Back

Advanced regionalization in LCA: An approach using GIS logic
Chris Mutel,*   ETH Zurich
Stephan Pfister,   ETH Zurich
Matthias Kestenholz,   ETH Zurich
Stefanie Hellweg,   ETH Zurich

Regionalization produces new insights and more accurate estimations of damage to human health and ecosystems, but is not widely practiced in the LCA community. However, several regionalized LCIA methods are available, and the only barriers to widespread regionalization are incorporation into current models and provision of accurate geographic information in inventory databases. Generic approaches to regionalization have been demonstrated, but it is possible to incorporate regionalization in a deeper sense in LCA.

We demonstrate a method to use regionalized LCIA methods with geographic boundaries which do not match the geographic units used in life cycle inventory databases. We integrate GIS technologies directly into the LCA calculation procedure, allowing for regionalization without any additional effort. We have integrated this method into our LCA software package, called Brightway. We also show how generic processes which are defined on a continent level can be replaced by commodities which are allocated to different geographic units. Data on the geographic distribution of commodity production is available, e.g. the new UN data site. We demonstrate regionalization using case studies on steel and electricity production in Europe, and water use for food production in various countries.

Because GIS systems allow for operations on arbitrary geographic areas, LCIA designers are no longer limited to grids or country boundaries. Instead, models can be designed to fit ecoregions, regional boundaries, watersheds, or other geographic areas. This allows for better LCIA methods, and a better fit between LCIA modelling results and characterization factors.

We discuss the challenges posed by detailed regionalization, including the uncertainty of input data, characterization factors, and the volatile nature of production data. Although geographic agglomeration is said to reduce uncertainty, it does so by losing important information on the specific affected geographic regions or ecosystems. Our LCA software will be made publicly available during the conference.


* corresponding author: mutel@ifu.baug.ethz.ch