IMPACT World+ / a globally regionalized method for life cycle impact assessment

IMPACT World+ is a life cycle impact assessment method which characterizes thousands of substances spanning across various compartments and sub-compartments of the environment. It differentiates 19 impact categories at midpoint level and 34 impact categories at damage level. For more information on IW+, refer to our website and scientific article. For information on the updates of IMPACT World+, you can register to the newsletter of CIRAIG.

The v2.1 update is the biggest update of the IMPACT World+ method in many years as it introduces new impact categories and updates many models with the latest available research.

IMPACT World+ comes in three interpretation levels: midpoint, expert and footprint. You can find explanations for these three interpretation levels here.

The expert and midpoint versions of IMPACT World+ also come with two different implementations regarding how to account for biogenic carbon. One with the traditional biogenic carbon neutrality approach (e.g., where biogenic carbon dioxide is set at 0 and biogenic methane is set at 27kgCO2eq for GWP100) and one including the uptake of biogenic carbon dioxide, where the release of biogenic carbon is therefore set at the same CFs as fossil carbon, but the uptake is with a negative sign, i.e., a -1/+1 approach for biogenic carbon (look for the files marked “(incl. CO2 uptake)”).

While we provide the -/+1 approach, we must make it clear to the users that this approach is heavily dependent on the quality of the inventory you are using, and that there are still issues currently with the LCI databases (even in ecoinvent 3.10). Furthermore, if you are using this approach, you either MUST adopt a cradle-to-grave approach to both account for the uptake and release of biogenic carbon (otherwise you will only account for the uptake of the carbon and have skewed results) or if you adopt a cradle-to-gate approach because you need to provide results to someone downstream of your supply chain you MUST communicate with that downstream user to tell them that they should account for the release of biogenic carbon in a -1/+1 approach also, otherwise, you and your downstream user will double count the benefits of using biogenic products, which is incorrect. This is especially true is the case of food products where the carbon emissions post consumption are typically not included in the inventories, which could result in a substantial under estimation of the impacts of this product over its life cycle.

Description of the files

- The dev file is a file useful for developers and maintainers of databases/datasets who wish to link IW+ 2.1 to their databases/datasets. It regroups all existing characterization factors of the IW+ LCIA method in an Excel format, using the terminology of IW+.

- The "ecoinvent" files are Excel files matching with "pure" ecoinvent and its flow name terminology (as in unaltered by various software) in an Excel table. This is useful if you are using ecoinvent outside of LCA software.

- The exiobase file links IW+ to the Exiobase GMRIO database. Once the file is read through pandas (pandas.read_excel()), the resulting matrix can directly be multiplied to the environmental extensions of exiobase (S, F or F_Y if using the pymrio package).

- The openLCA file can be directly imported in the openLCA software as a JSON-LD file.

- The SimaPro file can be directly imported in the SimaPro software as a CSV file.

- The brightway2 files are ecoinvent-version dependent, so you need to select the correct file to work with the correct version of ecoinvent. Else, some of the ecoinvent flows which name did change in between versions of ecoinvent would not be characterized, leading to underestimated results. To import a file, you need to pass through brightway2 itself (it cannot be done through the activity-browser for now). The function to import a .bw2package file is bw2.BW2Package.import_file().

- Finally, the source file regroups all the native information used by IW+ to derive the characterization factors. This file is primarily useful for the IW+ internal team. It is provided for transparency, as well as for curious users or users who wish to generate all these files themselves through the open-access code of IW+.

New indicators

- Plastic physical effect on biota

This new indicator measures the effect of plastic resins emitted in the water environments (both freshwater and marine) on biota, in PDF.m2.yr. It also comes with a midpoint indicator in CTUe. This indicator is the result of the work of the MariLCA working group.

- Fisheries impact

This new indicator measures the impact on biodiversity of fisheries activities. It is only assessed at the ecosystem quality damage level (in PDF.m2.yr). This is based on the work of Stanford-Clark et al.

- Marine ecotoxicity

In the v2.1 we decided to finally integrate these two ecotoxicity indicators, at the damage level only. These are based on an old version of Usetox (v2.02). As the v3 of Usetox is on the verge of being released, all ecotoxicity and toxicity categories will be updated in the next version of IW+.

- Terrestrial ecotoxicity

In the v2.1 we decided to finally integrate these two ecotoxicity indicators, at the damage level only. These are based on an old version of Usetox (v2.02). As the v3 of Usetox is on the verge of being released, all ecotoxicity and toxicity categories will be updated in the next version of IW+.

- Photochemical ozone formation

For this impact category, IW+ adopts what the ReCiPe methodology recommends. In their 2016 update, ReCiPe renamed the indicator "Photochemical oxidant formation" to "Photochemical ozone formation". In addition, they calculated the impact of this category on ecosystem quality. There are thus two corresponding impact categories at damage level: "Photochemical ozone formation, human health" and "Photochemical ozone formation, ecosystem quality"

Updated indicators

- All climate change indicators

IMPACT World+ v2.1 proposes the carbon neutrality approach (i.e., CO2-bio = 0) as well as the -1/+1 approach (CO2-bio uptake = -1 / CO2-bio release = +1). However, the latter is only available in the expert and midpoint versions. In the footprint version, the carbon neutrality assumption is still being used.

