Problems from including technospheric parameters in characterization factors for natural resources

Poster (konferens), 2022

The life cycle inventory (LCI) analysis generally considers a product system in the technosphere, whereas the life cycle impact assessment (LCIA) is generally concerned with impacts in nature. However, in LCIA methods for natural resources, we have noticed a tendency to include technospheric parameters. This practise, which deviates from the predominant use of parameters related to environmental processes in characterization factors for emissions, has not received much attention in the LCA community. Here, we discuss a number of problems arising from such inclusions. Three types of technospheric parameters found in characterization factors for natural resources were analysed: (i) extraction rates, (ii) recycled contents, and (iii) prices. Extraction rates vary over time, and frequent updating is therefore needed to avoid outdated characterization factors. Furthermore, the inclusion of extraction rates in the characterization factors creates an interdependency between the LCI analysis and the LCIA, since extraction rates are also part of the inventory modelling. We show that such interdependencies can potentially lead to counterproductive information. Regarding recycled contents, when inventory data with recycled content are matched with characterization factors also taking recycled content into account, the benefit of recycling is double counted. Furthermore, it introduces a risk of inconsistency: the recycled contents in the characterization factors may not match those in the LCI analysis. In addition, characterization factors based on recycled contents are also time sensitive. Prices are commonly used in economic allocation in the LCI analysis. When they are also used in characterization factors, there is a risk of inconsistency if these prices are not the same as those used in the allocation. In addition, prices are very time sensitive, potentially fluctuating notably even on a daily basis. There are possible solutions to some of these problems, such as frequent updating of characterization factors and avoiding economic allocation. However, these solutions come at a cost. For example, frequent updating of characterization factors is work intensive, and economic allocation may be otherwise recommendable in some studies. For the LCI-LCIA interdependency, we see no obvious solution. Considering the identified problems, we recommend further critical discussions on the inclusion of technospheric parameters in characterization factors for natural resources.