Location: GLO - Global
This dataset is based on two sources: the European Aluminium Association 2005 LCI data (EAA, 2008); and (2) the ecoinvent v2.2 dataset for the same activity (Althaus, 2009), itself largely based on EAA data from 2000. Priority was given to data contained in EAA (2008); Althaus (2009) was used only in cases where it was found to be more complete that EAA (2008). Note that both sources are European in scope - this GLO dataset is therefore an extrapolation.
Remelters use mainly reverbatory furnaces, and this dataset is based on this furnace technology only.
The dataset excludes, for lack of access to data, salt slag and dross recycling. According to EAA, this data gap represents approximately 10-20% of potential impacts, as calculated using the CML 2001 method.
The following emissions to air are reported in EAA 2008 but excluded from this dataset: CO (0.095 kg/tonne secondary aluminium), Dust/particulates (0.041 kg/tonne secondary aluminium), NOx (0.329 kg/tonne secondary aluminium), SO2 (0.051 kg/tonne secondary aluminium), and VOC (0.050 kg/tonne secondary aluminium). These are assumed to be combustion emissions or electricity production emissions, conclusion arrived at by comparing gate-to-gate emissions with the cradle-to-gate emissions also reported in EAA 2008. Their inclusion wouldhave led to double-counting.
While this dataset represents the remelting of post-consumer scrap aluminium, the inputs and outputs reported by EAA (2008) actually refer to refining of new scrap. In the ecoinvent database, both the remelting of post-consumer scrap and new scrap is represented by two distinct but mostly identical datasets to enable the appropriate modeling of aluminium mass flow through the economy. The use of new scrap remelting to represent prepared old scrap remelting is thought to be more accurate than the use of refining as a proxy for this activity. The extrapolation is accounted for in the uncertainty.
Undefined unit processes (UPRs) are the unlinked, multi-product activity datasets that form the basis for all of the system models available in the ecoinvent database. This is the way the datasets are obtained and entered into the database by the data providers. These activity datasets are useful for investigating the environmental impacts of a specific activity (gate-to-gate), without regard to its upstream or downstream impacts.