Quantifying the Dunkelflaute -- An analysis of variable renewable energy droughts in Europe

TL;DR

The paper addresses Dunkelflaute in Europe by quantifying VRE droughts across wind, solar, and portfolios using 38 years of data and a multi-threshold drought-mass approach. It combines the VMBT drought identification with a DIETER-based storage optimization across island, copperplate, and distributed interconnection scenarios to connect drought patterns to long-duration storage needs. Findings show that portfolio and geographical balancing significantly reduce drought severity and duration, with the most extreme events occurring as sequences of shorter droughts; the winter 1996/97 European super drought under perfect interconnection lasted 55 days, even as average portfolio availability remained substantial. The study argues for weather-year ensembles and a two-step planning framework, warning against single-year analyses and advocating planning horizons that capture cross-year seasonality to support weather-resilient, fully renewable energy systems.

Abstract

Variable renewable energy droughts, also called "Dunkelflaute", emerge as a challenge for climate-neutral energy systems based on variable renewables. Here we characterize European drought events for on- and offshore wind power, solar photovoltaics, and renewable technology portfolios, using 38 historic weather years and an advanced identification method. Their characteristics heavily depend on the chosen drought threshold, questioning the usefulness of single-threshold analyses. Applying a multi-threshold framework, we quantify how the complementarity of wind and solar power temporally and spatially alleviates drought frequency, return periods, duration, and severity within (portfolio effect) and across countries (balancing effect). We identify the most extreme droughts, which drive major discharging periods of long-duration storage in a fully renewable European energy system, based on a policy-relevant decarbonization scenario. Such events comprise sequences of shorter droughts of varying severity. The most extreme event occurred in winter 1996/97 and lasted 55 days in an idealized, perfectly interconnected setting. The average renewable availability during this period was still 47% of its long-run mean. System planners must consider such events when planning for storage and other flexibility technologies. Methodologically, we conclude that using arbitrary single calendar years is not suitable for modeling weather-resilient energy scenarios.

Figures (24)

• Figure 1: Drought patterns in 1996 and 1997 for selected regions. Identified drought patterns in 1996 and 1997 for all investigated relative thresholds $\tau_i$ with $i \in [0.1,..., 1]$ and selected regions. The dates shown on the bottom axis correspond to all technology-region-specific panels. For each panel, the stacked horizontal bands indicate drought occurrences for the color-coded threshold of a particular region. To illustrate persistent patterns, only droughts lasting longer than one day are displayed. Figure \ref{['fig:figure_SI2']} focuses on longer-lasting events by illustrating respective patterns for droughts lasting at least one week.
• Figure 2: Example of frequency-duration distributions of drought events. Cumulative frequency-duration distributions of drought events across all investigated relative drought thresholds $\tau_i$ with $i \in [0.1,..., 1]$, sorting the yearly frequencies of all events that are at least as long as a given duration. White space indicates the absence of droughts for given thresholds in the data. The contour lines represent the threshold-specific yearly frequency. For illustration, the distributions are truncated at 360 hours, i.e., they show events with a maximum duration of just above two weeks. Yearly frequencies of events lasting at least two days, one week, and a fortnight are marked for a relative threshold of $\tau_{0.75}$. An interactive version of this Figure which allows zooming and rotating is available online kittel_high_2025.
• Figure 3: Example of return period curves of extreme drought events. Return period curves of extreme drought events with an average annual frequency below 1 across selected relative thresholds $\tau_i$. Figure \ref{['fig:figure_SI10']} provides a more comprehensive perspective on return periods across all thresholds.
• Figure 4: Maximum duration of single drought events. Most extreme duration of single drought events across all investigated years for given relative thresholds $\tau_i$ with $i \in [0.1,..., 1]$. The year with the most extreme event duration varies across thresholds. The marked numbers indicate the longest drought durations found for a relative threshold of $\tau_{0.75}$.
• Figure 5: Example of maximum duration of single drought events across months. Most extreme duration of single drought events across all months in the data and all investigated thresholds $\tau_i$ with $i \in [0.1,..., 1]$. The contour lines represent the threshold-month-specific maximum duration. Note that the monthly assignment illustrates the median hour of respective droughts, while the duration of each event is plotted on the vertical axis. Events lasting longer than one month start (end) in previous (subsequent) months. The events with the highest and lowest duration across all months are marked for a threshold $\tau_{0.75}$.