AGRIBALYSE data perimeter
Perimeter - Key methodological hypothesis
Last updated
Perimeter - Key methodological hypothesis
Last updated
Analysing the production systems for all the food consumed in mainland France represents a considerable scientific challenge, given the scale and complexity of the production, processing and distribution systems in a globalised market. To carry out the calculations, we had to rely on a large amount of statistical data, supplemented by hypotheses and expert opinion.
Approximations and simplifications had to be made, and these are documented in the methodology report. Hypotheses were always constructed as "systematically" as possible in order to avoid bias, and by concentrating efforts on the structuring parameters with the greatest impact, known thanks to the consortium's expertise. Particular effort has therefore been focused on the agricultural stage, which for the majority of products represents the phase of the life cycle with the greatest impact, while the processing, logistics, packaging and use phases have been simplified.
The environmental indicators available on the impacts associated with agricultural production are calculated for 1 kg of raw agricultural product (e.g. 1 kg of wheat leaving the farm). The calculations take into account all the processes upstream (manufacture of inputs) and in the field (cultivation operations), stopping at the end of the field. Impacts relating to transformation processes, logistics, transport, packaging and product use are not taken into account in this agricultural section.
The data available for the agricultural section is rich: it concerns average products, and variations (integrated farming, organic farming, grass-fed farming, etc.), reflecting the variability of production methods and their relative environmental impacts. The "standard/average" practices are the result of the representativeness of these different production methods in France today.
A specific calculation tool for upstream agriculture
The analysis of agricultural systems using the Agribalyse® methodology is greatly facilitated by the use of INRAE-Cirad's MEANS-InOut software upstream of SimaPro. The MEANS-InOut software can be used to describe agricultural technical itineraries, and is available online after subscribing to a service contract. For more information: https://www.inrae.fr/means
The Organic LCA project has produced data on a large number of organic farming operations. These data are included in the current database. One of the aims of the work carried out as part of the Organic LCA project was to illustrate the diversity of organic farming practices. However, for certain products (such as beef), the project was unable to model an "average organic product" produced in France, due to the diversity of practices and a lack of statistics on organic farming production. As a result, the AGRIBALYSE database cannot currently be used to compare the environmental impact of the average conventional cow's milk in France and the average organic cow's milk in France. Work is currently underway to remedy this shortcoming.
Data based on "consumption mixes"
For the raw agricultural products used in the manufacture of food products, ‘consumption mixes’ have been calculated to take account of imported products. For example, the impacts of the ‘standard conventional tomato’ used in pizza represent the weighted average of the impacts of the tomatoes mainly used for processed products (i.e. 18% of tomatoes from French production, 46% of Italian tomatoes and 36% of Spanish tomatoes).
"Average/standard" products representative of products consumed in France
The impact indicators provided by AGRIBALYSE (in its simplified version) represent indicators calculated for "standard" products, the most widely consumed in France. For example, we can see the impact of a ‘standard’ Margherita pizza, made from "standard" conventional tomatoes, standard "conventional" Gruyère cheese and standard "conventional" ham, produced using the most common production systems today, and the most common packaging used for this type of product. The impacts of the "conventional standard tomato" in the pizza represent the weighted average of the impacts of the tomatoes most commonly used for processed products (i.e. 18% of tomatoes from French production, 46% of Italian tomatoes and 36% of Spanish tomatoes).
At this stage, this simplified version does not provide any "variations" at food level : there is only one set of environmental indicators for each of the 2,500 Ciqual products, and therefore no variation for each of these products. This means that the simplified version of Agribalyse® data cannot be used to compare food products in the same category (e.g. two Margarita pizzas from competing brands, or an organic Pizza Margarita and a non-organic Pizza Margarita). These data therefore reflect a ‘majority standard’ reality, and allow food products from different categories to be compared.
With the exception of a few "test products", there is no breakdown of the available indicators according to season or local produce sourcing.
To make the data easier to understand, details of the impacts for each of the 2,500 products are provided by major stage, in the simplified file: agricultural production, transport, packaging, distribution and use.
For compound products, the environmental impacts are also detailed by ingredient.
Focus on season and air transport
For a number of key products, a number of variations have been made: for example, the impacts for tomatoes and strawberries have been broken down for in-season tomatoes/strawberries and out-of-season tomatoes/strawberries. A breakdown has also been made for products transported by air (for example, green beans from Kenya versus ‘average’ green beans produced in France).
The current LCA framework provides structuring and useful elements to shed light on the environmental aspect of the sustainability of agricultural and food systems. However, this framework still needs to be consolidated and improved to better cover all the environmental issues facing the sector. The usual LCA indicators do not currently capture all the potential environmental impacts, and modelling needs to be improved for certain impacts. Among the main issues for the food sector, the limitations and development needs of the LCA and Agribalyse indicators relate in particular to :
Better representation of all pressures and practices favourable to biodiversity
An improved account of carbon storage/removal processes in soils
A better description of transformation processes and the use of by-products in the agri-food industry
A better description of the degradation of pesticides in the environment, and their impact on human health and ecosystems
A better spatialisation of water consumption
In the short term, this means that when LCA is used to compare the environmental impacts of different production systems (organic versus conventional farming, intensive/extensive livestock farming, etc.) and processing systems, these limitations must be explicitly highlighted. It may therefore be useful to use additional indicators, to be defined according to the context and the scale of the work. These additional indicators could be, for example, Areas of Ecological Interest for biodiversity, animal welfare indicators or socio-economic indicators.
Work to improve the LCA framework is underway at international, European and French level (via the GIS REVALIM in particular). Projects carried out as part of the AGRIBALYSE programme and then the GIS REVALIM have already proposed methodological improvements, but not all of these have yet been incorporated into the database. Their results will be gradually incorporated into AGRIBALYSE in the future. They will make it possible to consolidate environmental analyses to provide a more complete picture of the interactions between food and the environment.
