National and farm level carbon footprint of milk – Methodology and results for Danish and Swedish milk 2005 at farm gate
Schmidt J H, Dalgaard R (2012)
Arla Foods, Aarhus, Denmark.
Arla Foods wants to estimate and track the development in greenhouse gas (GHG) emission per kg raw milk – both at farm level, national level as well as corporate level which include emissions in several countries. The current report concerns a CF model for raw milk from cradle to farm gate.
The modelling of life cycle emissions for agricultural products is associated with several challenges. The production systems are most often characterised by having several co‐products, and the most significant emissions are related to biological processing, such as enteric fermentation and altering of nutrient balances as opposed to LCAs in other sectors where most emissions are related to the combustion of fuels (Schmidt 2010a). The modelling of co‐products is one of the major challenges in the modelling of life cycle emissions. The modelling of emissions in agricultural production systems involves a large number of activity and product parameters and the models (IPCC models for GHG‐emissions) are often related to significant uncertainties.
A key challenge for Arla is that different methods for calculating the carbon footprint (CF) are often used in the countries where Arla operates. The following relevant modelling approaches have been identified:
- Consequential modelling (CLCA), which is most often used in Denmark.
- Average/allocation or attributional modelling (ALCA), which is typically used in Sweden and the UK.
- In the UK a national CF guideline called PAS 2050 has been developed (PAS2050 2008; Dairy UK et al. 2010). PAS2050 is a sub‐set of the attributional modelling with some specific rules for specific activities.
- At industry level, the International Dairy Federation (IDF) has also completed a CF guideline specifically for milk and dairy products (IDF 2010). The IDF guideline is a sub‐set of the attributional modelling with some specific rules for specific activities.
Arla Foods therefore needs a flexible tool that enables different types of modelling depending on the context. It should be possible to calculate the CF at farm level and national level according to the used practises in the given country, but it should also be possible to compare results between countries and to calculate the aggregated CF at corporate level. The latter requires that the same model is used in all countries. The model developed in the present project, therefore have built‐in switches that enables to use the same data, but to get the CF results according to the different modelling approaches. Hence, the model makes it possible for Arla to compare results across markets as well as within markets. The purpose of the present project is to:
- Calculate a baseline for Denmark and Sweden for 2005 of the average CF for milk according to the four modelling approaches referred to above.
- Develop a tool to calculate the CF on farm level, which will help to follow the development in CF per kg milk according the same guidelines and approaches as for item 1.
Compared to a ‘normal’ CF model, the current model is generically described with input parameters and formulas. Then the same model can be used for calculating the CF baseline for different countries as well as farm specific CF. The generic model and country baseline results are described in the current report. All input parameters are described in an inventory report (Dalgaard and Schmidt 2012).
The special features and the generic nature of the Arla model require that the framework for the life cycle inventory is defined consistently. Therefore, before the actual CF model is described in chapter 4 to 9, the inventory framework is described in chapter 3.