Furthermore, we added CFs for temporary storage of biogenic carbon that can be used (e.g., Correction for delayed emissions, carbon dioxide, biogenic).

- Climate change, human health

In the v2.0.1, we updated the GWP100 and GTP100 indicators following the recommendations of the AR6 from the IPCC2021. Now in the v2.1, we are also updating our damage indicators for climate change to follow the AR6 recommendations. Notably, the cumulative AGTP500 used in the derivation of these CFs was recalculated with updated equations (which we obtained thanks to Yue He and Thomas Gasser from the International Institute for Applied Systems Analysis - IIASA). In addition, the effect factors were also updated. Previously it was based on data from the World Health Organization from 2003, it is now based on the WHO 2014 report as well as the work of L. Rupcic.

- Climate change, ecosystem quality

Similarly to the human health indicator the cumulative AGTP500 were recalculated. However, the effect factor was not updated yet for this impact category.

- Particulate matter formation

Those CFs were updated to the latest model from Fantke, et al. This is composed of a series of articles on updates to fate and effect factors

This model now provides regionalized characterization factors per town of more than 100,000 inhabitants. The CFs at the town-level are available in the source file, but in the dev file and in the different software versions, we only provide national/regional (e.g., RER) as well as global values, aggregated from the town-level factors.

- Water availability, human health

Those CFs were updated to the latest model of L. Debarre (2024) [publication in review, link will be added once published]. This model includes a harmonization of methodology between the domestic and agriculture water use, updates the exposition factors using the latest Gross National Income data and updates the EF. Overall, the values of the characterization factors of this category have dramatically decreased, by a minimum of 65%.

- Water availability, terrestrial ecosystems

While the original value of the characterization was not updated (e.g., 0.21 PDF.m2.yr in Netherlands), the regionalization was updated based on an estimation of depths of groundwater, based on Jasechko (2021).

- Water scarcity

Those CFs were updated to the latest update of the AWARE model Seitfudem (2024) [publication in review, link will be added once published].

- Fossil and nuclear energy use

The HHV values were updated to match the updated HHVs in ecoinvent.

- Ozone layer depletion

Those CFs were adapted to match the latest data from the World Meteorological Organization (2022). In addition, the time horizon has now been extended to the infinite instead of limiting it to 500 years.

Methodology

For more detail on the methodology behind each impact category, refer to our Github, in the future it will be available directly on our website.

Corrections

In this section we only provide information on the major corrections that were made. For a full report of all the changes please refer to out Github.

- There are challenges associated with using IMPACT World+ files across databases or software for which they were not specifically designed. For instance, this is why we now provide files adapted to particular ecoinvent versions in brightway2. Similarly, both SimaPro and openLCA periodically update the names of their elementary flows. When an impact assessment method has previously been imported into one of these tools, an embedded procedure in their update processes is supposed to adjust the characterization factor names in line with the new flow names, ensuring compatibility with the updated list. However, we lack detailed knowledge of this procedure, meaning we cannot guarantee that the updated versions of IW+ in these tools would align with our specific modeling choices. Likewise, the IW+ versions we provide for a given release of SimaPro or openLCA might be incompatible with previous or subsequent versions due to discrepancies in flow names and characterization factors. Consequently, users of these software should verify which flows are characterized and make adjustments if necessary.

- Harmonization of regionalized flows

Regionalized impact model do not operate at the same geographical granularity. In the previous version, some flows were characterized in one impact category but not in the other. For instance, the flow "Water, lake, US-TRE" was characterized for the "water scarcity" indicator but not for "Water availability, human health". All regionalized flows are now characterized for all the impact categories they affect. This is also true for the newest regionalized impact category (Particulate matter formation) where SO₂ for example is regionalized at a much more granular level than in the freshwater acidification impact category.

- Fossil and nuclear energy use

For the SimaPro version of the v2.0.1, some flows that only exist in SimaPro were not characterized, such as "Oil, crude, 43.4 MJ per kg". Now they are properly characterized using the energy content value specified in the name. Results obtained with SimaPro will thus differ from results obtained with brightway2 and openLCA, since the latter do not use such flows and only use flow such as "Oil, crude".

- Thermally polluted water

In the v2.0.1, the flows "Water, turbine use, unspecified natural origin" were not linked to the correct proxy, which meant they did not impact the Thermally polluted water category, which underestimated the impact of this category.

- The problem of "Nitrogen"

"Nitrogen" can mean two different things. It can literally mean the "N" element, but it can also mean the "N2" molecule. The issue is that it is not clear in the LCI databases, which meaning does "Nitrogen" have. Previously, our understanding was that "Nitrogen" meant "N" but since then, ecoinvent notably, added formulas to the elementary flows they provide and associated the formula "N2" to their "Nitrogen" flows, indicating that they understand it as "N2" and not "N". Furthermore, in SimaPro and openLCA, the associated CAS number is generally "7727-37-9" which again corresponds to "N2". Thus, we now consider that “Nitrogen” represents dinitrogen, and thus does not impact the "Marine eutrophication" impact category anymore. We also corrected a previous mistake: N₂O is not characterized anymore for this impact category.