Notebook to estimate end of life volumes by plastic polymer per region and year combination.
This notebook approximates the polymer-level volumes reaching end of life in each region within the Global Plastics AI Policy Tool by distributing waste in each fate proportionally to the size of each sector at the time of waste generation. Having attributed waste by sector, a static matrix converts to polymers which are then summed across sectors.
This is provided for convenience to the community but is not part of the "main tool" (see tool repository) or its publications.
The easiest way to use this notebook is to simply download its output: polymer_eol_approximate.csv. This artifact is regenerated by CI / CD through this repository. The amount column is in million metric tonnes.
For those looking to execute this locally, simply install python requirements with pip install -r requirements.txt. Users may also consider creating a virtual environment.
The notebook can be run by executing jupyter notebook while in this repository's directory and navigating to the PolymerEndLifeApproximation notebook.
No explicit development standards are enforced at this time.
Merging to the main branch will cause the artifact to be rebuilt and deployed to production where the file can be downloaded as polymer_eol_approximate.csv.
There are two important caveats for this software.
Users may notice that the region naming convention has changed from prior releases of our tool. We provide these values as Majority World (MW) intead of Rest of World (ROW) and North America (NA) instead of NAFTA. These labels encompass the same parts of the world regardless of labeling convention though we encourage use of MW and NA. Alternative data exports with legacy names can be found in the tool itself.
Note that this method introduces a slight approximation. Due to lifecycle distributions, the sector ratios at time of waste generation may not be the same as if one were to project through time to get the prior sector ratios which generated the waste in question. However, this use of "end of life year ratios" provides a sufficient estimation for many use cases in which approximation is acceptable. Alternatively, one may generate more precise values by simulating the lifecycle distributions and summing volumes per-polymer while projecting forward future waste. For more details on this, see the pipeline behind the Global Plastics AI Policy Tool.
This code is released under the BSD 3-Clause License. See LICENSE.md for more details. Note that the data themselves are subject to CC-BY-NC.