|
Impact categories, normalisation and weighting in LCA 1. Introduction
This guideline focuses at the optional steps in LCA, normalisation and weighting. Normalisation is a part of the EDIP methodology used to compare results (e.g. global warming) to well-known figures, and weighting aims at ranking different impact categories against each other in order to determine their importance according to political targets. The present guidelines present an update and extension of the original EDIP methodology in relation to normalisation and weighting. Further, the guideline is focused on choice of impact categories and choice of normalisation reference. It is also mentioned here that a parallel methodology has been developed in another subproject in the Danish LCA methodology and Consensus project. This methodology is termed EDIP 2000 and uses site-differentiated characterisation and normalisation. 1.1 Normalisation and weighting and standardisationIn the international standardisation organisation, ISO, a number of standards regarding LCA has been developed and approved. One of these standards, ISO 14042, Life Cycle Impact Assessment (LCIA), deals with the subject of normalisation (calculating the magnitude of the category indicator results) and weighting (assignment and calculation of different impact categories and resources reflecting the relative importance). This is based on value choices. According to ISO 14042, life cycle impact assessment includes several steps from the inventory to the interpretation:
In Life Cycle Impact Assessment, the first three steps; assignment of inventory results to impact categories, the characterisation, and the classification are mandatory, while normalisation, grouping and weighting are optional elements. 1.1.1 EDIP applied to ISO EDIP offers two ways of identification of the relative proportion of category indicators to reference values (normalisation in ISO terminology). One is to identify the relative proportion to current values. The other is to identify the relative proportion of category indicators to future values. In the original EDIP 1997 these ways are called normalisation and weighting, respectively. In order to avoid confusion, the terminology from EDIP 1997 is used in this update as well. However, it is recommended when presenting results from LCA work to use the ISO wording. 1.2 The concept of normalisation and weighting in EDIPBasically, the guideline takes off from the technical report (Stranddorf et al. 2005). This includes further development and extension of normalisation references compared to the original EDIP97 methodology as it is explained in Wenzel et al. (1997) and Hauschild and Wenzel (1998). In EDIP, along with the normalisation the concept of the person equivalent is developed. This is also used in the weighting process. The concepts of normalisation, person equivalent and weighting are explained below. In EDIP as well as in the ISO standards, the overall advantage of normalisation is an increased possibility of interpretation of indicator results compared to each other. Further, it also provides an option of control by unveiling extreme indicator results. Finally, normalisation is a preparatory element in the optional element of weighting. The goal of normalisation is to refer the impact scores to a common reference to enable comparison of different environmental impacts. It is an integral part of EDIP that the impacts are summarised for the area, which actually contributes to the current condition of the environment as it is experienced. For the global impacts it makes no difference where the environmental exchanges occur. But for regional and local impacts, only the exchanges occurring within the regions or local areas in question contribute to the current and future condition in the environment there. Due to this reasoning the normalisation and the subsequent weighting should ideally be carried out with normalisation references and weighting factors which, for each individual emission, are representative of the region where the emissions occur. However in the original EDIP 1997, the data for the global impacts are available while data for regional impacts, generally, are not available. Instead, EDIP use Danish normalisation references as default values for regional and local impacts. In contrast to EDIP 97 it has been chosen to regard human toxicity and ecotoxicity as being primarily regional impacts with the same normalisation reference being applied to all EU-15 countries. It is, however, also possible to use the updated Danish normalisation reference, e.g. in an assessment of specific Danish products or product systems, or as an element in the sensitivity analysis. The same possibility also exists for most of the other EU-15 countries. Due to lack of data no regional normalisation references for other regions e.g. Central Africa and the Middle East have been developed. For acidification, normalisation references are provided for EU-15 as well as for the individual countries within EU-15. The individual normalisation references can be found in Stranddorf et al. (2005). A different approach has been taken in a parallel project within the LCA methodology and consensus project, site-characterisation (Hauschild & Potting, 2005). The present guideline does not discuss the pros and cons of either method, but provides only an update of the original EDIP97 method with respect to normalisation and weighting. This project has provided a large number of normalisation references for EU-15 countries but substantial work still remains to be done before the ultimate goal, i.e. normalisation references for all relevant regions in the World, can be reached. 1.2.1 The person-equivalent The global impact will always be much greater than the impact from a particular region, irrespective of the type of effect to which the impact contributes. Use of global impacts as normalisation reference for the global impact categories, and regional impacts for the regional, and local impact categories will give an imbalance in the normalisation, and it will result in global impacts from the product system appearing as much less than the other impacts, because they are compared with the activity of the population of the entire world, while the others are compared only with the activity of, e.g. the European population. To correct this bias and ensure that the set of normalising references constitutes a common scale for all impact categories the normalisation references are calculated as the background impact over the course of one year per person in the area for which the impact is computed, irrespective of whether they are global or regional. This gives the normalisation references the unit "impact potential per person per year" for each individual impact category. Thus, when the impact potentials are normalised, they are expressed in person-equivalents, i.e. fractions of the contribution to the impact deriving from the average person in the affected area. 1.2.2 Establishing of normalisation references Normalisation references are established for Denmark, Europe (EU-15), and worldwide. For the relevant area surveys of emissions contributing to the different impact categories have been established. The normalisation references are calculated based on EDIP (Wenzel et al. 1997):
