After nearly 9 years, the Grid Optimization (GO) Competition is ending. PNNL began supporting the Competition soon after Tim Heidel's 2014 Advanced Bulk Power System Optimization Technologies Workshop, with funding beginning September 1, 2015. The first two years were devoted to the design and deployment of the three successful Challenges while the GRID DATA program worked on producing data. With data in hand the Challenges began.
Richard O'Neill, National Academy of Engineering member and former Chief Economist at FERC, who became the GO Competition ARPA-E Program Director on November 25, 2019, retired today after 4 years of dedicated service to the Competition. The idea for the Competition emerged out of O'Neill's ideas put forth at the FERC Software Conferences.
O’Neill initiated the series of annual FERC Technical Conference on Increasing Market and Planning Efficiency through Improved Software (FERC Software Conference) held at the FERC building in Washington, DC, in 2008, but there seems to be no on-line records until 2010, when there were 8 talks about AC Optimal Power Flow. Many of the people eventually involved, either as competitors or members of the support team, participated as early as 2011, when O’Neill and collaborators began a series of talks on Computational Approaches to the AC Optimal Power Flow (OPF) Problem. Other presentations in 2011 by Competition related individuals included Rothberg (Gurobi) and Bienstock (Columbia). The 2012 Conference had 7 Competition presenters and 2013 had 6. 2013 saw the release of 11 Optimal Power Flow and Formulation Papers by O’Neill et al. and discussions of how much of the $400 billion in energy sector revenues could be saved with improved optimization algorithms. O’Neill, noting that the transition from Lagrangians to Unit Commitment in the early 2000s saved approximately $30 billion, estimated better optimization algorithsm could save $10 billion. Tim Heidel told Steve Elbert he didn’t think it was that much. The 2014 Conference saw 6 presentations by Competition people with two more by O’Neill collaborators. Heidel told Elbert at this meeting that $10 billion probably was reasonable and on October 6, 2014 sent out invitations to the Advanced Bulk Power System Optimization Technologies Workshop in Arlington, VA, November 13-14, 2014. Heidel said he was motivated by discussions with O’Neill about the improved impact of optimization on the efficiency of the electric grid and by queries from Congressional staffers about why ARPA-E was not holding competitions like DARPA. The goal of the Competition would be to improve the state-of-the-art optimizers for Alternating Current Optimal Power Flow (ACOPF) and in general increase the efficiency of the North American Power Grid. Slide 12 from Heidel’s presentation indicates $11.5-$23 billion in savings per year are possible from faster, more robust OPF.
5. Physics-Informed Machine Learning for Electricity Markets, Robert Ferrando¹, Laurent Pagnier², Robert Mieth³, Zhirui Liang⁴, Yury Dvorkin⁵, Daniel Bienstock⁶, Michael Chertkov¹, ¹University of Arizona, Tucson, AZ, ²University of Arizona, Tucson, AZ, ³Princeton University, Princeton, NJ, ⁴Johns Hopkins University, Baltimore, MD, ⁵Johns Hopkins University, Baltimore, MD, ⁶Columbia University, New York, NY
---
Session SK02, Keynote: The Electric Grid in Evolution: Data, Optimization, and Risk-Taking, Sunday, October 15, 5:25 PM – 6:15 PM, CC-West 301ABC, Phoenix Convention Center, Phoenix, AZ
Daniel Bienstock, Columbia University, New York, NY (slides)
Session WA51, PERFORMS: Integrating Non-Traditional Resources and Reducing Asset Delivery Risk in Power Systems, Wednesday, October 18 8:00 AM – 9:15 AM, CC-North 229B, Phoenix Convention Center, Phoenix, AZ
Session WB51, PERFORMS: New Approaches to Risk Management in Power Markets, Wednesday, October 18, 9:30 AM – 10:45 AM, CC-North 229B, Phoenix Convention Center, Phoenix, AZ
A few of the people with a GO Competition connection after the Sunday morning presentations.
Dan Bienstock, Jim Ostrowski, Andy Sun, Carleton Coffrin, Dirk Lauinger, Mathew Brun, Robert Parker, Farnaz Safdarian, Hassan Hijazi, Constance Crozier, Sam Chevalier, Jesse Holzer, Hans Mittelmann
Results from Event 4 were posted today. Submissions were received from 13 teams: Artelys_Columbia, Electric-Stampede (University of Texas at Austin), Gatorgar (University of Florida), GOT-BSI-OPF (Global Optimization Technology, Inc.), GravityX (individual), LLGoMax (Lawrence Livermore National Laboratory), Occam's razor (individual), PACE (IncSys), PGWOpt (University of Pittsburgh), quasiGrad (individual), The Blackouts (University of Tennessee - Knoxville), TIM-GO (Massachusetts Institute of Technology), and YongOptimization (Mississippi State University).
