Life Cycle Impact Assessment

This chapter serves as an introduction to the presentation of the many aspects of life cycle impact assessment (LCIA) in this volume of the book series ‘LCA Compendium’. It starts with a brief historical overview of the development of life cycle impact assessment driven by numerous national LCIA methodology projects and presents the international scientific discussions and methodological consensus attempts in consecutive working groups under the auspices of the Society of Environmental Toxicology and Chemistry (SETAC) as well as the UNEP/SETAC Life Cycle Initiative, and the (almost) parallel standardisation activities under the International Organisation for Standardisation (ISO). A brief introduction is given on the purpose and structure of LCIA. As a common background for the 11 chapters dealing with the characterisation modelling of the most common impact categories, the chapter concludes with an introduction of the general principles and features of characterisation.

This chapter concerns ‘selection of impact categories’ and ‘assignment of LCI results to selected impact categories (classification)’. These elements are the first two mandatory elements of Life Cycle Impact Assessment (LCIA). They have largely been developed during the 1990s. In practice these mandatory steps are often performed using default lists of impact categories and default lists of inventory items classified to these default impact categories as part of LCA handbooks, guides and software tools. Despite these default lists, it is still important to pay sufficient attention to both these steps in any LCA case study. Every practitioner of LCA will always need to justify the completeness of default lists of impact categories and default classification lists for their study. In addition, the handling of missing information needs to be reported explicitly and transparently, and needs to be taken into consideration when developing conclusions and recommendations for the study at stake. After the 1990s, the attention to selection of impact categories and classification in LCA methodology studies and papers has been limited. Still, there are issues that deserve further attention from LCA method developers such as the harmonisation of naming impact categories while distinguishing or not between names for midpoint impact categories and names for endpoint category indicators, keeping default lists of impact categories manageable, and the classification of inventory results that relate to more than one impact category.

Climate change is defined as the warming of the climate system due to human activities. Emission of greenhouse gases (GHGs), which cause an increase in radiative forcing, is the main contributor, and the only climate forcing agent currently considered in life cycle impact assessment (LCIA) methodologies. The direct consequence is an increase in the temperature of atmosphere and oceans, which leads to several types of higher-level impacts such as sea level rise, extreme meteorological events and perturbations in rainfalls, which in turn cause damages to human health and ecosystem quality. All the LCIA methodologies use GWPs (Global Warming Potentials), developed by the Intergovernmental Panel on Climate Change (IPCC), as midpoint characterisation factors since they are based on state-of-the art and peer-reviewed publications and have a relatively low associated uncertainty. Some LCIA methodologies also propose endpoint characterisation factors. However, these factors are considered highly uncertain because of the complexity of the impact pathway so that further research is still needed to improve robustness of the models. Recent new developments are addressing the accounting of biogenic CO₂ emissions, the timing of GHG emissions, and the development of characterisation factors for terrestrial albedo changes induced by human activities.

The stratospheric ozone layer plays a critical role in regulating conditions on Earth, but has been substantially depleted by CFC (chlorofluorocarbon) and other halocarbon emissions. This has increased transmission of UVB radiation to the surface, and been implicated in a range of negative human and ecosystem health impacts.

Midpoint-level LCA has traditionally utilised the steady-state Ozone Depletion Potential factors that are prominent in policy making. Current ozone-depletion endpoint models incorporate skin cancer, cataract damages, and certain changes in ecosystem productivity caused by excess UVB exposure. Other health, ecosystem and agri-production impacts are still to be incorporated into the LCA framework.

As the ozone layer recovers following regulated halocarbon emission reductions, scientific attention turns to the question of longer term ozone layer management. While growing anthropogenic emissions of N₂O (nitrous oxide) might pose a threat to ozone layer recovery, the mitigating effects of CH₄ (methane) and CO₂ (carbon dioxide) emissions will more than compensate for this. Global stratospheric ozone is expected to exceed pre-industrial levels sometime this century, albeit with a very different spatial distribution. Predictions are that UVB levels will remain elevated in the tropics, but become depressed in other regions. That latter situation might increase the incidence of diseases associated with insufficient UVB exposure.

Whatever the policy response to these new challenges, it seems the interface of ozone layer science and management will become increasingly complex. It may be that the metrics used for ozone layer analysis will also need to evolve, if LCA is to remain relevant to this new management paradigm.

