Skip to main content

DOE announces competition for the best net load forecasting tool

The Net Load Forecasting Prize will be awarded to the team that can create the best probabilistic forecast that may later be used to create a tool for grid operators.
Will $600,000 in prize money be enough to spur engineers to create a probabilistic forecasting tool for better grid management? DOE hopes so.

As the energy transition gets further underway, grid operators and electric utilities need new tools to help them reliably deliver energy to the homes and businesses they serve. But as more distributed energy resources (DER) like solar PV and energy storage are added to the distribution grid, forecasting net load — the combination of the building’s load minus any solar generation it is using — becomes more difficult.

On February 7, during the DISTRIBUTECH International keynote address, US Energy Secretary Jennifer Granholm announced a new prize sponsored by DOE’s Solar Energy Technologies Office that she hopes will help grid operators understand how much generation is coming from DER. The Net Load Forecasting Prize will be awarded to the team that can create the best probabilistic forecast that may later be used to create a tool for grid operators.

“Providing an actual probability distribution is, we think, the way to deal with quantities that are variable and uncertain, like electricity demand and generation from weather dependent resources like solar,” said Becca Jones-Albertus, director of the U.S. Department of Energy (DOE) Solar Energy Technologies Office (SETO), in an interview. She wants grid operators to be able to use behind-the-meter solar PV generation more effectively.

“Dealing in those probabilities instead of worst-case scenarios can enable us to more effectively get more out of the solar resources that we have on the grid,” she added.

Jones-Albertus knows that using probabilistic forecasts is a shift for system operators. “So we’re hoping the prize can showcase how probabilistic forecasts can be very effective and help give grid operators and utilities more comfort using them.”

DOE has developed an open-source Solar Forecast Arbiter (under the stewardship of the Electric Power Research Institute (EPRI)), which Jones-Albertus said allows for the “objective and automated evaluation of forecasts and their quality.” This means that potential net load forecast developers can use the solar forecast arbiter to evaluate the forecast models they are developing. In addition, grid operators can use the arbiter to “evaluate forecasts that come to them,” said Jones-Albertus.


Subscribe today to the all-new Factor This! podcast from Renewable Energy World. This podcast is designed specifically for the solar industry and is available wherever you get your podcasts.


In all, six teams will be awarded with the winner receiving $200,000, the runner up receiving $150,000, and the prize money declines from there. A total of $600,000 will be awarded.

Jones-Albertus said DOE wanted to create a prize competition to spur the creation of net-load forecasting tools which are critical to being able to add more solar to the grid.  She envisions the traditional forecasting companies to enter the contest but said it is open to all.

“We like the opportunity to appeal to a broad audience and get folks, not just companies that are already developing forecasting tools, but maybe university teams or others, to get excited about the idea of a prize and it brings them into the space,” she said.

Representatives from EPRI and the National Renewable Energy Lab (NREL) are attending the DISTRIBUTECH event and are available to answer questions or provide more information in the EPRI booth (#5216).

In addition, there is an informational webinar scheduled for February 20 at 3PM eastern time. Register for the webinar here.

More information about the prize here.

Contest entrants need to register  by March 27 and will be able to test their forecasting software against a DOE-provided dataset in mid-April. The actual competition will begin in late April and last 28 days. Winners of the Net Load Forecasting Prize will be announced in June.

Data & News supplied by www.cloudquote.io
Stock quotes supplied by Barchart
Quotes delayed at least 20 minutes.
By accessing this page, you agree to the following
Privacy Policy and Terms and Conditions.