1 Life Cycle Impact Assessment (LCIA) Methods

1

Electronic Supplementary Material

1 Life cycle impact assessment (LCIA) methods

There are two categories of LCIA methods: midpoint-based and damage-oriented impact assessment methods (Jolliet et al. 2003, 2004). In the midpoint analysis, the environmental impacts are considered in relatively early stages of the cause-effect chain. The midpoint-oriented methods are associated with less uncertainty compared to the damage-oriented methods. The damage-oriented methods assess the environmental impacts at the endpoint in the cause-effect chain and these methods are associated with high uncertainties (Mizsey et al. 2009). The four midpoint-oriented LCIA methods are summarized as follows.

EDIP97/EDIP2003

The Danish method for Environmental Design of Industrial Products (EDIP) was developed in 1996 (Wenzel and Alting 1999). EDIP97 is a midpoint approach covering most of the emission-related impacts, resource use and environmental impacts. The basis for normalization is person equivalents; weighting is based on political reduction targets for environmental impacts and supply horizon for resources. Compared to the other methods, EDIP97 emphasizes the category of toxicity but gives little consideration to resource depletion.

TRACI

The tool for the reduction and assessment of chemical and other environmental impacts (TRACI) is developed by the U.S. Environmental Protection Agency (Bare et al. 2003). This method adopts 12 impact categories at the midpoint level. This approach has the benefits of (i) reducing uncertainties of prediction at the damage points, and (ii) decreasing the modeling complexity (Bare et al. 2003). The TRACI methodology reflects state-of-the-art developments and best practice of life-cycle impact assessment methods in the United States.

IMPACT2002+

IMPACT2002+ is one of the European LCIA methods, which represents a combination of midpoint and damage oriented approaches because it defines all types of life cycle inventory results in 14 midpoint categories and 4 damage categories (Jolliet et al. 2003). The midpoint categories are grouped and linked to the damage categories. In this way, users can better understand the cause-effect chain of the environmental impacts. IMPACT2002+ also improves upon IMPACT2002 by adding the considerations of human toxicity and ecotoxicity. Compared to the other methods, IMPACT 2002+ emphasizes the category of ecosystem quality but gives little consideration to resource depletion.

CML2001

CML was developed by the Institute of Environmental Sciences at Leiden University of the Netherlands. It is a midpoint-oriented method, which includes characterization and normalization in the impact assessment process. The CML2001 method defines three categories of impacts: (1) baseline impact categories which are normally used in almost all LCA studies, (2) study-specific impact categories which are used depending on the available data and goal and scope of the study, and (3) other impact categories which require further elaboration (Suebsiri 2010).

2 Comparison of human toxicity impact categories in four LCIA methods

The human toxicity potential is defined as one category in CML2001. In IMPACT2002+, this impact category is divided into the two groups of carcinogens and non-carcinogens.

In EDIP 97, the category of human toxicity is subdivided according to the media in which humans are exposed to different substances: air (human toxicity air), water (human toxicity water), and soil (human toxicity soil). TRACI divides human toxicity into six groups: (1) toxicity transmitted by air which have non-cancerous effect to humans (human health non-cancer air), (2) toxicity transmitted by air which have cancerous effect to humans (human health cancer air), (3) toxicity transmitted by water which have non-cancerous effect to humans (human health non-cancer water), (4) toxicity transmitted by water which have cancerous effect to humans (human health cancer water), (5) toxicity transmitted by soil which have non-cancerous effect to humans (human health non-cancer ground surface soil), and (6) toxicity transmitted by soil which have cancerous effect to humans (human health cancer ground surface soil). A comparison of the human toxicity category in the four methods is shown in Table 1.

Table 1 Comparison of human toxicity impact categories in CML 2001, EDIP 97, TRACI and IMPACT 2002+

3. Comparison of Midpoint Reference Substances used in Four LCIA methods

The four LCIA methodologies use different midpoint reference substances for the same impact categories, as shown in Table 2. For example, the category of photochemical oxidation in CML 2001 and EDIP97 use kg ethene-equivalent as the reference substance, while the IMPACT 2002+ uses kg ethan-equivalent and TRACI uses kg NOx-equivalent for reference substances. All the reference substances are the primary causes for photochemical oxidation. Although the four LCIA methodologies use different reference substances and different mechanisms to calculate the results, the generated results demonstrated similar trends.

Table 2 Different midpoint reference substances used in four LCIA methodologies