The solar eclipse of 12 August 2026 reduced the power delivered by photovoltaics in Europe by more than 9 GW, very close to the operators’ scenario assuming a drop to 9.7 GW. Despite such a rapid change, system frequency remained stable around 50 Hz. It is a good example of how forecasting, coordination and flexibility help manage a high share of PV.
What exactly happened to PV output during the eclipse?
For skywatchers it was an exceptional summer evening. For power system operators it was a planned test of how a grid with a high share of photovoltaics reacts when solar irradiance drops rapidly across many countries at once. ENTSO-E had earlier indicated the window of 19:15-21:30 CEST as the period of greatest impact on the European system. Under a clear sky, the drop in PV output was forecast to reach as much as 9.7 GW.
After the event, ENTSO-E reported that the maximum drop in power delivered by photovoltaics exceeded 9 GW across Europe. An important distinction here: 9 GW of installed capacity did not “disappear.” It was a temporary drop in current output as the Moon limited irradiance, followed by a change in output as the eclipse moved on and sunset approached.
The scale of over 9 GW is impressive, but it must be read in a broader context. According to ENTSO-E, solar energy accounted for about 13% of electricity generation in the EU in 2025. Short-lived, multi-gigawatt changes in PV output are therefore no longer an abstract scenario for the future. They are part of the operation of a system in which weather-dependent sources have a growing share.
Why did a drop of over 9 GW not destabilise the grid?
The simplest answer is: because the event was predictable. An eclipse is rare, but from the point of view of system planning it has a big advantage. Its timing can be calculated with very high accuracy. The uncertainty concerns mainly the weather and the actual PV output just before the eclipse, not the timing of the event itself.
Even before 12 August, ENTSO-E described specific preparations. Operators in control centres were informed of the situation in advance, market participants received data on the expected drop in solar output, and local operators additionally received updated PV forecasts from weather-data providers. During the eclipse, planned outages of grid elements were also limited to keep greater operational flexibility.
In practice this meant that a spectacular astronomical phenomenon was reduced to a well-known balancing task. The power grid does not require every source to generate continuously. It does require a constant balance between generation and demand. After the event, ENTSO-E reported that frequency remained stable with no significant deviation from 50 Hz.
Why was the time of day almost as important as the degree of the eclipse?
The percentage of the solar disc obscured does not translate directly into the same percentage of energy lost over the day. What matters for the system is the timing of the phenomenon, the height of the Sun above the horizon, the current PV capacity, the weather, the demand profile and which other sources are available at the time.
In 2026 the eclipse reached much of Western Europe in the evening. PV output was already naturally starting to fall before sunset. A Solcast analysis carried out after the event illustrates the importance of this effect well — locations with a very deep eclipse did not always lose the most energy over the whole day, because the phenomenon occurred late there. Reykjavik was a good example: the additional loss of the daily solar resource fell below 1% due to cloud cover.
What should installers, EPCs and investors take away from this event?
At the level of an individual project, the eclipse creates no new requirement for the mounting structure or a separate rule for switchgear design. The conclusion is broader: a PV plant operates within a system that is becoming more dynamic. A good design should therefore cover the whole architecture of generation and the electrical side, and not end at the sum of module power.
The mounting structure, cable routing, switchgear, inverters and monitoring are not independent items on a shopping list. They must form a single system that can be safely installed, predictably operated and serviced for years. The eclipse of 12 August was simply a continental example of the same engineering principle: first understand the operating conditions, then design the right level of flexibility for them.