where: i.e. for each of the substances contributing to the impact category, the emitted quantity is multiplied with the equivalence factor for the specific substance giving the potential impact caused by the specific substance. The potential impacts by all the contributing substances are summarised and finally divided with the number of capita in the considered area. The figures used for normalisation in the EDIP methodology are the potential impacts caused per capita in a defined geographical area e.g. potential greenhouse effect per capita per year. 1994 has been chosen as reference year. The actual normalisation references are in general expressed as equivalents, e.g. kg CO2-eq./capita/year, with human toxicity and ecotoxicity being exceptions from this. When normalised the potential environmental impact is expressed in person equivalents (e.g. PEW94 or PEEU94) - i.e. relative to the annual contribution to the effect by one person in the affected area. The index to the unit expresses that the reference region for the normalisation is either the world or EU and the reference year is 1994. In principle, the normalisation methodology allows different impact categories to be compared to each other. The present study has developed normalisation references for Denmark, Europe (EU-15) and worldwide and provides thus the opportunity of a more versatile impact assessment. 1.2.3 Weighting The figures used for weighting in the EDIP methodology are based on the political reduction targets for the individual substances contributing to the relevant impact category. 2004 has been chosen as the target year. The actual weighting factors are dimensionless. The weighting factors are linked to the above mentioned normalisation references with regard to the geographical area covered. The present study presents weighting factors for Denmark, Europe (EU-15) and worldwide. The procedure for applying and using weighting in this project consists of the following steps:
In this project 1994 has been chosen as reference year and 2004 as target year. The weighting factor (WF) for impact category x on a given geographic scale is thus defined as follows:
As the actual political targets for reducing certain pollutants in the different countries seldom are related to the target year selected for this project, it has been necessary to interpolate and extrapolate the data, so that the weighting factors for all effect categories relate to year 2004. The interpolation is done linearly. The unit of weighted impact potentials is the targeted person equivalent – the PETWEU2004 – expressing the environmental impact from an average citizen according to the political targets for each of the impact categories in the year 2004. The index expresses that the reference region for the weighting and the underlying normalisation is either the world (W) or EU (EU) and the target year used for the weighting is 2004. In line with the thinking applied in the life cycle impact assessment standard ISO 14042, the weighting of EDIP97 using political targets to derive weighting factors can be seen as a normalisation where the politically targeted level of impact is used as normalisation reference, representing the probable future environmental impact – a prediction of what the background impact will be in 2004. In ISO TR14047, which illustrates the use of ISO 14042, an example presents this interpretation which means that the use of the EDIP default weighting factors is acceptable according to ISO 14042, also in comparative assertions disclosed to the public, where weighting is not allowed. 1.2.4 Reference year In the original EDIP97, 1990 was used as reference year for normalisation, while 1990 and 2000 were chosen as the reference and target year for weighting. In this update, 1994 and 2004 have been chosen as new reference year and target year for normalisation references and weighting factors, respectively. The reason for choosing 1994 as reference year has been the relatively easy availability of compilations of data for most of the countries included in the EU-15 normalisation reference. In the recent years the availability of emission data has increased significantly e.g. by access to databases via Internet, and the delay caused by central compilation of data can be expected to be reduced in the future. Therefore, future updating can be expected to be faster with reference years closer to the actual year. 1.3 Purpose of the guidelineThe purpose of the guideline is to provide guidance on choice of impact categories in LCA. Based on a comprehensive list in the technical report, the guideline lists the most commonly used impact categories. In addition to the list, general considerations regarding quantitative and qualitative assessment of impact categories are presented. Further the purpose of the guideline is to
The principle of normalisation is - as mentioned above - that global impacts are normalised globally, the regional impacts and the local impact are normalised regionally. The new normalisation reference includes normalisation reference at both the regional and local levels for the relevant impact categories. Therefore, the guideline also provides guidance on when to apply the different normalisation references. 1.4 How to read the guidelineChapter 2 is focused on a comprehensive list of impact categories to be included in LCA. In the chapter the list is presented and some considerations regarding which impact categories to choose or exclude are presented as well. Chapter 3 presents the updated and extended normalisation references and it deals with the application of these normalisation references. The principles in weighting and a list of updated weighting factors are also presented. Chapter 4 to 10 deals with the impact categories for which updated normalisation references are developed. All chapters are structured in an equal manner: Each chapter begins with a short presentation of the environmental issue and what potential impacts the issue has. The contributions of different substances to the impact category are listed in a table and it is followed by a description of how the normalisation reference for the specific impact category is modelled. Correspondingly, the procedure of the weighting factor is also included. In the chapter regarding global warming (Chapter 4), a detailed description of the steps in the calculation is demonstrated. Recommendations for the use of normalisation references are presented and in order to illustrate the applicability of the new normalisation references, a common imaginary example is used in the guideline. The full example is shown in Chapter 11. In each chapter regarding the impact categories the relevant part of the example is brought. For each impact category, the chapter is finalised with a section with further literature. 1.5 ReferencesHauschild, M. & Potting, J. 2005. Background for spatial differentiation in life cycle impact assessment - the EDIP2003 methodology. Environmental Project no. 996, Danish EPA, 2005 Hauschild, M. & Wenzel, H. 1998, Environmental Assessment of Products. Volume 2 - Scientific background. First edition. Chapman & Hall, London. ISO TR 14047. Illustrative examples on how to apply ISO 14042 – Life cycle assessment – Life cycle impact assessment. International Organisation of Standardisation, Paris. Stranddorf, H.K., Hoffmann, L. & Schmidt, A. 2005. Update on impact categories, normalisation and weighting in LCA. Environmental Project no. 995, Danish EPA, 2005 Wenzel, H., Hauschild, M. & Alting, L. 1997. Environmental Assessment of Products.Volume 1 - Methodology,Tools and Case Studies in Product Development. First edition. Chapman & Hall, London.
|
