Page 121 - Materials Chemistry, Second Edition

106 A. Bjørn et al.

constructed unit processes are of the actual unit processes that they are models for.
Representativeness of LCI data can be understood in three interrelated dimensions:
geographical, time-related and technological. Based on the goal definition and
knowledge about the studied product system, the scope definition must provide
guidance and requirements for the inventory analysis with respect to representa-
tiveness of LCI data, as explained below for each dimension of representativeness.
Besides serving as a guide for carrying out the inventory analysis, the representa-
tiveness of data should also be used in the interpretation of the results to reflect upon
the extent to which the product system model corresponds to reality (Chap. 12).

8.7.1 Geographical Representativeness

The geographical representativeness reflects how well the inventory data represents
the actual processes regarding location-specific parameters. Geographical repre-
sentativeness is important to consider because two processes delivering the same
product output, but taking place in two different locations (e.g. nations), can be
quite different in terms of the other flows (elementary flows, energy flows, material
flows and waste to treatment). Differences between unit processes can be caused by
geographical differences, such as local climate and proximity to natural resources,
and regulatory differences, such as energy taxes and emission thresholds. In
addition, when a mix of processes (market mix for attributional LCA and mix of
marginal processes for consequential LCA) is used to model the background system
or perform system expansion, the location of the mix used in the model versus the
actual location of the mix must be considered. For example, the electricity mixes of
Denmark (mainly coal and wind power) and Sweden (mainly nuclear and hy-
dropower) vary quite a lot; despite the close proximity of the two countries, see
Fig. 8.14.
This can in part be explained by geographical differences (Sweden has moun-
tains and therefore a potential for generating hydropower—Denmark is flat) and in
part from social and political differences (Sweden has nuclear power plants—
Denmark does not, largely due to public resistance).
Due to the importance of geographical representativeness the LCA practitioner
must in the scope definition define the geographical scope of the processes, or
combinations of processes, taking place in the product system. The starting point
should be the foreground system, where the locations of processes are typically
known with high certainty. The LCA practitioner can then proceed to defining the
geographical scope of upstream and downstream processes that typically are more
uncertain the more “process steps” from the key processes they are in the model.
The appropriate resolution of the geographical scopes (e.g. city, region, nation or
continent) depends on factors such as the spatial coverage of regulation (typically
following national borders), geographical variations (e.g. weather, climate) and the
spatial extent of markets (some markets are very local, while others are global).