Page 112 - Materials Chemistry, Second Edition

L1644_C03.fm Page 87 Tuesday, October 21, 2003 3:11 PM

eco-indicator 99 approach (Goedkoop and Spriensma, 1999). In Japan, impact assess-
ment models are currently developed according to this approach (Itsubo and Inaba,
2000), which starts from the main values in society, connected with areas of protection.
From these values and connected endpoints the modeling goes back to the emissions
and resource consumption (Udo de Haes and Lindeijer, 2001).
Figure 3.6 shows the steps that can be involved if a practitioner wishes to take
an LCA study from the inventory stage to valued scores via midpoints and endpoints
in the impact assessment. Not all possible environmental loads can be considered
in the inventory because data are not available for all of them. Based on the inventory
table, two different routes to arrive at valued scores, representing the routes taken
when using midpoint and endpoint approaches, are presented (Bare et al., 2000).
On the one hand, the impact categories that can be expressed in the form of midpoints
are directly presented as valued scores; on the other hand, as far as possible according
to current knowledge, the impacts are expressed in the form of endpoints by relating
the midpoints to endpoints or by modeling effects directly from the inventory to the
endpoints. Then several endpoints can be aggregated to a valued score if the selected
weighting scheme allows it.
At the moment, the availability of reliable data and sufficiently robust models
to support endpoint modeling remains quite limited. Uncertainties may be very high
beyond well-characterized midpoints. As a result, a misleading sense of accuracy
and improvement over the midpoint indicators can be obtained. One of the biggest
differences between midpoint and endpoint approaches is the way in which the
environmental relevance of category indicators is taken into account. For midpoint
approaches, the environmental relevance is presented as a qualitative relationship,
while endpoint modeling can facilitate more informed and structured weighting
(Bare et al., 2000; UNEP DTIE, 2003).

Interpretation

? V Procedure
? I M E a with
n i ? n l
v d d u endpoint
? ? e
e p p approach
n o o ? d
t i i S
o n ? n Procedure
c
r ? t t with
? o
y s s r midpoint
? e approach
s
?

FIGURE 3.6 Some basic differences between the midpoint (lower row of swinging arrows)
and the endpoint approach (upper row of swinging arrows). (From Bare, J.C. et al., Int. J.
LCA, 5, 319–326, 2000.With permission)

© 2004 CRC Press LLC