Forecasting Solar Output: Why It Is Becoming Increasingly Important


By Peter Kelly-Detwiler - Storyteller in Residence

If you’ve been paying attention to the Texas energy scene, you know a solar storm is brewing. Tens of gigawatts of utility-scale solar projects are being lined up in the ERCOT interconnection queue, a development that will soon affect market prices and grid operations. At the same time, a smaller, but rapidly growing increase of distributed solar is coming online as well, which will affect both ERCOT and the state’s distribution utilities. Pretty soon, it is quite likely that California’s infamous ‘duck curve’ may have a Texas counterpart, already know to some as the ‘dead armadillo curve.’

And that’s a good thing. However, the challenge is that while solar generation is clean and increasingly cost-effective, it’s not dispatchable. The ‘fuel’ (in the form of sunlight, or solar radiation) shows up when the sun rises, disappears when the sun sets, and can be spotty whenever clouds appear. As solar energy becomes an increasing element of both ERCOT and other grids across the globe, it becomes increasingly important to accurately forecast that irradiance, so that we can integrate and optimize the corresponding amounts of solar energy.

On June 13, Husch Blackwell Partner (and TREIA board member) Chris Reeder hosted TREIA’s third webinar of 2019, a presentation from Dr. Nick Engerer of Solcast, to discuss the how and why of state-of-the-art solar forecasting

Engerer, Chief Technology Officer of Solcast, noted that the challenge of solar forecasting will be critically important here in Texas, where “massive amounts of solar will significantly change the ERCOT landscape.”

Cloudy forecasts must be accurate

Projecting solar energy output is, he said, “a meteorology problem, a cloud forecasting problem.” In a macro sense, overall output is affected by latitude and time of year, both of which are predictable. The real challenge is to optimize, plan for, and integrate the daily local solar resource that is affected by clouds. That requires accurately forecasting where the clouds are, how thick they are, and where they are going, and to “predict that from minutes to hours to days.”

Clouds can have a highly profound impact on overall energy output. Engerer commented that different types of clouds possess different levels of opacity. High opacity clouds, such as thunderheads, for example “can push solar production down to near zero.”

With accurate forecasts in hand, owners can develop bidding strategies in Day-Ahead and Real-Time markets, and grid operators can determine how to dispatch other generation to complement the solar resources at a local and granular level.