 jk )
i=0,j=p
i=0,j=0
e
ij= emission of the LCI item j for total consumption of input parameter i;
x
i= consumption of the input parameter i;
f
ij= emission of LCI item j per unit input parameter i;
d
k= total potential impacts in impact category k, expressed in equivalents c
jk= the characterization factor for LCI item j to impact category k
Normalization, grouping and weighting are three optional steps proposed through the LCA framework by ISO standards. Normalization refers to the application of factors, based on appropriate reference values, to impacts in each of the selected midpoint or endpoint categories that allow analysis of their respective contributions to the overall impact in each category (SAIC, 2006). The operational documentation by Guinée, (2001) recommends use of normalization data based on a single geographically and temporally well-defined reference system, preferably the world for one year. A number of normalization methods have been described, for instance by Sleeswijk (2008; 2010), Wenzel et al. (1997), Breedveld et al. (1999), Huijbregts et al. (2003), Strauss et al. (2006), Stranddorf et al. (2005a; 2005b), Bare et al. (2006), Lundie et al. (2007). The equations below express the normalization procedure (Baumann and Tillman, 2004;
Hammervold et al., 2009):
𝑚 k = 𝑑 k × 𝑛 k
m
k= normalized potential impacts for category k n
k= normalization factor for category k
Grouping sorts the characterization results into one or more sets to facilitate the interpretation of the results (Baumann and Tillman, 2004). ISO standards addressed two ways of grouping: 1) sorting indicators by characteristics such as air, water and solid emissions, or local, regional and global locations.
2) sorting indicators by ranking, such as high, low or medium.
Weighting is a qualitative or quantitative procedure where the relative importance of an environmental impact is weighted against all the other, the relative importance are expressed by their weighting factors (Baumann and Tillman, 2004; SAIC, 2006; Guneè, 2001; ILCD, 2010). The importance can be oriented based on monetary values, authorized targets, authoritative panels, proxies, technology abatement, etc.
However, according to ISO standards weighted impacts cannot be used for product comparisons due to the bias they may introduce in results. Since there is no societal consensus regarding these fundamental values, there is no reason to expect consensus either on weighting factors, or on the weighting method or even on the choice of using a weighting method at all (Finnveden G., 1999).
2.3.4 Interpretation
Interpretation refines the numerous LCA results into specific concerns with meaningful conclusions.
ISO 14040 (1997) defines that the interpretation phase of life cycle assessment in which the findings of
either the inventory analysis or the impact assessment, or both, are combined consistent with the
defined goal and scope in order to reach conclusions and recommendations. During this stage, the
limitations, drawbacks and issues of uncertainties should be clearly elucidated.
2.3 MAIN STEPS OF LCA
9
Table 2.2: Impact indicators in different methodologies (Nebel et al., 2009)
ReCiPe
Global warming, Ozone depletion, Terrestrial acidification, Freshwater eutrophication, Marine eutrophication, Human toxicity, Photochemical oxidant formation, Particulate matter formation, Terrestrial ecotoxicity potential, Freshwater ecotoxicity potential, Marine ecotoxicity potential, Ionising radiation potential, Agricultural land occupation potential, Urban land occupation potential, Natural land transformation potential, Water depletion potential, Mineral depletion potential, Fossil depletion potential
CML 2007
Abiotic depletion, Acidification, Eutrophication, Global warming (GWP 100), Ozone layer depletion, Human toxicity, Fresh water aquatic ecotoxicity, Marine aquatic ecotoxicity, Terrestrial ecotoxicity, Photochemical oxidation
Impact 2002+
Global warming, Carcinogens, Non-carcinogens, Respiratory inorganics, Ionising radiation, Ozone layer depletion, Respiratory organics, Aquatic ecotoxicity, Terrestrial ecotoxicity, Terrestrial acidification/nutrification, Land occupation, Non-renewable energy, Mineral extraction
