Wind Turbine Energy Conversion (2D)

Side view of a three-turbine row, rotor front view and the blade-element velocity triangle, with the same model as the 3D page.

Presets

Playback

Rotor

Wind and site

Control and farm

Layers

Model check

The science

Power in the wind and the Betz limit

The kinetic power flowing through the rotor disc of area A = pi R^2 is P_wind = 1/2 rho A v^3. A rotor extracts the fraction Cp of it. Momentum theory (actuator disc) with axial induction a gives Cp = 4a(1-a)^2 and thrust coefficient Ct = 4a(1-a); the maximum is Cp = 16/27 = 0.593 at a = 1/3 (Betz limit). Real rotors stay below it because of wake rotation, blade drag and a finite number of blades.

Cp(lambda, beta)

The tip-speed ratio is lambda = omega R / v and beta is the blade pitch angle. The page uses the widely used empirical fit Cp = c1 (c2/l_i - c3 beta - c4) exp(-c5/l_i) + c6 lambda with 1/l_i = 1/(lambda + 0.08 beta) - 0.035/(beta^3 + 1) and c = (0.5176, 116, 0.4, 5, 21, 0.0068). It peaks near lambda = 8.1, Cp = 0.48 at beta = 0. Pitching a blade out of the wind lowers Cp, which is how power is shed above rated wind.

Drive-train dynamics and control

J d(omega)/dt = T_aero - T_gen - T_loss with T_aero = 1/2 rho pi R^3 v^2 Cp / lambda and rotor inertia J ~ 2.5 R^4 (scaled by the slider). Below rated wind the generator torque follows T_gen = k omega^2 with k = 1/2 rho pi R^5 Cp_max / lambda_opt^3, which holds lambda at its optimum (variable-speed operation). Above rated, torque is limited to P_rated/(eta omega_rated) and a PI controller pitches the blades (rate-limited to 8 deg/s) to hold the rated rotor speed. Rotor speed is also capped by a tip-speed limit of 80 m/s. Electrical power is eta T_gen omega with eta = 0.95. Cut-in is 3 m/s and cut-out 25 m/s, where the blades feather and a brake parks the rotor.

Wind and energy yield

The hub wind is a mean plus an Ornstein-Uhlenbeck turbulence process (time constant 10 s, standard deviation = turbulence intensity x mean), low-pass filtered by the rotor-disc size. The long-term resource is a Weibull distribution f(v) = (k/c)(v/c)^(k-1) exp(-(v/c)^k) with c = v_mean / Gamma(1 + 1/k). Annual energy is 8760 h x integral of P(v) f(v) dv using the steady power curve; the capacity factor is that energy divided by rated power x 8760 h. Air density follows altitude and temperature: rho = p/(287.05 T) with p = 101325 (1 - 2.256e-5 h)^5.256.

Wakes in a row

The Jensen (Park) model gives the speed deficit behind a turbine: delta = (1 - sqrt(1 - Ct)) (R/(R + k x))^2 with wake-decay constant k = 0.075 (onshore). The third turbine sees the root-sum-square of two deficits. Row efficiency is total power divided by three times the power of an isolated turbine.

Honest simplifications

The Cp surface is an empirical fit for a generic large rotor, not a blade-element-momentum solution of this exact blade; no tower shadow, yaw error, wind shear, blade flexibility, gearbox or grid dynamics. Turbines 2 and 3 are quasi-steady (no own rotor dynamics). The Jensen wake is a top-hat engineering model. Annual energy ignores availability, electrical and wake losses beyond the row shown. Values are educational estimates, not a bankable yield assessment.

Model checks: a grid search of Cp over the whole (lambda, pitch) plane against an independent golden-section optimum and the Betz limit; the Weibull pdf integral, mean and cubic moment against Gamma-function closed forms; actuator-disc power from the kinetic-energy flux difference against Cp x wind power; and a headless 900 s steady-wind run of the dynamics.

Power coefficient Cp(lambda) for several blade pitch angles
Steady power curve and the Weibull wind distribution
Row efficiency versus turbine spacing (Jensen wake)
Wind at the hub (last 5 minutes)
Electric power (last 5 minutes)
Rotor speed (last 5 minutes)