Long-duration electricity storage needs for coping with Dunkelflaute events in Europe

This study analyzes Europe's power sector over 35 historical weather years to conclude that coping with extreme "Dunkelflaute" events in a decarbonized system requires a massive expansion of long-duration electricity storage, amounting to approximately 7% of annual demand, as geographical balancing and fossil-based backups offer limited mitigation.

The Gist

Imagine Europe's electricity grid as a massive, high-tech kitchen trying to cook a meal for millions of people using only two ingredients: wind and sun. The problem is, these ingredients are unpredictable. Sometimes the wind stops blowing, and the sun hides behind clouds for days or even weeks at a time. In German, this scary period of "no wind and no sun" is called a Dunkelflaute (literally "dark doldrums").

This paper is like a chef's manual asking: "If the wind and sun go on a long vacation during the coldest part of winter, how much food do we need to keep in the pantry to keep everyone fed?"

Here is the breakdown of their findings using simple analogies:

1. The "Pantry" Problem (Long-Duration Storage)

The researchers found that short-term batteries (like the ones in your phone or a regular car) are like a small lunchbox. They are great for a quick snack or a power outage lasting a few hours. But a Dunkelflaute is like a blizzard that lasts for weeks. You can't survive a blizzard on a lunchbox; you need a giant pantry.

The paper argues that to survive the worst possible "dark doldrums" in Europe, we need to build a massive long-duration storage system.

• The Size: They calculated that Europe needs a pantry capable of holding 351 terawatt-hours (TWh) of energy.
• The Analogy: That is enough energy to power the entire continent for about 7 days straight, or roughly 7% of all the electricity Europe uses in a whole year.
• The Technology: The most promising "pantry" they looked at is hydrogen. Think of it as turning extra electricity into a gas (hydrogen), storing it in giant underground caves (like natural gas storage), and turning it back into electricity when the sun and wind are gone.

2. The "Group Trip" Analogy (Geographical Balancing)

One way to solve the problem is to share. If it's cloudy in Germany, maybe it's sunny in Spain. If the wind dies in the UK, maybe it's blowing in Norway.

• The Analogy: Imagine a group of friends on a road trip. If one person runs out of gas, they can borrow from a friend who has extra.
• The Finding: The paper shows that connecting countries with stronger power lines and hydrogen pipes helps. It's like having a bigger group of friends to share with.
• The Catch: Even if Europe is perfectly connected (like a single giant super-grid), the "pantry" is still needed. Why? Because sometimes the entire continent gets hit by the same bad weather at the same time. The "group trip" helps, but it can't fix a storm that covers the whole map.

3. The "Backup Generator" (Nuclear and Fossil Fuels)

The researchers also asked: "What if we just keep a nuclear power plant or a gas generator running as a backup instead of building a giant pantry?"

• Nuclear: They found that nuclear power is like a reliable, steady friend who always brings a sandwich. It helps reduce the size of the pantry needed, but it doesn't eliminate the need for it. Even with a lot of nuclear power, you still need a massive pantry for the worst storms.
• Fossil Fuels + Carbon Capture: They looked at using oil generators that suck carbon out of the air to stay "green." They found this is only a good idea if the technology becomes incredibly cheap (almost free). If it costs normal amounts, it's not worth the money compared to just building the hydrogen pantry.

4. The "Weather Lottery" (Why 1996 Matters)

The researchers didn't just look at one year; they looked at 35 years of weather data.

• The Finding: They discovered that the year 1996/1997 was the "worst-case scenario." It was a winter where almost the whole of Europe had a massive, simultaneous drought of wind and sun.
• The Lesson: If you build your system based on a "normal" year, you will fail when the "1996-style" storm hits. You have to build your pantry big enough to survive the worst year in history, not just an average one.

The Bottom Line

To keep Europe's lights on in a world without fossil fuels, we cannot rely on the wind and sun alone. We need to build a massive, continent-sized pantry (specifically using hydrogen storage) to survive the rare but terrifying weeks when the weather turns against us.

