Rooftop solar has reached a level where it provides a substantial amount of renewable energy to the Danish electricity grid (up to 10% at peak production), but not all data from homeowners’ solar installations are submitted to grid operators.
A new project is looking into the potential of exchanging solar data between the Danish Meteorological Institute (DMI) and the grid operator Energinet (ENDK), with the purpose of better predicting power loads and thus planning ahead in the intermittent nature of a nation wide grid made up from evermore renewable energy sources.
Last year wind power contributed a total of 44% of all electricity production in Denmark, which in itself is a huge challenge for a national grid. Yet, the Danish grid is considered one of the most stable and resilient in the world with an average uptime of 99.99%. This is helped in part by high-capacity connections to neighboring countries Norway, Sweden, and Germany.
But what about solar? Is it just a matter of years before solar will reach the levels of wind in terms of nameplate capacity? The planned cancellation of import controls on solar panels and cells from China in September by the EU certainly suggests that this could happen sooner rather than later.
Last winter a trial was setup to look into the data that DMI and ENDK can provide. DMI has weather stations spread out all over the country that measure all the weather parameters you could think of, including sunshine. ENDK gets data from a few thousand rooftop installations which it uses to calculate an estimate of current electricity production from all installations combined. Turns out there is a robust correlation between the data, thus spurring the project to find out more.
Through digital analysis of a large number of historical data measured per municipality from both DMI and ENDK, it appeared that the potential of mutual advantage was present.
Lasse Borup from ENDK found that when they compared historical data, there was a high correlation between the observations made by DMI stations and the production that was measured from solar installations by ENDK. It showed that there could be a basis for all to benefit from cooperating further by exchanging data.
Ulrik Ankjær Borch, meteorologist at DMI, confirms that the preliminary results seem to open up new possibilities for better measurements of solar radiation in the country.
They both see how this could develop into a more permanent solution, where key advantages would be things like better prediction of solar electricity production and external control of each instance of measured data. In other words: connecting their individual systems would result in a more reliable unified system to ensure a robust grid.
I asked Ulrik Ankjær Borch how the project was coming along, and whether data from wind power would be used as well:
DMI and ENDK continue to investigate the possibilities of exploiting each other’s data to everyone’s advantage ie. ENDK production data and our weather data. The DMI’s forecast data is compared against ENDK’s data, but so far still only from solar. We have long-term cooperation with ENDK who daily (in their operation of the electricity grid) use our data from both solar and wind, so yes, wind is also part of our cooperation. And yes, my expectation is that we will continue to look into how DMI and ENDK can benefit from each other’s data and develop our businesses accordingly for the benefit of society.
Indeed, the data from wind and solar are inherently different in nature, and thus very different in how they are managed. Both of these renewable energy technologies have one single purpose which is to provide electrons to whoever needs them at the exact time they need them, and if managed correctly the intermittent nature of these technologies will eventually be invisible to the end-user.
I reached out to ENDK for a comment on the project, and press officer Jesper Nørskov Rasmussen had this to say about the status on their side:
We have not yet initiated an operational delivery between DMI and ENDK, but there is still goodwill to continue the cooperation.
DMI and ENDK have previously also looked at the possibility of exchanging wind data. However, due to data confidentiality in the production of wind turbine owners, etc., ENDK has not had the opportunity to contribute in this area. However, it is ENDK’s vision to make the most possible data available. We already have an Energy Data Service, where anyone can retrieve data about the energy system.
We are constantly working to release even more data through this portal.
Data is the key. No doubt about that. Ulrik Ankjær Borch from DMI elaborated a bit more on their status:
Focus right now is on how DMI can benefit from production data from ENDK. One advantage is better observations of solar data (check of abnormalities in DMI measurement data), while another is looking at whether production data can be used directly in the projection of radiation — as an alternative or support for DMI observations. The latter is something we are already working on.
The same thoughts can apply to wind data, where the completion of DMI’s model forecasts can be combined with ENDK’s needs even more than today. Which could result in DMI’s model not only predicting the wind, but also the power output directly from the wind turbines.
One of the differences between solar and wind that are not discussed as much as the “when the wind doesn’t blow” and “when the sun doesn’t shine” problems, is the fact that solar cells are electronics and wind is mechanics and thus solar can be switched on and off in microseconds as opposed to wind that rely on adjusting the angle of blades and changing gears in turbines.
Wind is essential for a stable grid today, but I think a future will emerge where low-cost solar will dominate completely, and the chemical nature of batteries will probably help marginalize wind over time. We will see, but in any case, cooperation between meteorological instances and utility instances will probably be essential for the overall system efficiency in the rapidly developing global renewable energy transition.