The following GO Competition related talks were presented at the 2023 ARPA-E Virtual Grid Software Annual Review
GravityX, presented by Hassan Hijazi, Los Alamos National Laboratory; Presentation
Electric-Stampede’s Approach: Fast and Robust Strategies for Large-scale mixed-integer SCOPF, presented by Javad Mohammadi, UT Austin and Hussein Sharadga, University of Texas - Austin
Fast AC Security-Constrained Optimal Power Flow with Unit Commitment on High Performance Computing, presented by Yong Fu, Mississippi State University; Presentation
TIM-GO, presented by Andy Sun, Massachusetts Institute of Technology; Presentation
The following talks were presented at the 2023 FERC Software Conference by individuals with a GO Competition connection:
Simulation of Wholesale Electricity Markets with Capacity Expansion and Production Cost Models to Understand Feedback between Short-Term Market Procedures and Long-Term Investment Incentives, presented by Jesse Holzer
Integration of DER Aggregations in ISO-Scale SCUC Models, presented by Brent Eldridge; Presentation
Energy Storage Participation Algorithm Competition (ESPA-Comp), presented by Brent Eldridge; Presentation
Mine Production Scheduling under Time-of-Use Power Rates with Renewable Energy Sources, presented by Daniel Bienstock;Presentation
Results from Event 3 were posted today. Submissions were received from 13 teams: Argonauts (Argonne National Laboratory), Artelys_Columbia, Electric-Stampede (University of Texas at Austin), Gatorgar (University of Florida), GOT-BSI-OPF (Global Optimization Technology, Inc.), GravityX (individual), LLGoMax (Lawrence Livermore National Laboratory), PACE (IncSys), PGWOpt (University of Pittsburgh), quasiGrad (individual), The Blackouts (University of Tennessee - Knoxville), TIM-GO (Massachusetts Institute of Technology), and YongOptimization (Mississippi State University).
Results from Event 2 were posted today. Submissions were received from 12 teams: Argonauts (Argonne National Laboratory), Artelys_Columbia, Electric-Stampede (University of Texas at Austin), Gatorgar (University of Florida), GOT-BSI-OPF (Global Optimization Technology, Inc.), LLGoMax (Lawrence Livermore National Laboratory), PACE (IncSys), PGWOpt (University of Pittsburgh), quasiGrad (individual), The Blackouts (University of Tennessee - Knoxville), TIM-GO (Massachusetts Institute of Technology), and YongOptimization (Mississippi State University).
Results from Event 1 were posted today. Submissions were received from 11 teams: Argonauts (Argonne National Laboratory), Artelys_Columbia, Electric-Stampede (University of Texas at Austin), Gatorgar (University of Florida), GOT-BSI-OPF (Global Optimization Technology, Inc.), LLGoMax (Lawrence Livermore National Laboratory), PACE (IncSys), PGWOpt (University of Pittsburgh), The Blackouts (University of Tennessee - Knoxville), TIM-GO (Massachusetts Institute of Technology), and YongOptimization (Mississippi State University).
In a press release, the Department of Energy (DOE) announced that at least13 teams from 11 states will compete in the Grid Optimization (GO) Competition Challenge 3, which requires them to develop software management solutions to address real-world power grid conditions. The GO Competition—managed by the Advanced Research Projects Agency-Energy (ARPA-E) and Pacific Northwest National Laboratory (PNNL) with support from Arizona State University, Georgia Tech, Los Alamos National Laboratory, the National Renewable Energy Laboratory, Texas A&M University, the University of Texas at Austin, and the University of Wisconsin-Madison—echoes the Biden Administration’s commitment to create a more reliable, resilient, and secure American electricity grid.
Access to the Challenge 3 Evaluation Platform is now available via a Sandbox submission. A Sandbox submission will run an entrants solver against selected Sandbox datasets using the Evaluation Platform. Entrants are strongly encouraged to make use of this capability to test the correctness of their runtime parameters before an Event. Datasets currently available include S0 and, soon, S1.
The Sandbox will remain Open for Submissions throughout Challenge 3 except when and Event is taking place.
Session MB61, October 17, 2022, 11:00 AM - 12:15 PM, Room Marriott 8 (second floor), Indianapolis Marriott Downtown, 350 West Maryland Street, Indianapolis, IN
Session ChairRichard Paul O'Neill, ARPA-E, Silver Spring, MD
The S0 Sandbox dataset (50.3 MB) was released for download. It contains of 5 networks consisting of 3-, 14-, 37-, 73- and 6049-buses. There are 27 datasets, 9 for each division. Each dataset consists of a JSON file and a TXT file. The JSON file contains the data and the TXT file contains summary information about the dataset.
The Grid Optimization Competition Challenge 3Problem Formulation has been released. ARPA-E has issued a Press Release and a Funding Opportunity Announcement (DE-FOA-0002690), Support Grants For Participation in ARPA-E Grid Optimization (GO) Competition Challenge 3. The Full Application Submission Deadline is 4/5/2022 9:30 AM ET.
Participation in the GO Competition Challenge 3 will be open to anyone that satisfies the applicable requirements in the Rules Document, not just those awarded under ARPA-E DE-FOA-0002690.
Richard (Dick) O'Neill, Advanced Research Projects Agency-Energy (ARPA-E) Distinguished Senior Fellow and GO Competition Program Director, has been elected to the National Academy of Engineering for "leadership and contributions to the liberalization and design of efficient natural gas and electric power markets."
