Page 182 - Materials Chemistry, Second Edition

MODELING THE AGRI-FOOD INDUSTRY WITH LIFE CYCLE ASSESSMENT 167

omitted. On the contrary, these phases have significant impact; the first one has
a great relevance, as it will be described in a following paragraph, the second is
absolutely important considering that, on average, more than fifteen per cent
of household food is wasted. In this sense consumer behaviour and kitchen
preparation cover an important role in food life cycle. More attention to this
aspect, in order to reduce the total environmental burden of food is necessary.
Even when studies stop at the farm gate, there are many data which are
excluded from the system boundaries. In commercial databases, very little data
are available for food processes and usually emissions are related to energy
consumption only, since specific plant emissions to atmosphere, water and soil
of food industrial plants are almost always not available. This leads to the data
quality issue that still remains one of the most important problems in food
LCA. Firstly, very few data are available in commercial databases. Secondly,
the strong site-specific nature of agricultural activities requires a region charac-
terisation which is in contrast to the site-independent nature of LCA. Research
is needed in order to have data sets applicable to the global food chains and
that are made available for LCA studies around the world. At the same time it
is necessary to increase the credibility of existing LCA data and it is also nec-
essary to implement the generation of more data, especially in the food and
drink sector. A project between UNEP and SETAC that addresses the need for
global guidance on LCI data collection and data processing into databases is
underway (Sonnemann, et ah 2011).
As far as allocation of food and drink product systems is concerned, stan-
dard principles should be applied, such as system expansion to include the
additional functions related to the co-products, or dividing the unit process
to be allocated into two or more sub-processes and collecting the input and
output data related to these sub-processes. If allocation cannot be avoided
through these methods, because no profiles for avoided co-products and/or
by-products or for sub-divided systems are available, then impacts have to be
allocated between them. Since agricultural systems are characterized by large
quantities of low value by-products (e.g. pomace from extra virgin olive oil,
stalks from wine) economic together with mass-based allocation is often per-
fectly suited for such systems.

7.2.3 Fertilizer and Pesticide Dispersion Models

For the development of plants and to increase crops, agriculture makes
extensive use of fertilizers and chemicals for weed and pest control. The
quantity and quality of these products, in terms of their production and the
resulting emissions, directly affect the results of LCI and LCIA. Fertilizer
and pesticide use contributes strongly to impact categories such as global
warming, acidification, eutrophication, human toxicity and ecotoxicity. For
the food LCA practititioner, retrieval of data on the production of fertilizers
and pesticides but also the estimation of the output of their use is particu-
larly difficult. Many studies report that one of the most complex phase of
a food LCA is just this.