Wind Turbine Yaw Misalignment Power Loss

A turbine pointing twelve degrees off the wind loses 6.41% of its output, and it is the cheapest fault in the industry to fix and one of the most commonly missed. On a 2,500 kW machine at a 40% capacity factor that is 562 MWh of 8,760 MWh a year, $22,473 at $40 per MWh, and across thirty machines $674,188 a year -- from a wind vane calibration. WHAT MAKES IT EXPENSIVE IS PERSISTENCE, NOT MAGNITUDE. A gust that swings the wind twenty degrees for a minute costs nothing worth measuring. A vane that reads a few degrees off, all year, costs a few percent of revenue on every hour the machine runs, silently, with no alarm and no visible symptom, and static misalignment is routinely found on machines that have been running for years. The field procedure is comparing nacelle position against the free-stream wind direction over a period and looking for a persistent BIAS rather than scatter: scatter is the controller doing its job, a bias is money. THE SHAPE OF THE CURVE TELLS A TECHNICIAN WHAT TO CHASE. Loss accelerates -- 1.14% at 5 degrees, 4.49% at 10, 9.88% at 15, 17.02% at 20, and 35.05% at 30 -- so at a 2% threshold anything past 6.6 degrees is worth a calibration and anything under it is not. THE EXPONENT IS AN APPROXIMATION AND BOTH FORMS ARE REPORTED. The cubed form follows from the rotor seeing only the along-axis wind component carried through the cubic power relation, but the true exponent is machine-specific and lies between 2 and 3, and real measured losses often fall closer to the squared form at small angles -- 4.32% rather than 6.41% at twelve degrees -- so the cubed result is the pessimistic bound rather than the answer. It does not account for the increased fatigue loading a persistent misalignment causes, which can matter more than the energy, or for wind veer and shear across the rotor which produce an effective misalignment no yaw correction removes. It does not distinguish static from dynamic yaw error, evaluate nacelle anemometer or vane transfer functions, or address cable twist, yaw drive loading, or yaw system faults. The turbine manufacturer yaw calibration procedure and the operator performance engineer govern.

retained power = cosine of the yaw error raised to the third power, from the rotor seeing only the along-axis wind component carried through the cubic power relation; the squared form is reported beside it, and the acceptable angle is the arccosine of the cube root of one less the acceptable loss.

The cosine-power yaw loss convention by name, with both the squared and cubed forms reported because the true exponent is machine-specific and lies between 2 and 3. The turbine manufacturer's yaw calibration procedure and the operator's performance engineer govern.

Trigonometry on the operator's own nacelle position and wind direction records; no manufacturer loss table is reproduced.

Estimate. AHJ and licensed professional govern.

Field names used by the API: yaw_error_deg, rated_power_kw, capacity_factor_pct, energy_price_per_mwh, turbine_count, acceptable_loss_pct, loss_cubed_pct, loss_squared_pct, annual_mwh, lost_mwh, lost_usd, fleet_lost_usd, acceptable_angle_deg

• Persistence is the cost, not magnitude a few degrees all year beats a large gust every timeoperations practice
• A bias is money, scatter is not scatter is the controller doing its joboperations practice
• The exponent is between 2 and 3 the cubed form is the pessimistic bound, not the answermeasured yaw loss studies