The varied paths to a more flexible, reliable power system
A clean and expanded power system, dominated by wind and solar, is reliable and resilient to extreme weather events.
Granular modelling reveals that Europe can operate a 95% clean power system by 2035 without compromising reliability and that the weather-dependent, intermittent nature of wind and solar does not pose a threat to the resilience of the grid, even when faced with unfavourable climatic conditions.
Enhancing system flexibility through a varied portfolio of technologies is key to cost-effectively integrating wind and solar, while maintaining the power system’s ability to supply growing demand. As the power supply transforms into one dominated by wind and solar, a parallel system transformation is required to provide for their distinct flexibility needs, and to efficiently integrate new types of power demand. Maximising system flexibility reduces dependence on thermal (gas) capacities for balancing. Enhancing system flexibility ensures that – if adequate wind and solar can be deployed – fossil assets can be phased out without compromising system reliability.
Fully leveraging demand flexibility enables the cost-efficient operation of the future power system. Electrification provides challenges but also opportunities if demand-side flexibility (such as smart charging EVs and flexible heat pumps) and battery storage, including that carried by electric vehicles, can be activated. This is particularly important for the integration of solar power, as shifting demand by a few hours can boost the alignment of demand with daylight hours. These flexibility services also enable peak shaving, a key tool supporting grid resilience and managing the growth of demand peaks.
Three key technologies emerge as the cornerstones of flexibility in a clean power system, maintaining system balance over a range of temporal scales: electrolysers, interconnections, and clean dispatchable generation.
By 2035, wind and solar output frequently exceed demand, at which point electrolysers convert excess supply into green hydrogen. The electrolyser fleet grows to 200-400 GW by 2035 and supplies 14-27Mt of green hydrogen, enough to cover the majority of estimated European domestic demand while maximising the value of renewables output. The REPowerEU plan broadly puts the EU27 on track for this by 2030, aiming for more than 65 GW of electrolyser capacity and 10Mt of hydrogen production. If green hydrogen is instead imported or produced off-grid, it is found that a smaller fleet of ~100 GW by 2035 would still provide sufficient flexibility to the clean power system.
Exchange over interconnectors enables system balancing when mismatch between supply and demand is geographic. The least-cost path for the European grid sees interconnections at least double by 2035 compared to 2020, enabling the cost-efficient expansion of wind and solar capacities by allowing their deployment in countries with the most favourable conditions.
New clean dispatchable power sources enter the system by 2035, but the complete replacement of declining fossil and nuclear capacities is not required. As such, the general trend in all modelled pathways is towards a smaller and cleaner fleet of dispatchable sources by 2035, despite increases in electricity demand (and peak demand). Maintaining the existing hydropower fleet through continued investment and modernisation is strongly recommended. New clean dispatchable capacities can take a variety of forms. Differences in system cost are small, but each technology has a unique risk profile which decision makers must consider.
The wind and solar deployment levels are unaffected by choices between dispatchable capacity options, which have bigger implications for Europe’s dependency on fossil gas. This reinforces that accelerating wind and solar deployment is the central challenge for power sector decarbonisation, as it remains essential across a range of possible system configurations.
Gas with CCS only plays a small role by 2035 in pathways that include it. The role of this technology becomes larger if interconnection expansion is limited, as wind power cannot be as effectively moved across the grid. This would compound two risk factors: the possibility that CCS technology will not reach maturity before 2035, and a prolonged gas dependence. Conversely, the need for gas CCS can be entirely replaced, at minimal additional cost, by a combination of additional solar, earlier deployment of hydrogen turbines, and some additional unabated gas capacity.
Bringing forward investment in clean dispatchable technologies can remove the need for any new unabated gas deployment after 2025. Alternative flexibility options, such as hydrogen turbines, gas with CCS and utility-scale batteries can be used, at minimal additional cost, to build a resilient and clean power system by 2035.
No new nuclear is found to be cost-competitive in modelled pathways, but sensitivity analysis reveals that developing new nuclear according to national plans does not incur significantly higher system costs. Doing so would quicken the transition away from gas in the medium term, and lower long-term reliance on this fuel by providing an alternative form of clean generation to abated gas. These benefits of course need to be weighed against safety risks and the issue of nuclear waste disposal.