This chapter reviews the human toxicological impacts of chemicals and how to assess these impacts in life cycle impact assessment (LCIA), in order to identify key processes and pollutants. The complete cause-effect pathway – from emissions of toxic substances up to damages on human health – demonstrates the importance to account for both outdoor and indoor exposure, including consumer products. Analysing the variations in intake fraction (the fraction of the emitted or applied chemical that is taken in by the consumer and the general population), effect factor and characterisation factor across all chemicals and impact pathways characterizes the contribution of each factor to the total variation of 10–12 orders of magnitude in impacts per kg across all chemicals. This large variation between characterisation factors for different chemicals as well as the 3 orders of magnitude uncertainty on characterisation factors means that results should by default be reported and interpreted in log scales when comparing scenarios or substance contribution! We conclude by outlining future trends in human toxicity modelling for LCIA, with promising developments for (a) better estimates of degradation half-lives, (b) the inclusion of ionization of chemicals in human exposure including bioaccumulation, (c) metal speciation, (d) spatialised models to differentiate the variability associated with spatialisation from the uncertainty, and (e) the assessment of chemical exposure via consumer products and occupational settings. As a whole, the assessment of toxicity in LCA has progressed on a very sharp learning curve during the past 20 years. This rapid progression is expected to continue in the coming years, focusing more on direct exposure of workers to chemicals during manufacturing and of consumers during product use.

The first section of this chapter outlines the complete cause-effect pathway, from emissions of toxic substances to intake by the population up to damages in terms of human health effects. Section 2 outlines the framework for assessing human toxicity in LCIA. Section 3 discusses the contributing substances and their coverage in LCIA methods. Section 4 provides an overview of the main LCIA methods available to address human toxicological impacts. Section 5 presents the range of variation of factor across chemicals, the main sources of uncertainty and good interpretation practice of results from human toxicity assessments. Section 6 finally discusses new developments and research needs.

This chapter deals with the causes and consequences of exposure from emissions of primary particles and secondary particle precursors on human health and how to deal with them in life cycle impact assessment (LCIA).

Following a short introduction and literature review, the first part outlines the complete emission-to-damage pathway, from emissions of primary particles and secondary particle precursors to damage on human health, so called ‘respiratory effects from particles’. It describes the assessment framework for quantifying respiratory effects from particles in the context of LCIA. The second part provides an overview of methods that have been available in LCA to address impact of particles on human health. We finally discuss variability and main sources of uncertainties, as well as future trends in modelling respiratory effects of particles in LCIA.

Anthropogenic ozone arises as the product of reactions in the atmosphere between OH-radicals, the anthropogenic air pollutants nitrogen oxides (NO_x) and different non-methane volatile organic compounds (NMVOC).

The photochemical oxidant of main interest within LCA is ground level ozone (O₃) caused by the emission of the air pollutants NO_x and NMVOC within life cycles of products and services. Several different LCIA methods have been developed in the last 20 years to characterise this impact category. Some provide midpoint and some endpoint characterisation factors, and there are site generic and spatially explicit methods. They all struggle with the highly non-linear dependence of ozone creation on background conditions regarding chemical substances and meteorology and also the fact that many response functions include thresholds and that resulting impacts depend on the ozone exposure, both for impacts on human health and on other living beings and even on materials. Hence, the modelled impacts due to ozone caused by anthropogenic emissions are subject to large variability and uncertainty.

