Identification of reliable locations for wind power generation through a global analysis of wind droughts

The increasing reliance on wind power for electricity generation necessitates a comprehensive understanding of wind droughts – prolonged periods of low wind speeds – and their potential impact on electricity system reliability. While wind power contributed to approximately 6.5% of global electricity in 2021, projections indicate a much larger role in the future, potentially exceeding one-third of global electricity needs by 2050. This significant increase in wind power dependency raises concerns about the consequences of wind droughts, which can severely strain electricity systems heavily reliant on this intermittent resource. Recent examples illustrate the disruptive potential of wind droughts: Northwestern Europe's low wind speeds in 2021 exacerbated existing energy challenges; the Western and Southern United States experienced record low windiness in 2015, impacting wind power generation; and India's wind power generation was significantly lower than expected during the 2020 summer monsoon. These events highlight the need for a thorough understanding of the global distribution, severity, and trends of wind droughts to inform effective energy system planning. Traditional intensity-duration-frequency analyses, focusing on individual low wind speed events, may not fully capture the cumulative effect of consecutive wind droughts. An integrated energy deficit metric, accounting for both depth and duration of low-wind periods, is necessary to comprehensively evaluate the impact on energy systems. Furthermore, investigating historical trends in wind droughts is crucial to determine whether recent events are due to climate variability or long-term trends and to improve projections of future wind power potential.

Previous research on wind variability and droughts has primarily utilized intensity-duration-frequency analyses, focusing on the duration of continuous periods below a certain wind speed threshold. Studies using these methods have shown that sustained periods of low wind generation can have return periods of around 10 years. However, these analyses fail to capture the cumulative impact of consecutive wind droughts separated by only brief periods of normal wind speeds. Ruhnau and Qvist (2022) used a storage requirement proxy to analyze the combined energy deficit from consecutive scarce periods of wind and solar generation. This approach offers a more realistic assessment of the impact of prolonged, consecutive periods of low generation from weather-dependent energy sources. Research on climate change impacts on wind resources is ongoing, with some studies suggesting increased energy density in Northern Europe and decreased energy density in the Southeast by the end of the 21st century. However, uncertainties related to global climate model projections and regional downscaling limit the usefulness of these projections for energy system planners. This study aims to address these limitations by employing an integrated energy deficit metric to analyze historical wind drought data and assess potential long-term trends, providing valuable insight for reliable wind-reliant electricity system design.

This study uses 44 years (1979-2022) of hourly wind speed, pressure, and temperature data from the ERA5 reanalysis dataset, produced by the European Centre for Medium-Range Weather Forecasts (ECMWF). ERA5 provides hourly estimates of various climate variables on a 0.25° × 0.25° grid, which translates to approximately 30 km spatial resolution in mid-latitudes. The study focuses on 100-m wind speeds, a height consistent with many modern wind turbine hub heights. From the wind speed data, an hourly wind power density time series is calculated for each grid cell using a formula that incorporates surface pressure and temperature. The study employs an energy deficit metric to quantify wind resource variability and wind droughts across three key aspects: seasonal variability, weather variability, and wind droughts. For each grid cell and year, the energy deficit is calculated by integrating the difference between a target generation profile (constant or climatological mean) and the actual power density time series. The target generation is normalized to have a unit mean value. The seasonal variability is assessed by comparing the actual power density to the climatological mean across all years. The weather variability examines deviations from the climatological mean within each individual year. Wind droughts are evaluated by comparing the actual power density of a given year to the year with the lowest mean power density. The maximum energy deficit over a year is then used to represent the severity of each of these three variables. Linear regressions are used to analyze trends in annual mean power densities, weather variability, and wind drought severity from 1979 to 2022. The analysis focuses on land and coastal areas, excluding Greenland and Antarctica. The study accounts for potential biases and errors in reanalysis data, acknowledging limitations in spatial and temporal resolution, terrain orography effects, and the accuracy of boundary layer parametrization.

The global analysis reveals distinct spatial patterns in wind resources. The American Midwest, Northeastern Canada, Australia, the Sahara, Argentina, parts of Central Asia, and Southern Africa show relatively high mean power densities (250-500 W/m²) and low seasonal and weather variability. In contrast, Northwestern Europe exhibits high power densities but considerably higher seasonal and weather variability, leading to a higher probability of severe wind droughts. The analysis of variability as a function of mean power density indicates that while mean power density varies significantly across land, coastal, and sea areas, the distribution of seasonal variability shows a narrower range, largely between 0.04 and 0.20 (fraction of a year). A composite percentile rank, combining power density, seasonal variability, and weather variability, further identifies the American Midwest, Australia, the Sahara, Argentina, and parts of Central Asia and Southern Africa as regions with the most favorable wind resources. The analysis of wind droughts reveals significant spatial and temporal variability. The most severe historical wind droughts occurred well before substantial wind power penetration, suggesting that historical data can effectively inform system design and planning. Northwestern Europe, the Mediterranean Sea, Northern Canada, and some parts of Northern Russia exhibited the highest probability of experiencing severe wind droughts resulting in more than 400 hours of energy deficit. Regions with the lowest probability include the American Midwest, the Sahara, Argentina, and Australia. An analysis of trends shows a broad increase in wind power density across equatorial regions, while India and Central Europe show decreasing power densities and increasing wind drought severity. Changes in weather variability are less significant, with minimal regional-scale changes over the study period except for some localized trends in the Amazon, North America, and Eastern Africa. The trend in wind drought severity largely mirrors that of wind power density, highlighting the more significant role of power density changes.

The findings demonstrate that certain regions offer abundant and reliable wind resources, characterized by high power densities and low variability. These regions are already being exploited for wind power generation, but the potential of other areas, particularly in Asia and Africa, remains largely untapped. The higher probability of severe wind droughts in Northwestern Europe, despite high power densities, underscores the importance of considering variability when planning wind power infrastructure. The fact that many of the most severe historical wind droughts occurred before widespread wind power deployment highlights the value of historical weather data in designing robust and reliable wind-reliant electricity systems. The relatively small observed trends in wind drought severity compared to the magnitude of historical events suggests that focusing on historical variability rather than solely on long-term projections may be a more effective strategy for designing resilient energy systems. Future research should focus on disentangling the contributions of natural climate variability and human-induced climate change to wind drought trends. Furthermore, investigating the influence of other factors such as land-use change and aerosol pollution is crucial for a more comprehensive understanding of regional variations in wind patterns.

This study provides a valuable global analysis of wind droughts and identifies regions with abundant and reliable wind resources. The findings underscore the importance of considering both power density and variability when planning for wind power integration into electricity systems. The observation that the most severe historical wind droughts predate significant wind power penetration emphasizes the utility of historical weather data in designing robust and resilient energy systems. Future research should focus on improving the understanding of the interplay between climate change, natural variability, and regional factors in shaping wind patterns and drought occurrences. This improved understanding will help optimize energy system planning and design to minimize the impacts of inevitable wind droughts.

The study acknowledges several limitations. The use of reanalysis data introduces potential biases and errors due to factors such as terrain orography, data assimilation limitations, and model resolution. The 0.25° × 0.25° spatial resolution may not fully capture smaller-scale topographical effects. The hourly temporal resolution may not capture more extreme, rapid events like wind gusts. The 44-year dataset may not be sufficient to capture multidecadal wind variability. The analysis focuses on grid-cell level assessment of wind droughts over yearly time periods, potentially overlooking the mitigating effects of spatial aggregation over larger regions or the potential complementarity of wind and solar power generation.