Sun and cloud forecasts provide benefits for solar energy production

How much energy can a solar energy plant produce? One factor that has an impact is the weather. In a research project, SMHI is developing forecasts for solar radiation and cloud cover, for use by solar energy producers both in Sweden and abroad.

For four years, researchers from SMHI have been working in the European DNICast project to develop solar radiation forecasts and forecasts for cloud cover. The focus has been on conditions around the Mediterranean, where the forecasts can be used to plan production at large thermal solar power plants. SHMI’s forecasting model has been evaluated and developed during this work.

“The challenge when it comes to solar radiation forecasts lies in getting the model to describe cloud cover. During the project we’ve been working to get the clouds in the model to correspond with what the clouds look like from the satellite. We’re now going to continue this work for Nordic conditions,” says Tomas Landelius, researcher at SMHI.

Special forecasts for Sweden

In a new project financed by the Swedish Energy Agency, SMHI will be developing solar forecasts to control and monitor the electricity system in Sweden.

“In the European project, we looked mainly at extremely short-term forecasts for a few hours ahead, but for planning in Swedish energy companies we need to develop forecasts for the next 1-1.5 days. The forecast can then help the energy companies to assess how much solar electricity they can produce and how much electricity they then need from other forms of energy,” explains Tomas Landelius.

Installation of solar panels on roof
In a research project, SMHI is developing forecasts for solar radiation and cloud cover, for use by solar energy producers both in Sweden and abroad.

Collaboration and evaluation

During the project, the solar forecasts will be evaluated against the measurements that the municipal energy company Tekniska verken i Linköping has from its solar electricity production. SMHI can also produce probability information for the forecast to help with the assessment of how reliable the forecast is.