Wind Power: From Kinetic Energy to Gigawatts

7.1 — From moving air to a rotating shaft

Wind is air with momentum. A turbine intercepts a moving column of that air, extracts part of its kinetic energy as torque on a shaft, and lets the slower air continue downstream. Nothing is burned and nothing is consumed — the rotor simply takes a tax on the momentum of the fluid passing through it.

Every design decision in a modern turbine follows from two consequences of that fact. The air must keep moving, so you cannot extract all of it. And the amount available depends violently on how fast the air was moving to begin with.

7.2 — The cube law: the most important equation in wind

Mass flow rate through the rotor is proportional to velocity, and the kinetic energy each unit of mass carries is proportional to velocity squared. Multiply them and power scales with the cube of wind speed.

This is the defining economic fact of the wind industry. A site with 10% more wind does not have 10% more energy. It has roughly 33% more.

7.3 — Swept area: why rotors keep getting bigger

The second lever in P = ½ρAv³ is the area the rotor sweeps. Since area scales with the square of diameter, a modest increase in blade length produces a disproportionate increase in captured energy.

This is the clearest trend in the industry. Rotor diameters that were 80–90 m in the 2000s now run 150–170 m onshore and beyond 230 m offshore. Turbines have not grown because bigger looks impressive. They have grown because the physics rewards area quadratically while tower, foundation and installation costs grow more slowly than the energy gain.

7.4 — Air density: the quiet third variable

Air density appears linearly in the power equation and is easy to overlook, but it varies meaningfully with altitude and temperature. Cold, dense, sea-level air carries more energy per cubic metre than hot, thin air at elevation.

A high-altitude site in a hot climate can lose 15–25% of available power relative to standard conditions at exactly the same wind speed. This is why energy yield models use site-specific air density rather than the standard 1.225 kg/m³, and why manufacturers publish density-corrected power curves.

7.5 — The Betz limit: why no turbine can exceed 59.3%

If a rotor extracted all the kinetic energy from the air passing through it, that air would come to a complete stop behind the rotor — and stationary air cannot make way for the air arriving behind it. Mass flow would cease and extraction would fall to zero. Conversely, a rotor that barely slows the air extracts almost nothing.

Between those two failures lies an optimum. Albert Betz showed in 1919 that it occurs when the rotor slows the wind to one-third of its upstream speed, capturing at most 16/27, or about 59.3%, of the available energy.

This is a limit of fluid mechanics, not of materials, manufacturing or budget. It is conceptually the counterpart of the bandgap ceiling on solar cell efficiency in Chapter 6 — a limit imposed by physics that no engineering programme will lift.

7.6 — Power coefficient and tip-speed ratio

A rotor’s power coefficient is not a fixed property. It depends strongly on tip-speed ratio, the ratio of blade tip speed to wind speed. Spin too slowly and much of the air slips between the blades untouched. Spin too quickly and the blades operate in each other’s disturbed wake, stalling and generating drag rather than lift.

Every rotor has an optimal λ, typically around 7–9 for modern three-bladed machines, where C_p peaks at roughly 0.45–0.50.

This is precisely why modern turbines are variable speed. As wind speed changes, the control system adjusts rotor speed to hold λ near its optimum, keeping C_p at its peak across a wide operating band. Fixed-speed machines hit optimal λ at only one wind speed and lose efficiency everywhere else — one of several reasons they have effectively disappeared from utility-scale deployment.

7.7 — Lift, not drag: how a blade actually works

A common misconception is that wind pushes the blades around like a child’s pinwheel. Modern turbine blades are aerofoils and work like aircraft wings: air moving over the curved surface generates a pressure differential and therefore lift, which acts perpendicular to the relative airflow and produces torque about the hub. Drag is the parasitic component, and blade design is largely an exercise in maximising the lift-to-drag ratio.

Because the blade is rotating, the airflow it experiences is the vector sum of the incoming wind and its own rotational velocity — and rotational velocity increases along the blade’s length. The tip moves far faster than the root, so it meets the air at a completely different angle. This is why blades are twisted along their span and why chord width tapers: each radial station is designed for the local relative wind it will actually encounter. A modern blade is not one aerofoil but dozens, blended along a single structural member.