Ecotoxicity impact assessment of chemicals in life cycle assessment (LCA) adheres to a number of underlying principles and boundary conditions: (1) a large number of emitted substances to cover (at least 100,000 potentially relevant elementary flows with current models covering around 2,500), (2) linearity of characterisation models, (3) conservation of mass and mass balance, (4) infinite time horizon, (5) additivity of toxicity, (6) assuming average conditions as best estimates to avoid bias in the comparison (including consideration of generic/average ecosystems and impacts). The cause-effect mechanism for ecotoxicity impacts of chemicals can be divided into four parts: (1) chemical fate (i.e. chemical behaviour/distribution in the environment), (2) exposure (i.e. bioavailability), (3) effects (i.e. affected species), and (4) severity (i.e. disappeared species). In terms of species represented, a freshwater ecosystem is described in this chapter by three trophic levels: (1) primary producers (e.g. algae), (2) primary consumers (i.e. invertebrates), and (3) secondary consumers (e.g. fish). Model uncertainty was estimated at about three orders of magnitude on top of important sources of parameter uncertainty such as degradation rates and effect factors. Current midpoint LCIA methodologies covering ecotoxicity include TRACI 2.0, and the ILCD recommended methodology, both employing the USEtox factors. Current LCIA methodologies covering midpoint and endpoint characterisation are ReCiPe, LIME, IMPACT 2002+, and IMPACT World+. Important research needs are (1) increasing substance coverage, (2) further developing marine and terrestrial ecotoxicity modelling for midpoint, (3) improving endpoint modelling for ecotoxicity towards biodiversity, (4) consideration of long-term emissions and impacts of metals, (5) importance of spatial and temporal variability, (6) mixture toxicity, and (7) decreasing model and parameter uncertainty.

This chapter outlines the cause-effect pathway of terrestrial and aquatic acidification from air emissions to ecosystem damage. Carbon dioxide is the main cause of (coastal) marine acidification, while nitrogen and sulfur inputs are underlying the damage due to freshwater and terrestrial acidification. Various life cycle impact assessment (LCIA) methods address parts of the impact pathway. Terrestrial acidification, caused by base cation leaching, has been addressed by a number of midpoint methods and several methods determining impacts to biodiversity. To decrease uncertainty in the ecological effect predictions, more insight needs to be gained in the stressor-response curves for many regions of the world. Moreover, research is needed regarding other indicators related to biodiversity than relative species richness as such. For freshwater acidification, only one midpoint and one endpoint method are available, with substantial options for improvement. To address ocean acidification in LCIA in the future, a carbon cycle model needs to be used to make the link to ocean acidification and stressor-response curves that assess impacts on marine biodiversity.

Anthropogenic increases in nitrogen and phosphorus inputs to terrestrial and aquatic ecosystems have driven increases in eutrophication, the occurrence of ecosystem changes due to over-supply of nutrients. Eutrophic water bodies exhibit changes in species composition that often include algal blooms and oxygen depletion, with occasionally arresting images of fish kills or dead zones. Though dramatic and subtle consequences of eutrophication itself have been described for over 100 years, understanding of nutrients as the main drivers for this phenomenon is more recent. Modelling nutrient fate has reached a basic level of operability, with a general rule that freshwaters are limited in phosphorus (and hence respond to its addition), and terrestrial and marine systems are nitrogen limited. However, understanding of ecosystems responses such as species shifts or changes in primary productivity is still growing. Future work should incorporate more comprehensive metrics to quantify impacts of eutrophication on ecosystems – and the human systems that depend on them.

Land use impacts are the effects caused by the use of land by humans, which range from changes in species composition and abundance to the disruption of ecosystem processes contributing to climate and water regulation. These impacts end up affecting key areas of protection such as human health, ecosystem quality, and natural resources. In life cycle assessment (LCA), land use impact assessment quantifies the difference between the land quality level of the studied system and a reference level over the duration and the area being used. For land transformations (change of land use), the time required for land to naturally regenerate is often considered as the duration when calculating impacts. Indicators to assess land use impacts on biodiversity in LCA focus on either species richness or on ecosystem metrics; whereas the indicators for land use impacts on ecosystem services are classed as pressure (describing land degradation processes) and state (describing overall quality) indicators. Many approaches to describe such impact indicators are presented in this chapter. The key areas for further research relate to (1) the description of the reference used to quantify the impacts; (2) aggregation of impacts on different ecosystem services; (3) the availability of inventory information to be able to inform the spatial differentiation level required for a more accurate impact assessment; (4) leveraging ‘big data’ processes to allow full utilisation of the data available on ecosystem quality; and (5) promote global consensus on impact indicators in order to facilitate comparison and stability of LCA results.

Water use impacts have two different dimensions: pollution (degradative use) and consumption (consumptive use). Degradative use is mainly tackled by impact assessment of pollutant emissions. This chapter focuses on water deprivation due to consumption. The impacts considered here address the case of ecosystems and human users being deprived of water, but also the depletion of stock resources, potentially depriving future users of water.

