How Can We Anticipate the Impact of Climate Change on Wind Energy Production?
- 1 day ago
- 6 min read
Wind energy is one of the pillars of the global energy transition. Governments, investors and utilities increasingly rely on wind power for greenhouse gases emissions reductions and energy security. But as surprising as it may seem, wind farm design, like many large scale industrial and infrastructure projects, still operate under an assumption of climate stability: future production is evaluated using past weather records and assuming they are representative of the climate of the next decades.
Evolving wind patterns, rising temperatures and shifting pressure systems raise an important question for the wind industry: as the climate is changing, can we still assume that future wind conditions will resemble those observed in the past?
Over the last fifteen years, researchers have published dozens of studies attempting to anticipate the impact of climate change on wind energy production resources and electricity generation. Their conclusions vary significantly depending on regions, climate models and methodologies but generally indicate a declining trend.
Understanding these methodologies is essential for interpreting the results correctly. Not all approaches measure the same thing, and some provide a much more realistic representation of future wind farm performance than others. This post offer a review of the main methodologies used in scientific litterature and industrial studies.
From Wind Speed to Energy Production: Approaches to Anticipate the Impact of Climate Change
Researchers typically use one of three families of indicators to evaluate the future evolution of wind energy resources under climate change:
Wind speed (WS)
Wind power density (WPD)
Actual energy production (AEP)
Each approach represents a different level of complexity and realism.
Wind Speed: The Simplest Indicator
The most straightforward method consists of analyzing projected changes in wind speed.
Climate models directly simulate wind speed at various atmospheric levels, making this variable relatively easy toi access for future projections. Researchers typically analyze changes in average wind speed or specific percentiles at either standard climate model heights or extrapolated turbine hub heights.
This approach offers several advantages. Wind speed is easy to calculate, easy to compare across regions, and directly available from climate projections. As a result, many of the earliest studies on climate change and wind energy relied primarily on this indicator.
However, wind speed alone provides only a partial view of future energy production: a turbine does not convert wind speed into electricity in a linear way. A 10% decrease in wind speed does not necessarily translate into a 10% reduction in production. Depending on turbine characteristics and local wind distributions, the actual impact can be significantly much larger, smaller or even non-existant.
Wind Power Density: Accounting for the Physics of Wind Energy
To better represent the energy available in the atmosphere, many studies use wind power density.
Wind power density measures the kinetic energy contained in moving air per unit area. Unlike wind speed, it incorporates the physical relationship between air movement and energy content.
Wind power density varies with the air densisty and the cube of wind speed.
This relationship means that relatively small changes in wind speed can produce much larger changes in available wind energy. A 10% increase in wind speed can increase kinetic energy by roughly 30%, while a similar decrease can produce a comparable reduction.
However, wind power density remains a resource indicator rather than a production indicator.
A wind turbine only capture a fraction of the available kinetic energy. According to Betz's Law, the theoretical maximum extraction efficiency is approximately 59%. Modern utility scale turbines typically convert around 40% of available wind power into electricity under ideal conditions.
Moreover, many studies simplify calculations by assuming constant air density. This assumption ignores one of the most certain and global consequences of climate change: rising temperatures.
As warmer air is less dense than colder air, even if wind speeds remain unchanged, increasing temperatures reduce the energy contained in the air mass. In usual conditions, a temperature increase of approximately 3°C reduce air density and wind energy density by around 1%.
Although seemingly small, this effect is expected to occur almost everywhere and therefore introduces a systematic downward influence on future wind resources.
Actual Energy Production Models: Bringing Turbines into the Equation
The most advanced studies attempt to estimate actual energy production rather than resource availability.
These approaches usually rely on simplified turbine power curves to convert projected wind conditions into electricity generation. Power curves describe how a turbine responds to different wind speeds.
Below the cut in speed, typically around 3 to 4 m/s, turbines produce no electricity. Production then increases as wind speed rises until rated power is reached. Above the cut out speed, generally around 25 m/s, turbines shut down to avoid mechanical damage.
Some simplified approaches estimate only the frequency of wind speeds falling within the operational range of turbines. More sophisticated methods use full turbine power curves to calculate expected production at each time step.
In both cases, these approaches provide a substantially more realistic estimate of climate change impacts because they account for the physical constraints of wind turbine operation.
The Remaining Challenges
The evolution of methodologies over the last decade reflects a broader maturation of the field. A comprehensive review published by Sara Pryor and colleagues in 2020 identified 29 studies investigating future wind resources under climate change. Only five used actual energy production and 11 used wind speed.
An updated review conducted by Callendar covering 14 additional studies published since 2020 reveals a clear methodological shift. Studies relying solely on mean wind speed projections have almost disappeared. Researchers increasingly favor wind power density or and energy production based approaches.
This evolution reflects growing recognition that simple wind speed indicators are insufficient for evaluating future wind energy performance. But despite considerable progress, significant limitations remain in the current state of the art.
Most Studies Ignore Actual Wind Farm Locations
Many studies assess future wind resources on climate model grids or across large regions. While useful for understanding broad trends, this approach often lacks operational relevance.
Wind energy assets are not evenly distributed across the landscape. They are concentrated in areas with exceptional wind resources such as the North Sea, Inner Mongolia or the Great Plains. Changes affecting these specific production zones may matter far more than global or regional averages.
Energy Production Modeling Remains Simplified
Only a limited number of studies explicitly model future electricity generation using turbine power curves.
Even when they do, turbine representations are often simplified and rarely account for differences between turbine power curve. As turbine designs continue to evolve, these assumptions can significantly influence projected outcomes.
Virtually All Models Remain Univariate
Perhaps the most important limitation is that most methodologies still rely primarily on wind speed, deriving their estimates from a single atmospheric variable.
Yet wind energy production is influenced by multiple interacting factors, including temperature impact on air density. Some studies introduce temperature based density corrections, but very few incorporate physically consistent multivariate climate projections capable of preserving the relationships between atmospheric variables at each time step.
Toward More Realistic Climate Risk Assessments for Wind Energy
The scientific community has made substantial progress in understanding how climate change may affect future wind resources. Methodologies have evolved from simple wind speed indicators toward increasingly realistic representations of energy production. Yet major uncertainties remain.
For wind developers, investors, operators, and policymakers, the next challenge is clear: moving beyond broad resource assessments toward methodologies capable of combining physically consistent climate projections with realistic representations of turbine behavior at the scale of actual wind farms.
To address these limitations, Callendar has developed a new methodology combining physically consistent climate projections, multivariate bias correction techniques and realistic wind farm production modelling at the asset level. Implemented in ClimateVision, Callendar's climatology platform, this approach can be applied to hundreds of wind farm locations worldwide, providing new insights into how climate change could affect future wind yields across the global wind industry or a specific portfolio.
The results of this global assessment will be unveiled on June 9 during a dedicated event in Paris, bringing together climate scientists, wind industry experts, developers and investors to discuss the implications of climate change for the sector and the strategies available to adapt. If you are involved in wind energy development, operations, financing or risk management, register here to join the discussion and discover the findings of this new research :
