Tools & Data I: Tools and Data for LCM I
Process on “Global Guidance for LCA Databases” – Just in time
1UNEP DTIE, France; 2SETAC North America, United States of America; 3World Steel Association, Belgium; 4US Environmental Protection Agency, United States of America; 5Technical University Berlin, Germany; 6ESU-Services, Switzerland; 7Nat. Inst. of Adv. Indust. Science a. Technol., Japan; 8Plastics Europe, Belgium; 9Federal Center of Technologic Education, Brazil; 10Sichuan University, China; 11Joint Research Centre, Belgium; 12individual, United States of America
UNEP and the Society of Environmental Toxicology and Chemistry (SETAC) are collaborating within the Life Cycle Initiative to address the need for global guidance on life cycle inventory (LCI) data collection and processing into databases for widespread use. The process was launched at the first Stakeholder Engagement Meeting, “Towards Global Guidance for LCA Databases”, in Boston on the 30th of September in 2009, where the high attendance confirmed the international interest in the UNEP/SETAC proposal and a majority of the participants agreed with the vision.
The vision is to help provide global guidance on the establishment and maintenance of life cycle assessment (LCA) databases, as an input for improved interlinkages of databases worldwide. The vision is expected to contribute to increasing the credibility of existing LCA data, to further foster the generation of more data (also for applications such as carbon and water footprint) and to enhance their overall accessibility.
It is therefore expected to help provide a sound scientific basis for product stewardship in business and industry and life cycle based policies in governments, and ultimately, aid in the advancement of the sustainability of products. This work will complement that of other initiatives.
The importance of the project is highlighted by the fact that a global network of data is required for managing supply and production chains in a global economy. To develop credible LCAs across such a scale, it is essential that databases have uniform data requirements to allow consistent modeling and reliable decision support. Additional disclosure of the information on a single operation or gate to gate level is seen as beneficial.
The seven stakeholder meetings following the launch in 2009 have informed stakeholders about this plan for the global guidance development. The central upcoming activity is the 5-day Pellston-type Workshop in 2011, which is being organized by the Secretariat of the Life Cycle Initiative on behalf of UNEP and SETAC. A Steering Committee (SC) of the process has been established and is co-authoring this paper.
What the open data movement means for the life cycle management community
Delft University of Technology, Netherlands
Life Cycle Management can be challenging due to the need to incorporate large amounts of information of a diverse nature, which can require a significant effort to compile. Because of this, some have argued that the act of performing an LCA is not necessarily eco-efficient itself, since the cost of the effort required may exceed the environmental benefits realized. Since this argument was first brought forward, we have seen the emergence of the World Wide Web dramatically change the economics of information gathering via lowering collection costs, and allowing for large-scale peer review and availability at an unprecedented scale. Further relevant developments are emerging with the growing push towards open data, as evidenced by high-profile initiatives such as data.gov in the US, data.gov.uk in the UK and the World Bank's data.worldbank.org.
We investigate these trends, and explore how the same technologies and philosophies behind these projects can also be applied to the benefit of Life Cycle Management practitioners. For example, this has the potential to address data quality issues in several ways. First, it would greatly speed the process of comparing different data sets describing the same objects. This could more easily highlight variances based on location and technology configurations, and may show flows or aspects that other researchers have missed. At a higher level, the global push for open data means that we are starting to have unprecedented access to a growing amount of information describing the state of the world. This can better inform the descriptions of stages in a life cycle, and can be used to provide a more informative context for issues such as resource scarcity, working conditions, and localized deforestation. At the same time, this highlights the need for researchers skilled in the multitude of means useful for managing and finding insights in large data sets.
Open data is not just about using certain Web technologies, but it has social dimensions that must be addressed as well. It will require actions such as rethinking of data licenses in light of the potential benefits, and it will likely lead towards a shift in business models from a focus on collection and maintenance of data to that of interpretation. Given this, we discuss some of the barriers to implementation, highlight promising initiatives already starting to emerge, and finally propose the next steps that are needed.
Comparison of inventory database -Effects of inventory data differences on the results of life cycle assessment
National Institute of Advanced Industrial Science and Technology (AIST), Japan
The inventory database is indispensable for implementing an LCA efficiently. The present time, there is some database for LCA in the world. Existing inventory databases of each country and each provider have each feature. The conceivable features of each database were including system boundary, and environmental load items. The features influence inventory analysis results greatly. In addition, LCA practitioners have to use their own country data, however if there is not a request data, they modify data from other countries. The assumptions and approaches used by different practitioners can lead to variable results. It is necessary to clear the availability of each database for the inventory analysis.
In this study, I compared inventory data of each database focused on system boundary and allocation methodology. And each database features were cleared.