TRACI Global warming (GWP100), Acidification, Carcinogens, Non-carcinogens, Respiratory effects, Eutrophication, Ozone depletion, Ecotoxicity, Smog
Eco-indicator 99 Carcinogens, Respiratory inorganics, Climate change, Radiation, Ozone layer, Ecotoxicity, Acidification, Eutrophication, Land use, Minerals, Fossil fuels
EPS 2000
Life expectancy, Severe morbidity, Morbidity, Severe nuisance, Nuisance,
Crop growth capacity, Wood growth capacity, Fish and meat production,
Soil acidification, Irrigation water, Drinking water, Depletion of reserves,
Species extinction
CHAPTER 2 LIFE CYCLE ASSESSMENT
10
Table 2.3 Mid-point impact indicators (ReCiPe, 2008; Sleeswijk et al., 2008)
Characterisation factor Abbr. Unit References
Global warming potential GWP kg (CO
2to air) IPCC (2013) Ozone depletion potential ODP kg (CFC-11 to air) VMO (1999) Terrestrial acidification potential TAP kg (SO
2to air) Van Zeim et al.,
(2007b, 2007c) Freshwater eutrophication potential FEP kg (P to freshwater) Struijs et al., (2007) Marine eutrophication potential MEP kg (N to freshwater) Struijs et al., (2007) Human toxicity potential HTP kg (1,4 DCB to
urban air) Huijbregts et al., (2005a, 2005b; 2007), van de Meent and Huijbregts (2005), van Zeim et al., (2007a) Photochemical oxidant formation
potential POFP kg (NMVOC to air) Van Zeim et al.,
(2007d) Particulate matter formation
potential PMFP kg (PM10 to air) Van Zeim et al.,
(2007d) Terrestrial ecotoxicity potential TETP kg (1,4 DCB to
industrial soil) Huijbregts et al., (2005a, 2005b; 2007), van de Meent and Huijbregts (2005), van Zeim et al., (2007a) Freshwater ecotoxicity potential FETP kg (1,4 DCB
freshwater) Huijbregts et al., (2005a, 2005b; 2007), van de Meent and Huijbregts (2005), van Zeim et al., (2007a) Marine ecotoxicity potential METP kg (1,4 DCB to
marine water) Huijbregts et al., (2005a, 2005b; 2007), van de Meent and Huijbregts (2005), van Zeim et al., (2007a) Ionising radiation potential IRP kg (U
235to air) Frischknecht et al.,
(2001) Agricultural land occupation
potential ALOP m
2×yr (agricultural
land) De Schrijver and
Goedkoop Urban land occupation potential ULOP m
2×yr (urban land) De Schrijver and
Goedkoop Natural land transformation
potential NLTP m
2(natural land) --
Water depletion potential WDP m
3(water) --
Mineral resource depletion potential MDP kg (Fe) --
Fossil resource depletion potential FDP kg (oil) Frischknecht et al.,
(2007)
11
CHAPTER 3 BRIDGE LCA MODEL AND THE DEVELOPED TOOL
The current Swedish bridge stock and BaTMan 3.1
Bridges are large-scale infrastructures, and require efficient management using Bridge Management System (BMS). BMS serves as a rational and systematic tool to organize and carry out all the relevant activities. BMS can integrate the factors of repair, rehabilitation and economic considerations in a holistic manner. This enables the relevant authority to make optimal decisions and budget allocations from a network perspective. Authorities in many countries have developed a BMS, with various levels of explicit detail, for example, Latvia (Lat Brutus), The Netherlands (DISK), Spain (SGP), Japan (JBMS), Ireland (Eirspan), Germany (GBMS), Finland (FBMS), Denmark (DANBRO), Switzerland (KUBA), and Sweden (BaTMan) (Mirzaei et al., 2012). However, most current BMS do not consider bridges’
environmental performance.
In Sweden, Trafikverket is the major owner of the bridge infrastructures. Since 1944, information about the condition of the national road network has been documented and stored in different archives in Sweden (Hallberg and Racutanu, 2007). Bridge and Tunnel Management (BaTMan) is the Swedish BMS operated by Trafikverket via internet. In 2004 the BaTMan was launched online, which is considered as one of the most comprehensive BMS in Europe (Safi, 2013). The current BaTMan covers approximately 31020 bridges, mostly governed by Trafikverket, including 24320 road bridges, 4542 railway bridges, 1895 pedestrian bridges and 263 bridges of other types, as summarized in Table 3.1 and Figure 3.1 (BaTMan, 2014).
