IEO2023 WEPS Source Code and Input Files

The World Energy Projection System (WEPS) is the model we use to create our International Energy Outlook (IEO). You can find more information about WEPS in the IEO overview and full documentation.

The WEPS download allows others to replicate the International Energy Outlook 2023 (IEO2023) results and discloses the source code and inputs we used. Users can also construct and analyze alternative cases, although doing so can require significant technical expertise and time.

Download WEPS ZIP

WEPS2023 now includes the option to create a package of files that can be submitted to a job queue suitable for processing on a cloud cluster. Although EIA uses the celery task manager with a RabbitMQ message broker, users can use whatever job queue they think works best.

Python

• We designed the WEPS archive for use as a 64-bit Python 3.10 application with Microsoft Windows. Installation files and additional Python libraries are listed below.

Installers

• Python 3.10.6 (https://www.python.org/ftp/python/3.10.6/python-3.10.6-amd64.exe)

Libraries

• numpy-1.22.4+mkl-cp310-cp310-win_amd64.whl
• numexpr-2.8.3-cp310-cp310-win_amd64.whl
• openpyxl-3.0.10-py2.py3-none-any.whl
• pandas-1.4.3-cp310-cp310-win_amd64.whl
• pyodbc-4.0.32-cp310-cp310-win_amd64.whl
• tables-3.7.0-cp310-cp310-win_amd64.whl
• tabulate-0.8.10-py3-none-any.whl
• xlrd-2.0.1-py2.py3-none-any.whl
• xlwt-1.3.0-py2.py3-none-any.whl
• XlsxWriter-3.0.3-py3-none-any.whl
• scipy-1.8.1-cp310-cp310-win_amd64.whl
• Bottleneck-1.3.4-cp310-cp310-win_amd64.whl
• docutils-0.19-py3-none-any.whl
• Jinja2-3.1.2-py3-none-any.whl
• scikit_learn-1.1.1-cp310-cp310-win_amd64.whl
• patsy-0.5.2-py2.py3-none-any.whl
• statsmodels-0.13.2-cp310-cp310-win_amd64.whl
• pytest-7.1.2-py3-none-any.whl
• lxml-4.9.0-cp310-cp310-win_amd64.whl
• pylint-2.14.5-py3-none-any.whl
• pwlf-2.2.1.tar.gz -- Optional
• celery-5.2.7-py3-none-any.whl
• flower-1.2.0-py2.py3-none-any.whl
• gevent-22.10.2-cp310-cp310-win_amd64.whl
• python-pptx-0.6.21.tar.gz
• Sphinx-5.1.1-py3-none-any.whl
• matplotlib-3.5.2-cp310-cp310-win_amd64.whl
• line_profiler-3.5.1-cp310-cp310-win_amd64.whl

FORTRAN

• Recompiling the WEPS Transportation module, written in Fortran, requires f2py (included with numpy) and mingw64.

GAMS

• We use GAMS for the International Electricity Market Module. For IEO2023, we used GAMS version 36.2.0 with the Xpress solver version 8.10. More information is available on the GAMS website.

AIMMS

• We use AIMMS for the International Coal Market Module and the World Hydrocarbon Activity Module. For IEO2023, we used AIMMS_Installation_Free_Releases/4.69.1.0-x64-VS2017 with the Cplex solver. More information is available on the AIMMS website.

Oxford Economics Model

• If you request files for the proprietary Oxford Economics Macroeconomic Model, we will send the files separately after you license the model from Oxford Economics. Contact the IEO team for details.

World Hydrocarbon Activity Model (WHAM)

• EIA uses 3 data sources with paid licenses for the World Hydrocarbon Activity Module (WHAM) contained within WEPS. The posted version of WEPS replaces these with proxy data based on publicly available information. As a result, the posted version of WEPS will not match IEO2023 results. Reproducing IEO2023 results will require licenses to the following data sources:

Bloomberg

• Purchase the Bloomberg Terminal and include commodity prices.

Rystad

• Purchase the Rystad UCube.

Oil & Gas Journal

• Purchase the 2022 Worldwide Refinery Survey.

Once EIA receives confirmation that these data sources have been purchased, we will provide the input files required to replicate IEO2023 results.

Email the IEO team with technical questions about the WEPS archive.