Integrating environmental decision making into the product innovation process
Unilever, United Kingdom
In November 2010, Unilever launched the Unilever Sustainable Living Plan (http://www.sustainable-living.unilever.com/). As part of this plan, there is a key target to halve the environmental footprint of our products across their lifecycle. In order to achieve this target, it is vital that Unilever is able to build a sustainability culture which is embedded into the product innovation process, to drive new product concepts that better meet consumer needs while also reducing impacts on the environment. This has been achieved by applying expert knowledge and understanding of the product portfolio by LCA specialists and data gained from calculating Unilever’s footprint for greenhouse gases (GHG), waste and water, in order to develop environmental assessment tools suitable for different stages of Unilever’s innovation process.
Consequently, the environmental assessment tools have been fully integrated into the Unilever’s innovation project management system to ensure that environmental impact assessment is now mandatory. This process has been supported with training developed for Project Leaders and the creation of environmental super-users, to provide tailored support. The environmental assessment tools are important to inform project teams about the potential impacts of their innovation across waste, water and GHG and to inform gate-keeping decisions. Within the first six weeks of launching the tools, over 150 innovation projects had assessed the impacts of their new product innovation. For the early stages of innovation, a qualitative assessment is completed, and then a more rigorous quantitative assessment is completed as the innovation reaches market launch.
This paper will explain the approach taken to develop environmental assessment tools suitable for guiding decision making as part of Unilever’s innovation process. In particular, this paper will discuss the challenges of building a sustainability culture within a large multinational company with a diverse portfolio of food, homecare and personal care products marketed in over 170 countries with varying consumer behaviours. For example the successful integration within Unilever required the utilisation of business champions assigned to product categories, who could ensure the correct application of simplified tools, supported by LCA specialists who are able to complete bespoke environmental assessments for more radical innovations. The lessons learned from this process will the shared and case studies will also be provided to give practical examples of how the tools have been applied to guide decision making and develop new product designs that will have less impact on the environment.
Modular extrapolation approach for crop LCA MEXALCA: Global warming potential of different crops and its relationship to the yield
1Agroscope ART, Switzerland; 2Unilever, United Kingdom; 3Johann Heinrich von Thünen Institut, Germany
Businesses wishing to analyse the environmental impacts of their products using Life Cycle Assessment (LCA) methods increasingly require large amounts of very detailed data which are rarely readily available. This is particularly true for companies operating global and rapidly changing supply chains with a large range of products and ingredients originating from all over the world. Thus, there is an urgent need for data that are sufficiently reliable but can be supplied without extensive, time consuming and costly collection.
The aim of the method MEXALCA (Modular EXtrapolation of Agricultural LCA) presented here is the geographical extrapolation of life cycle inventories and impacts in order to enable a simplified assessment of the environmental impacts and estimation of their variability worldwide. Detailed data are required for at least one typical production system that is then used as the baseline for the extrapolation to all other countries in the world producing the same crop. This base country Life Cycle Inventory is split into nine modules corresponding to the main farming operations and inputs known to dominate the environmental impacts of crop cultivation: essential cropping activities (including the minimum operations and inputs to grow and harvest the crop); variable machinery use; tillage; nitrogen, potassium and phosphorus fertilisation; plant protection; irrigation; and product drying. Mathematical functions relating global statistics of crop yields and agricultural production intensity (mainly FAOSTAT) in the baseline and target countries, respectively, are applied for the geographical extrapolation of each modular farming input. The full methodology and first validation using data from the ecoinvent database are described in Roches et al. (2010).
We applied MEXALCA to a variety of crops. The variability of the global distribution of the impacts depends on several factors: the amount of inputs in the original inventory together with the yield and intensity of agricultural production in the original country, the distribution of the yields of the considered crop, the distribution of the intensity indices for the producing countries, the contributions of the different producing countries to the global production, and the functional unit (ha per year or kg). The contribution of these influencing factors to the variability of environmental impacts as well as the differences observed between crops will be discussed. The method allows to estimate the global variability of environmental impacts of crops and can be a first approach to fill data gaps in situations, where detailed and specific data are missing.
The use of models to account for the variability of agricultural data in LCA studies
1CEMAGREF-ITAP, France; 2Solagro,France; 3CIRAD-Hortsys, France; 4INRA-LBE, France
LCA outputs are often presented as point estimates of potential impacts. Average impacts values may be misleading to rank different options with LCA, especially in the case of agricultural products owing to the inherent variability of farming systems and field emissions. Yet, accounting for emissions variability requires information on their distribution, which can be difficult to derive from small samples of field measurements, but that models can simulate. In this study, we compared two approaches to account for direct NH3 and N2O emissions variability for a LCA of slurry application techniques; one was based on experimental data from a literature review and the second on simulated data with mechanistic simulation models.