• Geographical sharing (connecting countries) helps shrink the pantry a little bit.
• Nuclear power helps shrink it a bit more, but not enough to remove the need for it.
• The pantry is non-negotiable. Without it, the lights go out during a Dunkelflaute.

The paper concludes that policymakers need to start building this "pantry" now, because it takes a long time to construct, and we can't wait for the next big storm to realize we were unprepared.

Technical Summary

Technical Summary: Long-duration electricity storage needs for coping with Dunkelflaute events in Europe

Problem Statement

The European Union aims to achieve net-zero greenhouse gas emissions by 2050, relying heavily on the massive expansion of variable renewable energy (VRE) sources, specifically wind and solar power. While these technologies promise declining costs and vast expansion potential, their integration exposes the power sector to significant weather variability. A critical challenge is the occurrence of prolonged periods of low wind and solar availability, known as "Dunkelflaute" or VRE droughts. These events, characterized by simultaneous low generation across large geographic areas and often coinciding with high winter demand, threaten the security of supply in a decarbonized system.

Previous research has established that increasing system flexibility is essential for managing VRE imbalances. However, there is a lack of distinct analysis regarding how extreme, pan-European renewable droughts specifically impact the sizing and operation of long-duration electricity storage (LDS). Furthermore, existing energy system models often rely on single or limited weather years, potentially leading to suboptimal capacity choices that fail to account for the inter-annual variability of extreme drought events. This study addresses the gap in understanding the specific LDS requirements needed to safeguard a fully renewable European power sector against the most severe historical drought events.

Methodology

The authors combine two open-source methods to analyze the interaction between VRE droughts and optimal storage investments across 33 European countries over 35 historical weather years (1982–2016).

1. VRE Drought Identification (VREDA): The study utilizes the Variable Renewable Energy Drought Analyzer (VREDA) to identify drought patterns based on availability time series for onshore/offshore wind and solar PV. The tool employs the Variable-duration Mean-Below-Threshold (VMBT) method, which iteratively searches for periods where renewable availability falls below specific thresholds. This approach aggregates drought patterns across multiple thresholds to compute a "drought mass" metric, identifying the most extreme events. The analysis covers both isolated "energy island" scenarios and a pan-European "copperplate" scenario (perfect interconnection).
2. Power Sector Modeling (DIETER): The Dispatch and Investment Evaluation Tool with Endogenous Renewables (DIETER) is used to determine least-cost capacity expansion and dispatch decisions. The model optimizes over contiguous hours of a full year (summer-to-summer planning horizon) under perfect foresight. It includes a simplified transport model for cross-border electricity exchange and a detailed module for hydrogen-based long-duration storage (electrolysis, underground storage, and reconversion via hydrogen turbines).

Scenarios: The analysis compares four interconnection scenarios:

• (1) Energy islands (no cross-border exchange).
• (2) Policy-oriented electricity exchange (TYNDP 2022 levels), no hydrogen exchange.
• (3) Policy-oriented exchange of both electricity and hydrogen (TYNDP 2022 levels).
• (4) Unconstrained pan-European copperplate (perfect integration).

Additional sensitivity analyses explore the roles of nuclear power (low and high capacity scenarios), fossil backup with Direct Air Carbon Capture and Storage (DACCS), and variations in technology investment costs.

Key Contributions and Results

1. Extreme Droughts Define Storage Needs

The study finds that the operation and investment sizing of long-duration storage are primarily defined by the most extreme renewable drought events. These events, which can last several weeks to months, often occur in winter and coincide with peak electricity demand. The analysis identifies the winter of 1996/97 as the period with the most extreme pan-European drought, affecting many countries simultaneously.

