Local data versus generic LCA databases: the case of cereal production and uses in Wallonia (Belgium)
With respect to sustainable development expectations, Life Cycle Analysis (LCA) can be a useful tool to assess agricultural products performances especially if local specificities are accounted for.
Based on a comprehensive description of the Walloon (southern Belgium) cereal sector (Delcour et al. 2014), the “ALT-4-CER” project studies food, feed, fiber and fuel uses of cereal resources through an exhaustive comparison of production and processing chains through environmental and socioeconomic LCAs adapted to the local context.
Environmental LCA focuses on local specificities regarding cultivation and processing in comparison with generic data from commonly used international databases. Cereal cultivation in Wallonia has been modeled from representative farms’ accounting data (DAEA, 2010) including areas, yields, inputs and machinery. Cultivation practices are based on interviews with farmers and experts. Nutrient in excess from previous crops in the rotation are accounted for. Fuel consumption is calculated from (Rabier et al. 2008). Inputs nutrient and heavy metal contents are based on (Piazzalunga et al. 2012). Field emissions were calculated from (IPCC, 2006) and ecoinvent models (Nemecek et al. 2007). Cereal processing is based on existing facilities (mills, animal feed industries, biofuel and biogas plants, etc.), on an operational model (Mathot et al. 2013) for animal husbandry and on literature for missing and future technologies. Life Cycle Impact Assessment is supported by a composite mid-point level method constructed according to (European Commission et al. 2011) with pertinent impact categories for agricultural products (Guinée et al. 2004).
Differences between specific local data for cereal cultivation and those commonly found in LCA databases are reflected in the results. Walloon wheat for example has smaller impacts in most categories. These results are partly due to Walloon larger yields (10 t.ha-1 at 15% humidity (DAEA 2010; DGSIE 2010), compared to 6.4 in ecoinvent data (Nemecek et al. 2007)). However these differences cannot be explained by yields only, as shown by introducing Walloon yields into ecoinvent practices (Nemecek et al. 2007) (Figure 1). It is therefore recommended not to directly extrapolate existing cultivation LCI data on a yield basis but use specific data wherever possible.
For many food products, the cultivation step weighs substantially in LCA results. The present work shows how important it is to use specific data for this step when assessing environmental impacts of crop-based products. For instance, differences between flour from Walloon wheat and from ecoinvent wheat (Nemecek et al. 2007) are highly or extremely significant for most categories, as concludes the Monte Carlo uncertainty analysis (Figure 2). Insignificant differences, i.e. terrestrial acidification and eutrophication, are due to field emissions based on IPCC models for both processes.
As LCA is an iterative process, further improvements could include the use of more specific emission models, such as CERES-EGC (Gabrielle et al. 2006) or PEST-LCI (Birkved et al. 2006), and missing toxicity factors in USETOX (Rosenbaum et al. 2008). Additional environmental impacts categories such as biodiversity or carbon stock changes may also be included provided reliable methods are workable.
Birkved M, Hauschild M (2006) PestLCI – A model for estimating field emissions of pesticides in agricultural LCA. Ecological Modelling 198(3-4): 433-451
DAEA (2010) Farm’s accounting data. SPW – Agricultural Economic Analysis Department, Namur, Belgium
Delcour A, Van Stappen F, Gheysens S, et al. (2014) Etat des lieux des flux céréaliers en Wallonie selon différentes filières d'utilisation. Biotechn Agron Soc Environ, in press
DGSIE (2010) Recensements agricoles de 1995, 2000, 2005, 2007, 2008, 2009, 2010, 2010. SPF Economie - Direction générale Statistique et information économique, Brussels, Belgium
European Commission, Joint Research Centre, Institute for Environment and Sustainability (2011) International Reference Life Cycle Data System (ILCD) Handbook - Recommendations for Life Cycle Impact Assessment in the European context. Publications Office of the European Union, Luxembourg
Gabrielle B, Laville P, Duval O, et al. (2006) Process-based modeling of nitrous oxide emissions from wheat-croppped soils at the sub-regional scale. Global Biogeochemical Cycles 20(4):1-13
Guinée J, Gorée M, Heijungs R, et al. (2004) Handbook on Life Cycle Assessment - Operational Guide to the ISO Standards. Kluwer Academic Publishers
IPCC (2006) Guidelines for National Greenhouse Gas Inventories. In: L.B. Simon Eggleston, Kyoko Miwa, Todd Ngara and Kiyoto Tanabe (eds) IPCC National Greenhouse Gas Inventories Programme, Intergovernmental Panel on Climate Change, 642 p
Mathot M, Loriers A (2013) Weden 1.0. Walloon Agricultural Research Centre, Libramont, Belgium
Nemecek T, Kägi T (2007) Life Cycle Inventories of Swiss and European Agricultural Production Systems. Final report ecoinvent V2.0 No. 15a., Agroscope Reckenholz-Taenikon Research Station Swiss Centre for Life Cycle Inventories, Zurich and Dübendorf, 360 p
Piazzalunga G, Planchon V, Oger R (2012) Evaluation des flux d'éléments contaminants liés aux matières fertilisantes épandues sur les sols agricoles en Wallonie - Rapport final CONTASOL. Walloon Agricultural Research Centre, Gembloux, Belgium, 201 p + annexes
Rabier F, Miserque O, Pekel S, Dubois G, Noël H (2008) Guide of running costs for farm equipment: a simple tool for decision making. In: Huyghebaert B, Lorencowicz E, Uziak J (eds) Farm machinery and process management in sustainable agriculture - III International Scientific Symposium, Gembloux, Belgium, pp 43-50
Rosenbaum RK, Bachmann TM, Swirsky GL, et al. (2008) USEtox – the UNEP-SETAC toxicity model: recommended characterisation factors for human toxicity and freshwater ecotoxicity in life cycle impact assessment. The International Journal of Life Cycle Assessment 13: 532-546