Dive Brief:

• Wind projects in Colorado currently dispatch weather data every five minutes to the National Center for Atmospheric Research (NCAR) in Boulder where artificial-intelligence-based software assimilates it with data from weather satellites, and weather stations to provide wind forecasts of unprecedented accuracy that are allowing increased capture of renewable energy at unforeseen low prices.
• The more accurate forecasts allow utilities and grid operators to reduce the impact of wind and solar variability without increasing recourse to fossil reserves, despite increasing penetration of renewables in their generation mixes.
• Forecast accuracy has caused Xcel Energy to reverse its anti-wind stance in Colorado and Xcel has begun to work with NCAR on the more challenging forecast of solar. 

Dive Insight:

NCAR is working to compensate for a lack of data from widely distributed rooftop solar installations by incorporating data from satellites, sky imagers, pollution monitors, and publicly-owned solar into mathematical modeling to project the amount electricity coming from the system’s solar and how that will be affected by weather changes.

Before adopting NCAR’s advanced forecasting methods, Xcel asserted a Colorado policy requiring it to obtain 10% of its power from renewables would increase ratepayers' costs by $1.5 billion over twenty years but it now supports a 30% renewables mandate and says it can handle even more.

Xcel supply and demand calculations include an assessment of real time and historical data that can predict power in 15 minute increments up to seven days ahead.

Xcel and NCAR are now working on how to use forecasts in demand response functions that could include smart appliances, rooftop solar systems, water treatment plants, and electric car charging.