2. Quantified Storage Requirements

• Island Scenario: Without cross-border exchange, the median long-duration storage capacity required across all modeled countries is 232 TWh, with a maximum of 378 TWh (approx. 7.6% of annual European demand) in the worst year.
• Policy-Relevant Scenarios: Under TYNDP 2022 interconnection levels (Scenario 3), the maximum storage capacity required to cope with the 1996/97 event is 351 TWh, corresponding to 7.1% of yearly European electricity demand.
• Theoretical Minimum: Even in a perfectly interconnected "copperplate" scenario (Scenario 4) where geographical balancing is unconstrained, a minimum of 159 TWh (3.2% of annual demand) remains necessary to handle the 1996/97 event. This represents the "no-regret" lower bound for storage in a fully renewable system.

3. Impact of Geographical Balancing

Geographical balancing via interconnection significantly reduces storage needs by smoothing regional variations in renewable generation. However, the study demonstrates that this effect is limited during extreme, synoptic-scale droughts where low availability affects many countries simultaneously. While policy-relevant interconnection (Scenarios 2 and 3) reduces storage needs compared to isolated islands, the reduction is modest for the most extreme events. Unconstrained balancing (Scenario 4) yields the lowest storage requirements, but the authors note that such levels of interconnection are not currently feasible.

4. Interaction with Other Flexibility Options

• Nuclear Power: The inclusion of nuclear power reduces storage needs, but the effect is limited. Low levels of nuclear power (24 GW) reduce median storage needs by only 4–8%. Even with high levels of nuclear power (102 GW), substantial long-duration storage investments remain optimal in the least-cost solution (reductions of 26–37% depending on the scenario). Nuclear power displaces VRE capacity but does not eliminate the need for storage to cover residual demand during extreme droughts.
• Fossil Backup with DACCS: Sensitivity analysis shows that deploying oil-fired backup capacity with Direct Air Carbon Capture and Storage (DACCS) only reduces system costs and storage needs if DACCS costs are extremely low (100 EUR/t CO2). At more realistic cost estimates, the substitution effect is marginal, and storage energy capacity remains largely unaffected.
• Short-duration vs. Long-duration: The model reveals complex interactions where short-duration batteries and pumped hydro support long-duration storage. For instance, batteries may cycle to maximize electrolyzer utilization during charging periods, and short-duration storage handles brief renewable surpluses within longer drought events to minimize round-trip losses.

5. Sensitivity to Weather Year Selection

The study highlights that the choice of weather year significantly impacts optimal storage sizing. The 1996/97 year requires storage capacities up to 47% higher than the upper bound of capacities required for the years typically used in current planning tools (e.g., TYNDP 2022 uses 1995, 2008, 2009). This underscores the risk of under-sizing infrastructure if models rely on a limited set of weather years that do not include the most extreme drought events.

Significance and Claims

The paper claims that transitioning to a renewable European energy system requires significant long-duration storage capacities, likely in the order of several hundred TWh, to ensure reliability during extreme "Dunkelflaute" events. The authors argue that:

• Storage is Indispensable: Long-duration storage is a critical technology for coping with pronounced renewable droughts in systems with significant seasonal demand and supply variability.
• Interconnection Limits: While cross-border interconnection helps, it cannot fully substitute for storage during extreme, widespread droughts. Policymakers must prepare for rapid expansion of LDS even with ambitious grid expansion plans.
• Planning Horizon: Current energy system models that rely on single or few weather years may lead to suboptimal capacity choices. The authors advocate for modeling approaches that incorporate multiple weather years, specifically including the most extreme drought events, to identify weather-resilient system configurations.
• Refining Definitions: The authors propose refining the definition of "Dunkelflaute" to focus on extended periods (weeks to months) where renewable generation falls short of demand, rather than short-term fluctuations, as these extended periods are the primary drivers of long-duration storage needs.

The study concludes that while firm zero-emission technologies like nuclear can partially mitigate storage needs, they do not eliminate the requirement for large-scale long-duration storage. Consequently, system planners and policymakers should prioritize early action to enable the rapid scaling of hydrogen-based storage and other LDS technologies to safeguard the renewable energy transition.