Provision of LCI data in the European aluminium industry Methods and examples

The development of robust and up-to-date generic life cycle inventory data for materials is absolutely crucial for the LCA community since many LCA studies rely on these generic data about materials. LCA databases and software usually include within their package such generic LCI datasets. However, in many cases, the quality of these data is poor while the methodology and the models used for their development are rarely accessible or transparent. This paper presents the development of robust European LCI datasets for the production of primary and recycled aluminium ingots and for the transformation of aluminium ingot into semi-finished products, i.e. sheet, foil and extrusion.

The environmental data have been collected through an extensive environmental survey, organised among the European aluminium industry, focusing on the year 2005 and covering EU27 countries as well as EFTA countries (Norway, Iceland and Switzerland). From this survey, European averages, i.e. foreground data, have been calculated for the direct inputs and outputs of the various aluminium processes. Using the GaBi software, the foreground data have been combined within LCI models integrating background LCI data on energy supply systems, ancillary processes and materials. For the primary aluminium production (smelters), a specific model for the electricity production has been developed. The methodology for the data consolidation and for the development of the various models is explained as well as the main differences between the new modelling approach and LCI models used in the past. An independent expert has critically reviewed the entire LCI project including data collection, models development, calculation of LCI data and associated environmental indicators.

As confirmed by the critical review, the new LCI datasets for aluminium ingot production and transformation into semi-finished products have been developed though a robust methodology in full accordance with ISO 14040 and 14044. Most significant environmental data and LCI results are reported in this paper with an emphasis on energy use and the major emissions to air. The full environmental report, including the critical review report and the calculation of environmental indicators for a pre-set of impact categories, is available on the website of the European Aluminium Association (EAA 2008). Whenever possible, the updated European averages and the new LCI data are compared with previous results developed from two past European surveys covering respectively the years 2002 and 1998. For the aluminium processes related to primary production, European averages are also benchmarked against global averages calculated from two worldwide surveys covering the years 2000 and 2005.

While some data evolutions are directly attributed to the variation of foreground data, e.g. raw materials consumption or energy use within the aluminium processes, modifications related to the system boundaries, the background data and the modelling hypotheses can also influence significantly the LCI results. For primary aluminium production, the evolution of the foreground data is dominated by the strong decrease of PFC (perfluorocarbon) emissions (about 70% since 1998). In addition, the electricity structure calculated from the refined electricity model shows significant differences compared to previous models. In the 2005 electricity model, the hydropower share reaches 58% while coal contributes to 15% only of the electricity production. In 1998, the respective share of coal-based and hydro-electricity were respectively calculated to 25% and 52%. As a result, the electricity background LCI data are then significantly affected and influence also positively the environmental profile of primary aluminium in Europe. For the semi-production processes, the reduction of process scrap production, especially for extrusion and foil, demonstrates the increase of process efficiency from 1998. In parallel, a significant reduction of energy use is observed between 1998 and 2005. However, this positive trend is not fully reflected within LCI data due to the significant contribution of the background electricity data. The choice of the electricity model plays also a critical role for these transformation processes since electricity production contributes to about 2/3 of the consumption of the non-renewable energy and to about the same level of the air emissions. In such a case, the move from the UCPTE electricity model used in the past towards the EU25 electricity model used for the development of the updated LCI data has a detrimental effect on the environmental profile of the three LCI datasets respectively related to sheet, foil and extrusion. In addition to energy and process scrap reduction, the reduction of the VOC (volatile organic compounds) emission is also a major trend in foil production. Finally, for old aluminium scrap recycling, the new LCI data show a dramatic improvement regarding energy efficiency, reinforcing the environmental soundness of promoting and supporting aluminium recycling within the aluminium product life cycles.

This paper shows the development of generic LCI data about aluminium production and transformation processes which are based on robust data, methodologies and models in full accordance with ISO 14040 and 14044 standards, as confirmed by the critical review. The publishing of these LCI datasets definitely shows the commitment of the European aluminium industry to contribute in a transparent, fair and scientifically sound manner to product sustainability in a life cycle thinking perspective.

Software houses and LCA practitioners are invited to update their generic European data on aluminium with the herewith datasets. Even if the quality and the completeness of these LCI data reach a high standard, some areas for data improvements have been identified, as described within the review report. Land use, water use and solid waste treatment appear as three priority areas for data refining and improvement. The land use dimension, particularly meaningful for bauxite mining, is not covered in the current LCI data while it is now integrated within many LCA studies. Up to now, the reporting of meaningful and robust data on water origins and use have not been possible due to the huge discrepancies between the surveyed sites combined with the difficulty to report coherent input and output water mass flows. The development of water data, only focussing on water-stressed areas, will most probably make more sense in the future. Finally, collecting more qualitative information about solid waste processing and treatment will help to include such operations within the system boundaries and to model their associated air, water and soil emissions.