Feb 1st, 2024
What is a Life Cycle Impact Assessment (LCIA)? Which Methods are Available in the ecoinvent Database?
Learn more about Life cycle impact assessment (LCIA) and LICA methods implemented in ecoinvent here
Why do LCIA Results of a Product Change Between Versions?
Modeling of environmental impacts depends on an overwhelming number of assumptions, and it is sometimes hard to know which ones are the most important. It is equally difficult to know what are the consequences of changing an assumption. When confronted with an impact score changing across different versions of the database, a detailed analysis is required to know the exact sources of the change. However, the potential reasons can be grouped in three categories:
Dataset Updates
When newer or better quality data become available, or when an error is detected, values within the dataset at hand are changed. These can be quantities or nature of certain emissions, natural resource and intermediate products. If the dataset is multi-output and economic allocation is applied, a change in price will affect the results of the dataset. The allocation method might also have changed.
Supply Chain Changes
LCA takes into account the impact of the whole supply chain of products and services. If a dataset is changed, the impact of the other datasets will also be changed. This is especially true for commonly used datasets like electricity, transport and infrastructures. Linking a change in a dataset to a change of score in another dataset require a software tool and careful inspection of the supply chain.
Change in Impact Assessment Methods
The characterisation factors provided by the method developers are regularly updated, and sometimes, there is a mistake in the implementation made by ecoinvent. The impact scores are obviously sensitive to a change of characterisation factors.
To help analyse changes in scores, a change report is produced for each new version of the database. The spreadsheet in annex of the report lists all the dataset changes that affect scores, and the word document clarifies the reasons for those changes. This report is available in the file section of ecoQuery.
What Does LT and w/o LT Mean?
In a nutshell: LT stands for long-term and classifies emissions, usually from landfills, which are release into the air or leach to the groundwater more than 100 years after the landfilling happened.
What’s the Difference?
LT stands for long-term. An emission is classified as “long-term” if it is released to the environment more than 100 years after the activities considered in the life cycle took place. Decisive for the classification “long-term” is thus the moment at which an emission is released in the environment and not the moment at which it causes its impact. It is, therefore, different from long-term impacts that would be caused, for example, by the bioaccumulation of a pesticide in the food chain.
Why Does ecoinvent Distinguish Between LT and w/o LT?
So far no consensus has been reached among LCA experts if and how long-term emissions should be taken into account. Until the debate is settled, ecoinvent allows users to decide themselves whether to include it. Therefore, long-term emissions are reported separately via two subcompartments explicitly labeled as “long-term”, namely “air, low population density, long-term” and “water, ground-, long-term”. Those exchanges are exclusively present in waste treatment datasets, where it is assumed that the active landfill maintenance ends after 100 years. Since long term emissions are in the environment 100 years less than short term emissions, their CF should be lower. However, most LCIA methods do not provide a distinction of time horizons. Thus, two options are possible:
Attribute the same CF to both short term and long-term emissions, leading to an over-estimation of the impacts. Attribute no CF to the long-term emission, leading to an under-estimation of the impacts.
ecoinvent provides the results for both options, allowing you to test the sensitivity of your conclusions. The results without long-term emissions are available in methods explicitly labeled as “w/o LT” or “no-LT”.