The life cycle impact assessment (LCIA) uncertainty estimation due to the choice of LCIA methods

In Life Cycle Impact Assessment (LCIA), life cycle inventory (LCI) is converted to environmental impacts with different characterization factors provided in an LCIA method. There are multiple LCIA methods in current LCIA... [ view full abstract ]

In Life Cycle Impact Assessment (LCIA), life cycle inventory (LCI) is converted to environmental impacts with different characterization factors provided in an LCIA method. There are multiple LCIA methods in current LCIA practice; they differ in many aspects, such as impact categories, characterization factors, and normalization factors. LCIA methods are often built in LCA software: the LCIA results are calculated based on a particular LCIA method that is chosen by the user. The uncertainty due to the choice of different LCIA methods based on particular products has been analyzed for years. In our study, we seek to evaluate the uncertainty of LCIA results of all processes in the US LCI database, by using all LCIA methods provided in the SimaPro software.

First, the uncertainties due to the connections to characterization factors are evaluated. It is often assumed by the users that the uncertainties in the impacts are caused by the uncertainty in different characterization factors. However, same impact assessment category in different LCIA methods can have their characterization factors for different environmental effects. For example, one LCIA method (TRACI) have characterization factors for 94 chemical components in such as Global Warming Potential category, while another method (IMPACT2002) has 81. It indicates that the uncertainty can be caused by merely dis-connections.

Then, the uncertainty in LCIA impacts due to “cut-off” criteria are analyzed. Cut-off values are the minimum reported values in LCI; the values are often percentage values of the total amount of emission. Trace amounts of emission are often not reported in the inventory data. However in LCIA, the contributions to impacts from emissions are also determined by their corresponding characterization factors. For example, according to the inventory data in the US LCI database, to produce 1000 kg Single-ply, 0.05 g HCFC-22 and 0.003 g Halon 1301 are emitted to the air. The characterization factors of ozone depletion are 0.02 and 16 kg CFC-11 eq, resulting in 0.001 and 0.048 g CFC-11 eq. If fixed cut-off value (i.e. 0.01 g) is used, the impact from Halon 1301 is ignored in the inventory; however, its ozone depletion value is 48 times larger than the value for HCFC-22.

The uncertainties of environmental impacts due to the connections, cut-off values and other factors will be evaluated and represented separately in ranges. The outcomes of the study can help the LCA software users to better understand the uncertainty of LCIA; it can also provide guidance to future inventory data reporters, such as using LCIA-driven cut-off values.