The American-Made Net Load Forecasting Prize is designed to increase adoption of the state-of-the-art in net load forecasting. The net load of an electric grid is the difference between the total electricity demand, and the electricity generation from behind-the-meter resources such as solar and other distributed generators. Having advanced knowledge of the net load allows grid operators to use solar generation and other energy resources more effectively.
Forecasts submitted by competing teams will be compared to benchmark models using the DOE-funded Solar Forecast Arbiter platform. The Prize incentivizes forecast providers to demonstrate the quality and performance of their probabilistic forecasting tools to the larger forecasting industry, while also promoting the adoption of probabilistic forecasts in grid operations.
The Net Load Solar Forecasting Prize builds on the Solar Forecasting Prize, which incentivized solar forecast providers to develop and potentially commercialize tools that generate probabilistic forecasts of solar irradiance, and the Solar Forecasting 2 funding program, which, among other things, developed probabilistic forecasts that enable grid operators to calculate when and how much reserve power is necessary to maintain a healthy grid.
Overview
The American-Made Net Load Forecasting Prize is open to forecasting industry organizations that cater to utilities, system operators, and power plant owners, as well as academic teams with machine learning capabilities that are interested in forecasting. Competitors are required to submit daily day-ahead probabilistic net load forecasts, with an hourly resolution, over the course of four weeks and across multiple distinct climatic locations in the United States. The Solar Forecast Arbiter will be used to evaluate how well each competitor’s forecasts performed compared to a benchmark forecast. By using the Arbiter for comparing forecast performance, the Prize helps demonstrate the feasibility of fair, transparent, and high-quality evaluations of probabilistic net load forecasts using a publicly available open-source platform.
This prize offers up to $600,000 in cash prizes, with three anticipated winners and three anticipated runners-up.
Timeline
- The competition opened on February 7, 2023.
- An informational webinar was held on February 20, 2023.
- Competition registration materials were due March 27, 2023.
- Solar Forecast Arbiter Test Week is June 5-9, 2023.
- The Forecast submission period is June 18-July 16, 2023.
- DOE announced winners on September 12, 2023.
Winners
DOE selected three winners and three runners-up to receive cash prizes in the amounts of $200,000 (first place), $150,000 (second place) and $100,000 (third place). Runners-up will receive $50,000 each.
- First Place: Reliable Autonomy (Basking Ridge, NJ)
- Second Place: Shifted Energy (Honolulu, HI)
- Third Place: Teraton Partners HQ (San Francisco, CA)
- Runner-up: Leaptran (San Antonio, TX)
- Runner-up: Linear Intelligence LLC (Apex, NC)
- Runner-up: Innowatts Team (Houston, TX)
Additional Information
- Download the rules for the competition.
- Join the American-Made Network.
- For questions about the Net Load Forecasting Prize, email [email protected].
- Learn about other American-Made Challenges in the Solar Energy Technologies Office (SETO).
- Learn about SETO's open funding opportunities.
The American-Made Net Load Forecasting Prize is a part of the American-Made Challenges and is administered by the National Renewable Energy Laboratory.