Challenge 1 solved a security-constrained (AC) optimal power flow (SCOPF) problem. Algorithms were tested on complex, realistic power system models, and Entrants were scored on how well their algorithms perform relative to other Entrants’ algorithms.

Ten teams were awarded $3.4 million.


Challenge 2 expanded upon the SCOPF problem posed in Challenge 1 by adding price-responsive demand, ramp rate constrained generators and loads, fast-start unit commitment, adjustable transformer tap ratios, phase shifting transformers, and switchable shunts.

Nine teams were awarded $2.4 million.


This Event focuses on finding improved solutions to the security-constrained optimal power flow (SCOPF) problem introduced in Challenge 2. Teams generated solutions on their own with no software or hardware constraints and no time limits. Solution Files were uploaded to PNNL for solution evaluatipon.

This event kicked off on January 3, 2022 and concluded October 31, 2022. This competition rewarded the teams that found the better solutions first as well as teams that found solutions with >1% improvement over Challenge 2 results.

Two teams were awarded $440,000.

See the C2-MoM Leaderboard for results.


Challenge 3 focused on multiperiod dynamic markets including advisory models for extreme weather events, day-ahead markets, and the real-time markets with an extended look-ahead. These problems will included active bid-in demand and topology optimization.

Challenge 3 FOA released February 16, 2022.

Challenge 3 Event 1 (no prize money) submission window was January 25-27, 2023.

Challenge 3 Event 2 (no prize money) submission window was April 13-14, 2023.

Challenge 3 Event 3 (prize money) submission window was June 15-16, 2023.

Challenge 3 Event 4 (prize money) submission window was August 31-September 4, 2023.

Eight teams were awarded $3.0 million.

See the Challenge 3 Leaderboards for results.


The sandbox is a testing environment that allows Entrants to try the submission, evaluation, and scoring process using datasets that may be small for quick turnaround or debugging as well as the actual Trial Event datasets from previous Trials for algorithm development.