DOE, EPRI, NREL Advance DER Innovation With Net Load Forecasting Prize

The U.S. Department of Energy’s (DOE) American-Made Net Load Forecasting Prize is open to competitors. The latest American-Made program challenge will encourage competition between non-profit organizations, companies, and institutions to develop advanced solutions for cultivating distributed energy resources (DER) into America’s evolving energy system infrastructure. U.S. DOE Secretary Jennifer Granholm announced the competition launch at Distributech International 2023 in San Diego this week. Competitors can enter to compete for the Net Load Forecasting Prize through the HeroX platform.

 

The U.S. DOE is encouraging competition between innovators from across the nation to develop key enabling technologies for the clean energy revolution. Image used courtesy of Pixabay

 

Renewable Energy Integration  

The U.S. government has ambitious goals to reduce carbon emissions and become carbon-neutral by 2050. To meet these goals, the U.S. needs to update its existing energy system infrastructure and replace fossil-fuel-derived energy with renewable-based energy supplies.

DERs are various small-scale, modular, and often renewable energy sources that generate electricity. They use grid-edge hardware in the form of physical elements such as solar photovoltaic cells, wind turbines, battery storage systems, and other technologies. Locating DERs next to areas of electricity consumption can reduce transmission losses and increase the reliability of the power supply in the grid network. DERs are also more efficient than traditional power plants as they take up less space and require less maintenance.

Additionally, they can provide a more reliable energy source since they do not rely on a single source for their power supply. By integrating DERs into the existing energy system infrastructure, the U.S. can reduce its dependency on fossil fuels and move towards cleaner energy sources.

 

Net Load Forecasting

The challenges facing the integration of grid-edge technologies include establishing appropriate physical and cybersecurity, robust control schemes that support grid resiliency, and more accurate net load predictions.

A microgrid is a subset of DERs connected to the main power grid or disconnected from the grid to provide power locally. Net load forecasting is a key enabling technology for integrating microgrid systems with the main power grid.

Net load forecasting is predicting future electricity demand by analyzing historical data. It uses mathematical models, statistical techniques, and machine learning algorithms to estimate the electricity needed in a particular region or area at any given time. Net load forecasting helps energy providers plan for future demand, allowing them to better prepare for peak usage periods and manage their resources accordingly. This planning and management extend to using DERs, which can be balanced with using the main power grid to meet these peak usage periods.

The Director of the DOE Solar Energy Technologies Office, Becca Jones-Albertus, believes that improving the accuracy of net load predictions can help strengthen grid reliability and security. Additionally, improving predictions allows for cost-effective planning, energy generation, and storage dispatch. 

 

Net Load Forecasting Prize

The DOE-funded America-Made program incentivizes the next generation of innovators from across the U.S. to develop new technologies to advance the clean energy transition. 

The National Renewable Energy Laboratory (NREL) and the Electric Power Research Institute (EPRI) administer the Net Load Forecasting Prize. Collectively, prize competitors can win up to $600,000, comprising three cash prizes for winners and another three prizes for runners-up. 

 

The Net Load Forecasting Prize offers competitors a chance to win up to $600,000, split into different cash prizes for winners and runner-ups. Image used courtesy of Pixabay

Net Load Forecasting Prize competitors will use their skills and knowledge to develop probabilistic models for predicting net load amounts each day before the forecast. Competitors will also advance the adoption of probabilistic forecasts and associated evaluation tools.

Competitors will use EPRI’s open-source Solar Forecast Arbiter tool to access historical load data for forecast model training. Across 28 days, competitors will submit their model results for a minimum of four locations in the U.S. to the Solar Forecast Arbiter platform. Models will be assessed for forecasting performance and compared against a benchmark forecast.