Grid headroom without new grid: Europe's hydro hybridisation finding

The finding

Ember's July 2026 analysis of hybrid power projects reports that Austria, Bulgaria, France, Italy, Portugal, Romania and Spain could add around 25 GW of new wind and solar without any new grid infrastructure. The capacity connects behind grid connection points that existing hydropower plants already hold, a practice Ember calls hybridisation. The seven countries account for 93 GW of the European Union's 131 GW hydropower fleet, so the connection points to share are already in place.

The constraint this addresses is not generation cost. Ember notes that a large volume of planned renewable capacity across Europe is held up waiting for grid connections rather than for a business case. Sharing an existing connection sidesteps that queue for the portion of the build that can physically co-locate.

Why sharing a connection works

A hydropower plant rarely exports at its full connection capacity around the clock, so the connection sits partly idle. Adding solar or wind behind the same point fills more of that headroom without exceeding the export limit the connection was built for. Ember measures the effect as a capacity factor gain: reservoir and pumped hydro plants move from an average 19 per cent to 36 per cent when solar is added, and run of river plants from 29 per cent to 47 per cent. In some locations, Ember reports, the addition could more than triple utilisation.

Solar pairs more productively than wind in both plant types, because its midday peak lands in the hours a reservoir is most often holding water back. The connection carries more energy over a year, and the grid beyond it does not have to be enlarged to take it.

What the 25 GW represents

Set against the roughly 138 GW of wind and solar the seven countries expect to deploy between 2026 and 2030, the 25 GW is about 18 per cent of the planned build that could connect without waiting on new grid. It is a share of a real pipeline, not a theoretical ceiling for hydropower co-location.

A working example: Pracana

The mechanism is already running. In January 2026 the Portuguese utility EDP commissioned what it describes as Portugal's first onshore hydro and solar hybrid, at its Pracana site on the Ocreza river. A 48 MW solar array of about 90,000 panels was built behind the connection of the existing 41 MW Pracana hydropower plant, which has operated since 1951. The two share one grid connection, for a combined 89 MW, and no new transmission capacity was added to accommodate the solar.

EDP reports the combined complex generates about 140 GWh a year, enough for roughly 51,800 households. The solar contributes the larger share of that energy from a plant that added nothing to the grid it connects to.

The transferable principle

The specific numbers are European, but the mechanism is grid agnostic. Wherever new generation is waiting on connection capacity rather than on cost, an existing connection that runs below its export limit for part of the day is spare capacity that new solar or wind can occupy. Hydropower is the clearest case because reservoirs hold water, and therefore connection headroom, through the sunniest hours. The same logic extends to any partly used connection point.

Australian grids carry the same kind of constraint, with connection queues and hosting-capacity limits shaping where new capacity can land in both the National Electricity Market and Western Australia's South West Interconnected System. This report does not put an Australian figure on the opportunity, which would need its own connection-level analysis, but the Ember finding is a reference point for the modelling: the cheapest new connection is often one that already exists. It sits alongside our reading of where storage is being added behind existing meters, in Global Behind-the-Meter Storage: The 2025 Data.