Page 292 - Materials Chemistry, Second Edition

278 R.K. Rosenbaum et al.

result (i.e. with a small standard deviation) is not necessarily meaningful if it comes
with low accuracy regarding the information one is actually looking for.
In the LCA context, this can be illustrated using the different time horizons of the
global warming potential (GWP). When intending to capture potential impacts from
global warming of greenhouse gas (GHG) emissions, the GWP is integrated over
20, 100 and until the 4th IPCC Assessment Report (IPCC 2007) even over
500 years. It is intuitive that precision decreases with an increasing time horizon
due to the assumptions necessary to model and predict far into the future, but does
accuracy also automatically decrease with longer time horizons?
In order to answer that question, we need to consider that most GHGs stay much
longer in the atmosphere than 20 years. GWP20 is a very precise and probably
accurate indicator for the cumulative radiative forcing (i.e. the capacity to absorb
energy, which can be measured in the lab) of a molecule during 20 years, but it
neglects that this molecule may still be active long after. It is thus a very inaccurate
indicator for the total potential contribution of the molecule to global warming,
which is what we are usually interested in for an LCA study (unless the goal and
scope definition requires a focus on short-term impacts). Therefore, implicitly
assuming that GWP20 quantifies the (total) potential contribution of an emission to
global warming bears a risk of interpreting LCA results wrongly in spite of using an
indicator that is very precise, as it is inaccurate for the objective at hand (Fig. 11.3).
This example may seem somewhat obvious, but there are many other instances
of exactly this type of confusion that can be found in current LCA practice. Another
example is the comparison of the uncertainty of indicator results from different
impact categories. The GWP is generally perceived as a fairly certain midpoint
indicator whereas human toxicity is seen as a very uncertain midpoint indicator, an
argument that is sometimes used to justify the omission of toxicity characterisation
from an LCA study. It is worth reflecting whether this direct comparison of
uncertainties makes sense by looking at the environmental relevance of what both
indicators are actually quantifying.
We discussed in Chap. 10 that GWP is the time-integrated radiative forcing of a
substance per unit mass emitted. The input data required to calculate it are relatively

Fig. 11.3 GWP20 more
precise but less accurate from
an LCA perspective than
GWP100