Skip to Main Content U.S. Department of Energy
Energy and Environment Directorate
Page 180 of 934

Research Highlights

Highlights Archive

Perfecting Power Projections

New tool increases the accuracy of power forecasts

August 2015
Perfecting Power Projections
PNNL's Power Grid Integrator has demonstrated up to a 50 percent improvement in forecasting future electricity needs over several commonly used tools. Project lead Luke Gosink, right, consults on the use of the new tool, which could save millions in wasted electricity costs.
Enlarge Image

Accurately forecasting future electricity needs is tricky, with sudden weather changes and other variables impacting projections minute by minute. Errors can have grave repercussions, from blackouts to high market costs. Now, a new forecasting tool delivers up to a 50 percent increase in accuracy and has the potential to save millions in wasted energy costs.

The Challenge

Grid coordinators have the daily challenge of forecasting the need for and scheduling exchanges of power to and from a number of neighboring entities. The sum of these future transactions, called the net interchange schedule, is submitted and committed to in advance.

Until now, forecasters relied on a combination of personal experience, historical data, and often a preferred forecasting model. Researchers at PNNL theorized that they could develop a method to guide the selection of an ensemble of models with the ideal, collective set of attributes in response to what was occurring on the grid at any given moment.

Strength in Numbers

The resulting Power Model Integrator tool, developed with funding from PNNL’s Future Power Grid Initiative, has the ability to adaptively combine the strengths of different forecasting models continuously and in real time. This addresses a variety of scenarios that impact electricity use, from peak periods during the day to seasonal swings. Performance of the tool was tested against five commonly used forecasting models processing a year's worth of historical power system data.

"For forecasts one-to-four hours out, we saw a 30-55 percent reduction in errors," said Luke Gosink, project lead and PNNL staff scientist. "It was with longer-term forecasts—the most difficult to accurately make—where we found the tool actually performed best."

The advancement is featured as a best conference paper in the power system modeling and simulation session at the IEEE Power & Energy Society General Meeting in Denver. For more information, read the PNNL press release.

Page 180 of 934

Energy and Environment

Core Research Areas