Page 199 - Materials Chemistry, Second Edition

10 Life Cycle Impact Assessment 185

or spatialized archetypes (e.g. city-specific emissions, formation and back-
ground concentrations of particulate matter and related mortality rates)

Or

• Modelling impacts with a certain degree of spatial resolution (e.g.
sub-continental, country-level, sub-water-shed level or GPS grid-based),
allowing for a characterisation which can be specific to any given place of
emission or extraction

Both solutions require that the place of emission/extraction is known for each flow
in the inventory—either explicitly (e.g. by country or geographical coordinates such
as latitude and longitude) or regarding the most representative archetype. In order to
support a spatially differentiated impact assessment, the life cycle inventory must thus
not be aggregated to present one total intervention per elementary flow since this will
lose the information about location of the interventions which is needed to select the
right CF. Otherwise, generic global average CFs need to be used, leading to a higher
uncertainty due to the spatial variability not considered in the characterisation. In
contrast to the site-generic LCIA method, which provides one CF per combination of
elementary flow and intervention/emission compartment, the spatially differentiated
characterisation method provides one CF per combination of elementary flow,
intervention/emission compartment and spatial unit. For grid-based methods, this
may amount to thousands of CFs for each contributing elementary flow.
It depends on the impact category and emission situation to evaluate whether a
spatial or archetypal setup will give the more accurate solution (e.g. urban/rural
differences in particulate matter-related health effects might not be captured by spatial
2
models with typical resolutions lower than 10 10 km at the global scale, whereas
an archetypal model distinguishing between urban and rural emission situations
would capture such differences). It should be noted that country-based characterisa-
tion is not meaningful from a scientific point of view, as most impacts are not influ-
enced by political borders, although from a practical data-availability point of view
this currently not unusual practice is understandable and normally an improvement to
not considering the spatial variation at all. It should furthermore be noted that most
currently available LCA software fails to support spatially differentiated characteri-
sation, and therefore most LCAs are performed using the site-generic CFs.

10.2.3.11 The Units?

The unit of CFs for midpoint impact categories is specific for each category and
LCIA method chosen, and therefore discussed in detail in the corresponding section
dedicated in detail to each LCIA method in Chap. 40. However, two different
approaches can be identified—expression in absolute form as the modelled indi-
cator result (e.g. area of ecosystem exposed above its carrying capacity per kg of
substance emitted for acidification) or expression in a relative form as that emission