Debate on electricity storage tends to centre on batteries. Yet the need emerging from a heavily renewable system is not primarily short-duration storage, but very long-duration, high-capacity storage. On that segment the gas transmission network holds a structural advantage.
75 TWh of surplus by 2050
Our hourly modelling, based on the French energy agency’s 2050 scenario, 60 GW of solar, 70 GW of wind, 20 GW of nuclear and consumption reduced to around 380 TWh, produces generation surpluses of close to 75 TWh a year, some 15 % of current French output, excluding local congestion.
Those surpluses occur across 5,000 to 6,000 hours a year. In counterpart the system faces around 27 TWh of shortfall over 3,000 to 4,000 hours, which means flexible capacity remains necessary at peak.
The need is for long storage, not short
This is the decisive point. In the short term, intermittency calls for intraday storage. In the long term, surplus periods lengthen: several consecutive days, sometimes whole weeks.
Our characterisation of surplus periods shows that around 80 % of the volume comes from periods of 12 consecutive hours or more. Episodes lasting three days to a week account for 26 TWh on their own.
That structure rules out most storage options. Electrochemical batteries reach at best a few tens of MWh. Pumped hydro caps out at a time constant of 30 to 40 hours, and additional French potential is limited, around 4.3 GW for 80 GWh. Adiabatic compressed air storage was still at development stage.
Within that frame, hydrogen is the only route capable of storing large volumes of energy over periods from several days to several months. It also carries an advantage neither pumped hydro nor compressed air has: electrolysis depends on no particular geography, and can therefore be distributed across the country to address local congestion as well.
Grid injection, the best use alongside transport fuel
Three routes existed for this must-run, distributed hydrogen: storage for material or transport fuel use, storage for later reconversion to power, or injection into the natural gas grid.
Our technical and economic analysis ranks grid injection highest, on a par with transport fuel. The reason is straightforward: it builds on transport infrastructure and storage capacity of considerable scale, operational today, and so avoids costly investment in dedicated pure hydrogen storage.
Injection and transport fuel are complementary rather than competing. Vehicle fleet consumption is steady, whereas must-run hydrogen production is inherently irregular. The grid absorbs the peaks mobility cannot take.
An order of magnitude equivalent to two EPRs
In the scenario modelled, electrolysis could absorb around 25 TWh a year of surplus, leading to close to 20 TWh of hydrogen injected into the gas grid, the energy equivalent of two EPR reactors and around 7 % of gas network consumption at that horizon.
That figure depends directly on the assumptions used: a less ambitious scenario on renewables and efficiency would produce a much smaller need, possibly none. Even so, several countries were already anticipating the shift, with E.ON’s electrolyser connected to the Ontras network at Falkenhagen and the UK GridGas project launched by ITM Power.
To go further, download our full analysis below or contact our experts.
The downloadable PDF of this study is available in French only.

