Thanks to @kevinbookdc for the conversation + @JosephMajkut @CSISEnergy for providing an excellent forum to highlight the decisive acceleration underway at @EIAgov

This afternoon at 3pm Eastern, I look forward to a candid conversation @CSIS with @EIAgov Administrator Tristan Abbey @EIA_One about the future of the vital agency he is leading at a critical time. Please join us: https://t.co/KQUtm4lRtp

Thanks to our hardworking team, the backup system we prepared last night (see prior post) successfully published the WNGSR without incident earlier today.

Today EIA failed to publish WPSR on time. We are investigating the causes of this incident, which appears to have been technical in nature, as the report itself was ready for publication. We are preparing a backup system to support tomorrow’s scheduled publication of WNGSR.

Looking forward to a frank and substantive conversation about an important agency with a vital mission.

Tune in this Thurs for a conversation with @kevinbookdc and @CSISEnergy on the future of @EIAgov.

Our dedicated workforce has many talents, but one shared mission: to collect, publish, and analyze energy data for the American people. It is an honor to lead this team.

Good afternoon! I'm Tristan Abbey, new Administrator of @EIAgov, nominated by @POTUS in February and sworn in by @SecretaryWright in September. The timeline before this post (12/01/2025) will remain for historical archiving of previous administrators.

This account is inactive while we await the arrival of our next Administrator. Please direct any questions to @EIAgov.

To address emerging energy issues, @EIAgov needs a next generation model that is modular, flexible, transparent, and robust. We are committed to a transparent model development process. If you’re a modeler, please consider providing feedback on this early stage prototype.

If you have a chance to experiment with the tool, let us know how we can improve it and expand the data coverage.

Caveats: 1. This is a python-based Dash app that runs on your local machine. We don't have a live dashboard on the website. Maybe someday. 2. We're pulling past projections from our API, which only goes back to 2015. 3. Only key market indicators are currently included.

It's important for modelers to examine the accuracy of their past projections. Accurate past performance is no guarantee for the future, as structural elements of the system can change rapidly. But we can still learn a lot.

In an effort to increase transparency, we have developed a python-based benchmarking tool to compare our AEO projections against real world outcomes. It includes all past cases and allows you to visualize the pathways. It's available in GitHub: https://t.co/Rl863L7E9C

@EIAgov already produces an AEO Retrospective: https://t.co/ER2DhptI64 However, it largely consists of statistical estimates and only makes use of the Reference case.

When we release the AEO, there's always plenty of discussion about the quality of our projections. Those discussions should be informed by a rigorous look at our our past performance.

It's important to condition model insights on a careful consideration of future uncertainty. There are many ways to do this, but @EIAgov considers uncertainty through a set of cases in the AEO, each with differing assumptions about the future.

Producing high quality long-term energy projections is hard. Uncertainty looms large, and there are inevitable surprises that our models do not capture.

For all the long-term modelers out there: we just released an open source benchmarking tool that allows you to plot projections from past editions of the Annual Energy Outlook (AEO) against actual outcomes. Let me explain why this is important. 🧵

Update: battery electric vehicle sales represented 7.1% of the U.S. light duty vehicle market in 2Q24, about the same as last year. Today in Energy (8/26/24) 👇 https://t.co/ILRzF8mSZm