LCA Food 2012 in Saint Malo, France! - Manifestations et colloques ...

PARALLEL SESSION 3A: LAND USE CHANGE 8 th Int. Conference on LCA in the
Agri-Food Sector, 1-4 Oct 2012
clude the additional functions related to the co-products) and allocation based on physical (e.g. mass, energy)
or other (market price) relationships between the co-products.
In the substitution method, the system is expanded with “avoided” processes to remove additional functions
related to the soybean meal (Guinee et al., 2009). Two alternative scenarios for the substitution of soybean
meal were implemented. Soybean meal is mainly used as animal feed and these scenarios took into
account two product systems that are currently displaced by soybean meal co-produced with soybean oil in
Portugal according to experts from the soybean oil industry.
Substitution scenario “Imported Soybean meal” (ISM) describes the case in which soybean meal
co-produced with soybean oil in Portugal displaces direct soybean meal imports from Argentina
for animal feed purposes. Substitution credits for GHG emissions were quantified based on the
GHG intensity of soybean meal produced in Argentina given by Dalgaard et al., (2008): 721 g
CO2eq/kg meal (no LUC).
Substitution scenario “Imported Soybean” (IS) describes the case in which soybean meal coproduced
with soybean oil in Portugal displaces direct soybean imports from Brazil (also used as
animal feed in Portugal). The substitution credits for GHG emissions were quantified based on the
GHG intensity of soybean produced in Brazil (calculated in this paper: 432 g CO2eq/kg soybean,
no LUC and no-tillage) and assuming a displacement ratio of 0.85 kg soybean meal per kg of soybean,
based on an equivalent protein content.
In allocation, the multifunctional process is split up into two single functional processes (oil + meal)
based on specific relationships. In this paper, a sensitivity analysis of allocation based on mass, energy
(lower heating value, LHV) and market prices was performed. Table 2 shows the mass, energy and economic
data (based in 2010 monthly market prices from IMF, 2011) used to calculate the allocation factors for soybean
oil and meal. Three alternative economic allocation factors were calculated based on: the average annual
prices for oil and meal (“Av”); the monthly prices for which the price ratio between oil and meal is the
lowest (“Min”) and it is the highest (“Max”). It should be noted that soybean oil and meal have different
characteristics, being used for completely different purposes (e.g. energy purposes, food and feed production),
which complicates the selection of a particular allocation method.
Table 2. Soybean oil and meal mass ratios, LHV, market prices and corresponding allocation factors.
Mass Energy Market price
Mass ratio
(t/t oil)
Allocation
factor
LHV
(MJ/kg)
Allocation
factor
Prices a (US$/t)
Min Max Av
Allocation factor
Min Max Av
Oil 1.0 20% 36.6 36% 837 1208 925 38% 43% 41%
Meal
a
IMF (2011)
4.1 80% 16.3 64% 335 388 331 62% 57% 59%
3. Results and discussion
Fig. 2 presents the results of a sensitivity analysis conducted to illustrate the implications of different
multifunctionality approaches on the GHG balance of soybean oil (excluding LUC). It can be observed that
the results obtained for different approaches present significant variation: between -0.36 and 1.12 kg
CO2eq/kg. The lowest value was obtained when the substitution scenario “Imported Soybean meal” (ISM) is
adopted. In contrast, soybean oil had higher emissions for market-price-based allocation. This large variation
in results demonstrates the critical influence on the results of the method selected for handling co-production.
This also justifies the need to perform a sensitivity analysis, as recommended by ISO 14044:2006 (ISO,
2006).
Regarding allocation approaches, the GHG emissions calculated based on energy or market price allocations
are almost double compared results obtained using mass allocation. The results do not significantly
vary in consequence of annual market price variations, since soybean meal has a relatively high mass share
of the extraction process, and its price (mostly driven by the livestock feed industry) does not vary as much
as the oil price (Fig. 2). Furthermore, it can be said that it is soybean meal demand that determines the production
volume of the soybean oil extraction process.
264