Submissions for the Challenge 2 Monarch of the Mountain (Ch2-MoM) opened January 3, 2022, for registered teams and will remain open until October 31, 2022. See Solution Uploading for details. The Ch2-MoM Leaderboard will be updated every Wednesday.
The Grid Optimization (GO) Competition Post Challenge 2 Event named Challenge 2: Monarch of the Mountain or Ch2-MoM, will focus on finding improved solutions to the security-constrained optimal power flow (SCOPF) problem introduced in the Challenge 2 Event.
The 2021 FERC Technical Conference on Increasing Market and Planning Efficiency through Improved Software took place virtually June 22-24, 2021. Talks of interest included
HIPPO Concurrent Optimizer – Enhanced Simultaneous Feasibility Test (SFT) and Migration to Private Cloud, presented by Jesse Holzer; Presentation
Pricing in Dynamic ISO Markets with Unit Commitment, presented by Richard O'Neill; Presentation
Efficient Prices under Uncertainty and Non-Convexity, presented by Brent Eldridge; Presentation
Challenge 2 Trial 1 Event Leaderboard results have been posted and the synthetic datasets published.
The set consits of 11 synthetic networks composed of 63 scenarios. The 4 industry datasets, composed of 20 scenarios, will not be made publicly available.
The Timeline document was updated. Teams wishing to participate in Challenge 2 Trial 1 must have their Registration Documents approved by ARPA-E by December 1, 2020.
Twelve network models comprising 34 scenario datasets are now available for testing in the Challenge 2 Sandbox that is open to Challenge 2 registered Teams.
Challenge 1 teams must reregister for Challenge 2.
Stehen Elbert and Hans Mittelmann presented a talk entitled Analysis of GO Competition Challenge 1 Final Event Problem Difficulty at the 2020 FERC Technical Conference regarding Increasing Market and Planning Efficiency and Enhancing Resilience through Improved Software.
A meeting was held at the Renaissance New Orleans Arts Warehouse District Hotel to introduce the winners of the GO Competition Challenge 1. The agenda included talks by Lane Genatowski, Director of ARPA-E, Key Industry Speakers, and a Panel discussion with selected entrant teams: LLNL, Lehigh, Georgia Tech and CU-Boulder.
U.S. Energy Secretary Dan Brouillette announced that the first round of winners has been named in the Department of Energy’s Grid Optimization (GO) Competition. The GO Competition, managed by DOE’s Advanced Research Projects Agency-Energy (ARPA‑E) is a series of challenges to develop software management solutions for challenging power grid problems. The competition’s intent is to create a more reliable, resilient and secure American electricity grid. The winners will share a total of $3.4 million, which is to be used to further develop their respective approaches and pursue industry adoption of their technologies.
Richard O'Neill took over as the ARPA-E Program Director of the Grid Optimization (GO) Competition today. He served as the Chief Economic Advisor at the Federal Energy Regulatory Commission (FERC) before coming to ARPA-E.
The GO Competition Platform will lock-down its software environment on September 2, 2019 and will remain locked for the duration of the Challenge 1 competition. Competition teams should submit any requests for changes before that date to the GO Competition Administrator.
The GO Competition will now host a third trial event designed to give teams more opportunities to work with large datasets similar to those from Trial Event 2. Trial Event 3 will test a smaller number of scenarios than Trial Event 2 and is optional, but all teams are encouraged to participate.
GO Competition related talks at the 2019 FERC Technical Conference regarding Increasing Real-Time and Day-Ahead Market Efficiency and Planning Efficiency and Enhancing Resilience through Improved Software held at the FERC Building in Washington, DC.
Fast Evaluation of Security Constraints in a Security Constrained Unit Commitment Algorithm, presented by Jesse Holzer; Presentation
HIPPO: A Concurrent Optimizer for Solving Day-ahead Security Constrained Unit Commitment Problem, presented by Feng Pan; Presentation
The ARPA-E Grid Optimization (GO) Competition: Challenge 1 and Beyond, presented by Kory Hedman; Presentation
Performance-based Energy Resource Feedback, Optimization, and Risk Management: PERFORM, presented by Kory Hedman; Presentation
In order to maximize the development time for the competitors, with open access to the sandbox for testing, we will shorten the submission window from fourteen days to three days and slightly reschedule the Trial 2 Event to the end of the work week. The new Trial 2 submission window will be from July 17th – July 19th.
Challenge 1 - Trial Event 1 is open from April 1 00:00 PDT until April 15 23:59 PDT; no prizes are associated with this event. Both the Sandbox and other Challenge 1 datasets will remain open and available during this time.
Each team is allowed only one submission for this event. Teams can submit to Trial Event 1 by submitting a Challenge 1 submission and selecting the Trial Event 1 dataset. Good luck!
A webinar with the topic "File Formatting and Solution Evaluation " was held at 1:00 pm ET. The webinar recording is available on YouTube. The slides, questions, and answers will be available soon.