Table 3.1 Swedish bridge stock statistics (BaTMan, 2014)
Bridge types Concrete Steel Wood Stone Other
materials
Slab bridge 5831 48 150 176 71
Beam bridge 3004 2795 248 12 34
Slab fram bridge 9956 2 3 5 2
Beam fram bridge 1120 3 0 2 2
Culvert bridge 314 4184 0 1 27
Arch bridge 595 138 18 1086 9
Cable stayed bridge 2 11 9 0 0
Moveable bridge 0 105 0 0 4
Other bridges 243 112 41 10 6
3.1.1 The feasibility of integrating LCA into BaTMan
This section mainly discusses the feasibility of combining LCA with the current BaTMan, although the real implementation finally requires the legislation and policy support from the authorities. A bridge LCA model encompasses thousands of processes and material types. Lack of input data is one main obstacle hindering LCA implementation (Du, 2012; Du and Karoumi, 2014), but BaTMan stores most of the fundamental data required for LCA analysis. Such information includes bridge identification, location, dimensions, details of bridge structural elements and related material, drawings, traffic data, inspection plans and maintenance history, cost and budget plans, and deterioration predictions.
Collectively, this information provides the required systematic basis for integrating LCA into BaTMan.
CHAPTER 3 BRIDGE LCA MODEL AND THE DEVELOPED TOOL
12
In this sense, it is clearly feasible to integrate LCA into BaTMan, thereby making it a substantially more comprehensive management system.
Figure 3.1 Indicated age classes of bridges in Sweden (BaTMan, 2014) 3.1.2 A systematic Bridge LCA model
The LCA implementation on bridges is very rare comparing with other industrial sectors (Du and Karoumi, 2014). One recognized limitation is the lack of authorized guidelines and unified criteria.
Therefore, this research has developed a practical framework, which may serve as an operational recommendation to the analyst. The framework can either be implemented on the whole bridge or on a specific life cycle stage or only part of the structural components. The LCI database, which comprises thousands sets of upstream and downstream processes, would be assigned into the selected scenarios.
The selected LCIA method is implemented to evaluate the inventory releases in accordance with the ISO standards. It results into the specific impact indicators to the human health, eco-system and resource depletions. The life cycle of the bridge can be divided into four stages: material manufacture stage, construction stage, use and maintenance stage and EOL, see Paper I and Paper VI.
3.1.3 Monetary evaluation of environmental impacts
LCA modelling can result into a wide range of impacts associated with human health, ecosystem quality and resources, which are not straightforward for stakeholder and decision-makers to illustrate and assess at the governing level. In order to comprehensively aggregate the impacts for an intuitive comparable set, weighting is adopted to convert the impacts into monetary values with a common unit. However, weighting of environmental impacts is being debated in LCA. As the ISO14040 standards (2006) and ILCD Handbook (2010) noted, value-based weighting is not permitted for comparative analyses that support decisions in open tendering processes. Nevertheless, Ahlroth et al. (2011) and Ahlroth and Finnveden (2011) observed that weighting is still widely used to meet practitioners and decision-makers need, as illustrated by several authors, e.g. Mahgoub et al. (2010), Contreras et al. (2009), Kiwjaroun et al.
(2009), Liu et al. (2010), Tsoutsos et al. (2010), and Zackrisson (2005); it is recommended to use several weighting sets and compare the outcomes to reduce risks of overlooking important factors. More specifically, Ahlroth et al. (2011) thoroughly discussed the feasibility of evaluating the economic value of environmental impacts in a whole-life perspective. They showed that one way to include external environmental costs in LCC is to use monetary-weighted results obtained from environmental system analysis (such as LCA); several examples of such applications are available in the literature (Carlsson Reich, 2005; Nakamura and Kondo, 2006; Kicherer et al., 2007; Lim et al., 2008; Hunkeler et al., 2008).