7.8 — The power curve: four operating regions

The power curve is the single most important document in a turbine’s specification, mapping wind speed to electrical output. It has four distinct regions, and the boundaries between them explain most of a wind project’s behaviour.

7.9 — Wind shear and the economics of hub height

Friction with the ground slows air near the surface, so wind speed increases with height in a profile governed by surface roughness. The exponent α is small over smooth water and large over forests, buildings and complex terrain. Because power follows v³, even a modest speed gain from a taller tower compounds into a substantial energy gain — which is why hub heights have marched from 80 m toward 120–160 m onshore.

7.10 — Turbulence intensity and turbine class

Mean wind speed determines how much energy a site holds. Turbulence intensity — the standard deviation of wind speed divided by its mean — determines how hard that site is on the machine. Turbulent inflow imposes rapidly fluctuating loads on blades, hub, drivetrain and tower, driving the fatigue damage that governs component life.

The IEC 61400 standard classifies turbines by reference wind speed — Class I for high-wind sites through Class III for low-wind sites — and by turbulence category, A for high through C for low. A Class III turbine, with long blades, lighter structure and low specific power, is engineered to harvest low-wind sites efficiently and will not survive a Class I regime.

7.11 — The drivetrain: geared versus direct drive

A large rotor turns slowly — typically 5–15 rpm — while conventional generators want hundreds or thousands of rpm. Three architectures resolve that mismatch.

Neither architecture is simply better. Offshore, where a nacelle repair may require a jack-up vessel and a weather window, high reliability commands a premium and direct drive is attractive. Onshore, where a crane can reach site in days, the capital saving from a geared machine often wins.

7.12 — Pitch and yaw: how a turbine protects itself

Two control systems govern how the rotor meets the wind. Yaw rotates the nacelle about the tower axis to keep the rotor facing into the wind, guided by a nacelle-mounted anemometer and wind vane. Pitch rotates each blade about its own long axis to change its angle of attack.

Pitch control does the critical work. Below rated wind speed, blades sit at an angle that maximises lift and therefore energy capture. Above rated, the controller progressively pitches blades toward the wind to reduce lift and hold output flat — the mechanism that creates Region III of the power curve. Above cut-out, blades are feathered nearly edge-on to the wind, minimising aerodynamic loading so the structure can ride out the storm.

Pitch is simultaneously the turbine’s power regulator and its primary aerodynamic brake, which is why pitch systems are safety-critical and typically triple-redundant with independent backup power per blade. A turbine that cannot feather in a storm is a turbine that does not survive one.

7.13 — Generators and power electronics

The industry has moved decisively toward full-converter architectures, largely because grid codes have tightened. As Chapter 5 established, a converter-interfaced generator provides no inherent rotational inertia to the grid. As the synchronous fleet retires, system operators increasingly require wind plants to supply fast frequency response, synthetic inertia, reactive power and voltage support. Delivering those services requires full control of the power electronics, which a full converter provides and a DFIG only partially can.

7.14 — Specific power: the design dial that sets capacity factor

Specific power is the most useful single number for understanding a turbine’s character, and it is routinely overlooked. It expresses how much generator you have bought relative to how much rotor you have bought.

A low specific power machine — a big rotor on a modest generator — reaches rated output at a lower wind speed, spends more hours at full output, and therefore delivers a much higher capacity factor and a flatter, more predictable production profile.

This is the direct analogue of the DC/AC ratio decision in Chapter 6. In both technologies, deliberately oversizing the energy-collecting surface relative to the power-conversion equipment raises utilisation, smooths output and lowers levelised cost — at the price of spilling some energy in the very best conditions.

7.15 — Wind resource assessment: Weibull, wind rose and P50/P90

A site is not characterised by a single wind speed but by a distribution of wind speeds. That distribution is conventionally fitted with a Weibull function described by a scale parameter, roughly how windy, and a shape parameter, how variable. A wind rose adds the directional dimension, showing how much energy arrives from each compass sector — the input that ultimately drives turbine layout.