Four slurry application techniques were compared: broad spreading, band spreading, harrowing after surface application and direct injection. NH3 and N2O emissions at field level for each technique were expressed relatively to the emissions observed with band spreading in the same conditions. Field emissions from the literature review were measured in contrasted conditions regarding the soil moisture, air temperature, slurry pH, or cultivated crops. The simulated emissions were calculated for 5 sites in France for 7 years.
In the literature approach, little data was available and it was not possible to estimate their representativeness. As a consequence, a conservative approach consisted in assuming a uniform distribution between the maximum and the minimum data values and propagating the total interval on LCA impacts results. However, owing to possibly large uncertainty margins, this approach can hamper the power of the comparison between the studied alternatives. In the modelling approach, more specific distributions were obtained by computing a variety of input parameters, offering more chance to discriminate the studied alternatives. Since we selected each set of input parameters, it was easier to assess the representativeness of the simulation outputs. Interestingly, the range of emission factors obtained either with literature data or simulated data were comparable. Moreover, distributions obtained by simulation were multi-modal, indicating revealing distinct emissions values for different types of situations. In that case, factorial analyses and/or analyses of variance (ANOVA) allow identifying the most sensitive input parameters (such as machine, soil and climate conditions…) on which to base a typology of farming practices and situations.
Development of an LCA tool for the evaluation of environmental performances and eco-design of drinking water treatment plants
1Université de Toulouse, France; 2Centre International de Recherche Sur l’Eau et l’Environnement, France; 3Public Research Centre Henri Tudor/CRTE, Luxembourg
Nowadays, industry leaders and politicians care about environmental issues more than ever. Nevertheless, adapted tools for tackling them are necessary and still missing for some industrial sectors like water treatment, while being confronted with high process variability. Besides, decision-makers need to believe first and foremost in the technical feasibility before they get interested in its LCA study results.
An LCA tool is currently developed precisely for this reason. The objective is to provide a user friendly tool, able to assess the environmental and sanitary performances of drinking water treatment plants. Although existing plants could be studied, the major interest relies on the predictive approach for an eco-design perspective.
The working environment is based on the Umberto® software. Unit processes are modelled as input-output models and coded into Python™ scripts. These models, highly parameterized, are saved into a library which constitutes the EVALEAU tool. In order to carry out an LCA, background processes and evaluation methods can be taken from the Ecoinvent database.
Several models have already been developed for each water treatment step. After reviewing the existing models, the most relevant have been programmed into the corresponding scripts. Modelling is flexible in order to fit any engineering design related to project constraints, i.e. our tool is able to calculate each particular project case. It exhaustively designs the processes from raw water quality and a few user-defined parameters. The calculation can also be run with real LCI data coming from an existing plant. Finally, this library makes it possible to simulate a process chain within any specific project context.
Moreover, Python™ scripting can be linked to other software tools. For instance, water quality changes along the process chain are calculated using PHREEQC® (hydrogeochemical software). So, the water composition after some chemical addition is accurately estimated. All the unit process models are also linked to a spreadsheet in order to save technical results such as the size of facilities, water composition or pathogen elimination rate. Scripting also enables one to program mathematical algorithms for parameters sensitivity analysis. Process parameter values are changed according to some predefined mathematical method (e.g. Morris method). Technical parameters which have an influence on the environmental impact simulation results are detected and tagged as priority action levers for the project under consideration.
Such detailed modelling can make the connection between technical changes at the process scale, resource variability and improvements of environmental impacts at the project level.
The eco-efficiency development of economic sectors in Europe
1Karlsruhe Institute of Technology, Germany; 2TAURUS Eco Consulting, Germany
The development of environmental use from economic activities is widely discussed among stakeholders. Eco-efficiency aims at linking economic efficiency with environmental efficiency. The main impetus of the concept is to identify and implement activities to enable simultaneously economic efficient and cleaner production. An important need is to describe concurrently the environmental and economic performance of industries. For a stakeholders’ decision the analysis of the driving forces of eco-efficiency within an industrial class also compared to competing industrial classes as a decision base is necessary. Although the aim of sustainability has to be reached also on a macroeconomic scale, to establish a decision-making a detailed sophisticated information base is necessary, i.e. the indicator should describe the activities of industries and should also include the whole economy comprehensively. To describe years of development and compare different sectors in different countries a regularly updated and harmonized data base is necessary.
This presentation discusses a specific eco-efficiency indicator on the level of disaggregated industrial sectors. The presented indicator – environmental impact, in E99-points, per economic performance, in Euro, of an industrial class – make use of the emission data base European Pollutant Emission Register (EPER) respectively the Pollutant Release and Transfer Register (PRTR) in combination with the life cycle impact assessment method Eco-Indicator 99. The approach enables to compare an industrial class in a country of the European Union with competing industrial classes on the home market but also in other member states for different impact categories. Due to still existing data constraints the contribution to the conference describes the development of sectors between 2001 und 2004 and discusses the results.