In this work, two monetary weighting systems: Ecovalue08 with updated Ecovalue12 weightings
0 1000 2000 3000 4000 5000 6000
2000- 20 14 1990- 19 99 1980- 19 89 1970- 19 79 1960- 19 69 1950- 19 59 1940- 19 49 1930- 19 39 1920- 19 29 1910- 19 19 1900- 19 09 1890- 18 99 Be fo re 1890
Nu mb er o f B rid ge s
Year of Construction
3.2 AN LCA BASED COMPUTATIONAL TOOL: GREENBRIDGE
13
(Ahlroth and Finnveden, 2011; Ahlroth et al., 2011; Finnveden et al., 2013) and Ecotax02 (Finnveden et al., 2006), have been applied and compared. The Ecovalue monetary weighting set has been developed for evaluating mid-point environmental impacts based on willingness-to-pay, with particular focus on Swedish conditions, while the Ecotax set is based on environmental taxes and fees levied by a focal society. Table 2 in Paper IV illustrates these two weighting sets. It should be noted that some impact categories cannot be weighted, due to the limitations of the available weighting factors.
An LCA based computational tool: GreenBridge 3.2
LCA analysis for bridges is time-consuming, costly and requires expert knowledge of both bridge engineering and LCA. Some commercial LCA software is available, but none of them is specific designed for assessing bridges. In order to realize the analysis into daily practice, a Matlab-based LCA tool ‘GreenBridge’ is developed in this study to facilitate the modelling involved. Figure 3.2 shows the front page of GreenBridge.
The GreenBridge tool is particularly designed for bridge structures, following the LCA framework also developed in this study. It enables the modeling of road bridges and railway bridges, with ReCiPe (H) methodology that covers over 1000 substances data extracted from Ecoinvent. A major advantage of GreenBridge is its flexibility; it enables automatic quantification of environmental indicators at any user- defined detail level, and the possibility to detect the relative importance of any scenario activities in any life stage. The database can be flexibly replaced according to the analysts’ desire. Figure 3.3 presents the general structure of GreenBridge, with each step displayed in a separate excel sheet.
Figure 3.2. Front Page of GreenBridge LCA tool
GreenBridge Scope
GreenBridge
Life Cycle Assessment (LCA) Tool for Bridges
Copyright: Guangli Du © all rights reserved Contact Lnfo: guangli.byv@kth.se Supervisors: trofessor Raid Karoumi
trofessor Håkan Sundquist
Bridge life cycle Material manufacture phase Construction phase Use and Maintenance Phase Demolition and waste treatment Raw material extraction
Sorting and processing Transportation Waste disposal and treatment
Resouces Energy Input Data
Ecoinvent v2.2 ReCiPe (H)
AirCWaterCSolid Releases:
SO2, NOx, CO2, CO, HC, CH4, NH4, BOD, COD, NMVOC, P, Cd, Cu, Pb, Fe, Zn, Particulate matters, etc.
Climate change Ozone depletion Human toxicity Photochemical oxidant formation Particulate matter formation
Ionising radiation Terrestrial acidification Freshwater eutrophication Marine eutrophication Terrestrial ecotoxicity Freshwater ecotoxicity
Human Health Ecosystem Resouces
EndPoint Substances
Level
weighting Monetary value
ReCiPe (H) Ecoinvent v2.2 Ecovalue Ecotax
CHAPTER 3 BRIDGE LCA MODEL AND THE DEVELOPED TOOL
14
Figure 3.3 User Input sheet in the GreenBridge LCA tool
Table 3.2 Structural components and processes covered by GreenBridge Life cycle phases Structure Structural components
Material manufacture phase
Substructures Piles, embankments, abutments, wing walls
Superstructures Slabs, beam, trusses, arches, cables, bracing, steel girders, bearings
Railway Track Railway tracks, sleepers, ballast, rail fasteners
Bridge equipment Formwork, dehumidification machines, railings, bearings, expansion joints, bitumen sealing, painting Earthwork Excavation, landfill
Construction phase
Machinery usage Electricity, diesel consumption
Transportations Truck/train/ship/passenger car transportation Traffic disturbances during the construction activities
Maintenance phase
Repair or replace all involved structural
components Edge beams, repainting, railway track, fasteners, sleepers, ballast, earing, railings, bitumen sealing, expansion joints Traffic disturbances Truck/passenger car disturbance