Combining the wind speed distribution with the turbine’s power curve, bin by bin, produces the annual energy production estimate. The result reveals something genuinely counter-intuitive about where a wind farm’s revenue comes from.

7.16 — Wake effects: why turbine 40 earns less than turbine 1

This is the loss mechanism with no equivalent in solar. A turbine that extracts energy from the wind necessarily leaves behind a wake — a region of slower, more turbulent air. Any turbine standing in that wake sees a lower inflow speed, and because power follows v³, even a small velocity deficit produces a large power deficit. Downstream machines also endure higher turbulence, accelerating fatigue and raising maintenance cost.

Wakes recover as ambient air mixes back in, but recovery takes many rotor diameters — and recovery is slower offshore, where low ambient turbulence over smooth water means less mixing.

7.17 — Layout optimisation and spacing

Turbine layout is a genuine multi-variable optimisation, and one of the highest-leverage decisions a developer makes. Spacing turbines further apart reduces wake losses but requires more land or seabed, longer internal cabling, more access roads and a larger project footprint. Packing them tighter raises energy density per hectare but compounds wake losses and fatigue loading.

Typical practice places turbines roughly 5–9 rotor diameters apart along the prevailing wind direction and 3–5 diameters crosswind — but those are starting points, not answers. Real layouts are optimised against the site-specific wind rose, terrain, noise and shadow-flicker constraints, environmental exclusion zones, landowner boundaries and cable routing cost.

The objective function is not maximum energy. It is minimum levelised cost subject to a long list of constraints.

7.18 — Onshore versus offshore

Offshore is not simply “onshore, but at sea”. It is a marine construction industry that happens to install generators, with a cost structure, risk profile and financing requirement closer to offshore oil and gas than to onshore renewables.

7.19 — Offshore foundations

Monopiles — a single large-diameter steel tube driven into the seabed — dominate shallow water up to roughly 40 m and are the cheapest option where geology permits. Jackets, lattice steel structures borrowed from offshore oil, extend viability into deeper water and heavier turbines. Gravity bases rely on mass rather than penetration and suit specific seabed conditions.

Floating foundations are the strategically important category. Fixed foundations become impractical beyond roughly 60 m depth, which excludes most of the world’s coastline — including waters off Japan, much of the US West Coast, the Mediterranean and large parts of the Atlantic seaboard.

7.20 — Export cables and offshore substations

An offshore wind farm generates electricity tens or hundreds of kilometres from where it is consumed. Turbines connect via medium-voltage inter-array cables to an offshore substation, which steps voltage up for the export cable to shore.

The transmission choice follows directly from the physics of Chapter 5. Beyond roughly 80–100 km, HVAC submarine cable becomes impractical because the cable’s own capacitance generates large charging currents that consume conductor capacity and produce reactive power that must be compensated. Long-distance offshore projects therefore use HVDC, which has no such charging-current penalty and lower losses over distance — at the cost of expensive converter stations at both ends.

7.21 — Balance of plant

As with solar, the turbine is only part of the project. Balance of plant covers foundations, towers, inter-array cabling, substations, transformers, switchgear, SCADA, access roads or vessel infrastructure, grid connection works and civil engineering.

In offshore projects, balance of plant and installation frequently exceed the cost of the turbines themselves. A marginally cheaper turbine can easily produce the more expensive project once installation and foundation implications are counted — the same trap as judging a solar plant by module price in Chapter 6.

7.22 — From gross to net: the wind loss cascade

Gross energy from the power curve and the wind distribution is never what reaches the meter. Losses stack multiplicatively.

Applied together these commonly convert a gross estimate into a net estimate 12–25% lower. Curtailment is the entry to watch most closely: unlike the others it is not primarily an engineering variable but a grid and market variable, and — exactly as Chapter 5 argued — it can become the dominant loss on a congested network long before anything mechanical is at fault.

7.23 — Availability, O&M and degradation

Availability is the share of time a turbine is capable of operating when wind is present, and it is a contractual centrepiece: turbine supply agreements typically warrant availability in the high nineties, with liquidated damages below that threshold.

