Page 114 - Materials Chemistry, Second Edition

8 Scope Definition 99

equations, which makes it difficult to model what will actually happen, short-term
and long-term, when, for example, a light-bulb is turned on, compared to a situation
where it is not turned on. There is therefore a risk of using wrong marginal pro-
cesses and this is problematic because LCA results are often quite sensitive to the
choices of marginal process (e.g. natural gas vs. wind for electricity supply). These
considerations have lead ILCD to pragmatically recommend using average pro-
cesses in the background system. It must be mentioned that some LCA experts
prefer to pursue the ideal by using marginal processes in Situation A studies, which
conflicts with the presented ILCD recommendations.

Situation B
Situation B concerns meso/macro-level decision support (see Chap. 7). ILCD
recommends the same modelling choices as for Situation A, with the exception that
background processes in the studied product system that have been identified as
being affected by structural changes as consequence of the analysed decision are
recommended to be modelled as mix of the long-term marginal processes. The logic
behind this exception is that marginal processes for suppliers that experience
structural changes are easier to identify than marginal processes for suppliers that
just experience changes in terms of the volume of products they deliver. The reason
for the focus on the long-term marginal is that the consequences studied under
Situation B are generally long term. Still, identifying the correct long-term marginal
processes in Situation B can be challenging and this is why it is pragmatically
recommended to use a mix of possible long-term marginal processes, rather than
actual long-term marginal processes, such as the mix for electricity shown in
Fig. 8.9. Chapter 9 addresses the calculation of such a mix. In light of the uncer-
tainty involved, we advise to model the LCI using a range of different mixes to
analyse how sensitive results are to the estimated mix (see Chap. 12). As for
Situation A studies, some LCA experts prefer to pursue the ideal of using a fully
consequential approach by only using marginal processes (either single process or a
mix) in Situation B studies, which conflicts with the presented ILCD
recommendations.
Situation C
Situation C relates to accounting, meaning that studies are not to be used to directly
support decisions and are of purely descriptive nature, often describing what has
already happened. Situation C1 considers interactions with other systems and ILCD
recommends handling this interaction via system expansion (for solving multi-
functional processes where subdivision is not possible) and use of average pro-
cesses in the background system. This means that the recommendations of ILCD in
practice are similar for Situation A and C1, even though the modelling ideals of
Situation A and C1 are different. Situation C2 disregards interactions with other
systems and ILCD therefore recommends that allocation be systematically used to
solve multifunctional processes, provided that subdivision is not possible. Note that
this conflicts with the ISO hierarchy, according to which system expansion should
be performed when possible instead of allocation.