End of life
Demolition,
Transportation, steel recycling, concrete crushing
Truck/train/ship/passenger car transportation Electricity and diesel consumptions
The main features of GreenBridge are:
Covers 27 types of environmental indicators, including atmospheric emissions, mid-point indicators and the cumulative energy demand (CED). 8 types of emission substances are named as CO
2, SO
2, CH
4, CO, NO
x, NH
3, NMVOC, PM10; while 18 types of mid-point environmental impact categories are global warming (GWP), ozone depletion (ODP), human toxicity (HTP), photochemical oxidant formation (POFP), particulate matter formation (PMFP), ionizing radiation (IRP), terrestrial acidification (TAP), freshwater eutrophication
1) User input 2) Run the
GreenBridge tool 3) Life cycle impact
assessment 4) Results and Graphs
presented in word and excel LCI database
unit process and
emissions data
3.2 AN LCA BASED COMPUTATIONAL TOOL: GREENBRIDGE
15
(FEP), marine eutrophication (MEP), terrestrial ecotoxicity (TETP), freshwater ecotoxicity (FETP) and marine ecotoxicity (METP), etc.
Enables to convert the environmental impacts into monetary value by 2 approaches. The environmental impacts are further aggregated by the monetary weighting factors Ecovalue 08 (Ahlroth and Finnveden, 2011) and the updated value of Ecovalue 12 (Finnveden et al., 2013) as well as EcoTax 02 (Finnveden et al., 2006), with a focus on the Swedish condition.
Enables the LCA comparison for up to 10 different bridges simultaneously, and identification of the most dominated structural components and life cycle stage. The tool allows the user to replace the default LCI database with flexible user-defined detail level.
An excel file and word document are eventually created automatically to store all the numerical results and graphs.
Table 3.3. Material inventory data
Material type The inventory data types Specification
Concrete C50/60 C50/60 exacting concrete 0.4 water/cement (w/c) ratio 375 kg/m3 cement content
Reinforcement and Structural
Steel Steel, low-alloyed, at plant 63% primary with 37% secondary
steel from electric furnace route Painting Epoxy paint layer and anti-corrosion
zinc coating Degreasing, pickling, fluxing,
galvanising and finishing
Steel Railing electric, un- and low-alloyed steel Hot dip galvanized after fabrication Bitumen sealing Hot bitumen adhesive compound To protect roads and roofs against
water intrusion
Formwork by timber Scandinavian softwood 8 cm to 10 cm thick
Truck transportation 1 tkm transport, lorry 16-32 t, EURO3 Material transportation from factory to site
Rail transportation Freight train Material transportation from factory
to site
Ship transportation Freight ferry Material transportation from factory
to site
Passenger car transportation 1 personkm transport, passenger car Petrol driven car per person per km
Diesel Diesel burned in building machine Consumed in construction
machines
Electricity Electricity, low voltage, at grid Consumed in construction machines and heating system Edge beam replacement C50/60 exacting concrete and
reinforcement 0.4 water/cement (w/c) ratio 375
kg/m
3cement content
Steel railing replacement Steel, low-alloyed, at plant 63% primary with 37% secondary steel from electric furnace route Concrete crushing Diesel and electricity consumption 16.99 MJ diesel and 21.19 MJ
electricity per ton of concrete
3.2.1 How to use GreenBridge
Filling the sheet of ‘user-input’ is the only step required to run GreenBridge, which is the basis for the
further evaluation. The input sheet includes the quantity of materials, transportation and energy
consumption of each structural component considered through the life cycle stages. The level of analysis
detail is governed by the user, which can vary from filling only few main structural components to the
entire structural system with complete auxiliary components in detail. Table 3.2 presents the structural
components and scenarios covered in the tool, while Figure 3.4 illustrates the flow chart of the tool
execution.
CHAPTER 3 BRIDGE LCA MODEL AND THE DEVELOPED TOOL
16