Note the distinction from solar. A wind turbine is a machine with rotating parts, bearings, gearboxes, hydraulics and blades under fatigue loading, so its maintenance burden is fundamentally higher than a solar plant’s. This is the single largest structural difference between the two technologies’ operating cost lines.

Offshore, the cost driver is access rather than parts. A component replacement requiring a jack-up vessel may involve heavy mobilisation cost and a wait for a suitable weather window, so a modest failure can produce weeks of lost production. This is why offshore operators invest heavily in condition monitoring and predictive maintenance — the objective is to convert unplanned failures into planned interventions that can be scheduled into good weather.

Blades also degrade. Leading-edge erosion from rain, hail and airborne particulates progressively roughens the aerofoil, measurably reducing the power curve over time and requiring periodic repair or protective coating.

7.24 — LCOE and the cost of capital

Wind shares solar’s essential financial structure: high upfront capital, effectively zero fuel cost, and a long operating life. That makes levelised cost acutely sensitive to the discount rate, exactly as set out in Chapter 4. Wind carries one significant complication relative to solar — a materially higher operating cost, because there is real machinery to maintain, particularly offshore.

7.25 — Capture price: what wind actually gets paid

Wind faces the same cannibalisation dynamic set out in Chapter 6, with a different shape. Solar generates in a tight, predictable daily block, so it depresses prices in the same few midday hours every day. Wind is correlated across very large geographic areas and blows for extended multi-day periods, so a windy front can depress prices across an entire market for days at a time — including overnight, when demand is lowest and there is least room to absorb it.

The consequence is that in high-penetration wind markets, periods of very high wind and low demand can drive wholesale prices to zero or negative. A wind project’s capture price therefore falls below the market average and continues to decline as penetration grows. The strategic response is the same as for solar: storage, transmission that connects to less-correlated demand, flexible loads, and long-term contracts that shift price risk onto a counterparty.

7.26 — Wind + storage, and complementarity with solar

Batteries do different work for wind than for solar. For solar, storage principally shifts a predictable daily surplus a few hours into the evening. For wind, storage is used more for ramp smoothing, delivering firm contracted profiles, capturing intraday price spreads and reducing curtailment during congested high-wind periods.

Multi-day lulls remain outside the economic reach of lithium-ion at current costs. Bridging them is a job for transmission, flexible generation, demand response or long-duration storage — not a bigger BESS.

Wind and solar are also usefully complementary in many regions: wind frequently peaks overnight and in winter, precisely when solar output is absent or weakest. A hybrid plant sharing one grid connection, one substation and one set of interconnection rights can achieve a much higher combined utilisation of that connection than either technology alone.

7.27 — Advantages and limitations

7.28 — The complete wind stack

Read it downward and each layer constrains the one below. Read it upward and each layer decides whether the one above was worth building. Layers 1 and 2 are given to you; layer 3 is capped by physics; and layers 5 and 6 are where most of the money is actually won or lost.

7.29 — The core wind mental model

• •Resource — what is the wind speed distribution, not just the mean?
• •Height — what does the shear profile give you for each extra metre of tower?
• •Rotor — how much area are you sweeping, and at what specific power?
• •Machine — what does the power curve look like, and is the turbine class right for the site?
• •Layout — how much energy do the turbines take from each other across the full wind rose?
• •Losses — wakes, availability, electrical, degradation, curtailment?
• •Grid — is there interconnection capacity, and is the network strong enough?
• •Revenue — what is the capture price, not the average market price?
• •Uncertainty — how wide is the P50-to-P90 gap, and what does that do to debt sizing?
• •Financing — what is the cost of capital, and what does it do to LCOE?

The recurring theme across Chapters 6 and 7 is worth stating plainly: with solar and wind you are no longer buying fuel. You are buying capital equipment and a statistical resource, then financing them. That is why the decisive variables are not the ones the technology press writes about, but capacity factor, capture price and cost of capital.

Chapter 8 turns to a generator that inverts almost every assumption in these two chapters: one that carries no resource risk at all, runs at ninety percent capacity factor, and buys that firmness with the most extreme capital intensity in the energy system. Nuclear power.