The short-term water cycle is dominated by evaporation from sea, precipitation on land and runoff in rivers. Groundwater and lakes play a longer-term role but are crucial for the assessment of effects of water use competition. While the global water cycle is not heavily influenced by human activities, the impacts can be significant in specific regions, and therefore regional water cycles are relevant. The temporal variability of the hydrological processes is also important for considerations of water use effects on environment.

The existing methods consider the main features of the hydrological cycle but still many improvements in the global, regionalised data are required for proper integration of relevant aspects in life cycle impact assessment (LCIA) of water use. Moreover, the available life cycle inventories have generally low data quality for the relevant flows for most processes and often lack global coverage.

The impact assessment methods can be grouped into four main categories: water scarcity as midpoint, impacts on human health, impacts on ecosystem quality, and resource depletion. They cover a variety of impact pathways, while still many important issues such as in-stream use of dams are currently missing.

Abiotic resource use in life cycle assessment (LCA) deals with the environmental concerns due to the use of resources such as metals, minerals, fossil energy, nuclear energy, atmospheric resources (e.g. argon), and flow energy resources (e.g. wind energy). Land and water may also be considered as abiotic resources, but these are dealt with elsewhere in the book series in dedicated chapters (Chap. 11 Land use by Llorenç Milà i Canals and Laura de Baan and Chap. 12 Water use by Stephan Pfister). Methods that evaluate ‘abiotic resource use’ in LCA were divided in three categories: (1) Resource accounting methods, which are methods that account for the overall natural resource use along the life cycle of a product; (2) Resource depletion methods at the midpoint level, which are methods that address the scarcity of resources (and therefore damage to the area of protection Resources), but at a midpoint level; and (3) Resource depletion methods at the endpoint level, which are methods that address the scarcity of resources at an endpoint level. Numerous methods are presented in this chapter, with different concepts and approaches. However, several gaps still exist in the evaluation of abiotic resource use in LCA, and more research is needed.

Defined as an optional step in the ISO14044 requirements, normalisation in LCA relates the characterised impact indicator scores of an analysed system to those of a reference system. By putting the LCA results in a broader perspective, it can facilitate their interpretation and communication, and allow checking whether their magnitude looks reasonable. This chapter provides a comprehensive overview of the two major normalisation approaches, internal and external normalisation, encompassing for the latter both production-based and consumption-based inventory methods. Pros and cons are addressed for each approach. Because of its wide use and usefulness, emphasis is put on external normalisation. The chapter details the calculation of external normalisation references, including their scoping, the collection of data for building their associated production-based or consumption-based normalisation inventories, their computation by use of adequate sets of characterisation factors, and their resulting uncertainties. The chapter provides insights in the application of normalisation in practice. After listing the past efforts of establishing external normalisation references for different regions in the world, the use of different normalisation approaches and their possible interpretation in relation to the goals and scope of an LCA study are discussed, along with the potential uncertainties and biases in the normalised scores.

In the ISO 14044 standard 2006, weighting is an optional step in life cycle impact assessment (LCIA). It enables the user to integrate various environmental impacts in order to facilitate the interpretation of the life cycle assessment (LCA) results. Many different weighting methodologies have been proposed and several are currently being used regularly. Most existing studies apply the average of the responses obtained from the people (i.e. the decision makers) that were sampled. Others believe that weighting factors should be based on the preferences of society as a whole so that LCA practitioners can successfully apply them to products and services everywhere. This chapter classifies methods of weighting into three categories: proxy, midpoint, and endpoint methods. Results using proxy methods, such as MIPS (Material Input Per Service), CED (Cumulative Energy Demand), TMR (Total Material Requirement), Ecological Footprint, and CExD (Cumulative Exergy Demand), are fairly easy to understand because physical quantities such as weight and energy are used. The advantages of midpoint methods include compliance with the ISO framework and how it permits weighting that uses characterisation results. Endpoint methods allocate weights to Areas of Protection (AoP) rather than at midpoints, reducing the number of subject items and simplifying interpretation. Recently, weighting with endpoint methods has attracted attention due to the advancement of characterisation methodologies of this type. This chapter presents the different features of weighting and integration approaches applied in LCIA. The important differences and future problems concerning five key endpoint weighting methods are described. It concludes with a brief summary of the key features of the weighting methods introduced herein.