Floating wind turbines had a dangerous wobble that engineers finally tamed with a single “ear” already on every machine
Europe has set itself an ambitious target: renewables covering at least 42.5 percent of its electricity by 2030, with wind already supplying nearly a fifth of demand today. Floating offshore turbines are central to that plan — they can operate in deep waters like the Mediterranean and Atlantic where fixed-foundation designs simply can’t reach. But moving turbines onto floating platforms creates an unexpected problem that engineers have struggled to solve without sacrificing performance.
Why floating turbines play by different rules
On land, blade-pitch controllers have a straightforward job: when wind pushes the rotor too fast, the blades feather to shed load and bring generator speed back to its target. Move that same turbine onto a floating platform, though, and the same logic can make things dramatically worse.
When the platform pitches backward under aerodynamic thrust, the controller reads a drop in rotor speed and feathers the blades — reducing thrust, which lets the platform pitch back even further.
The new SIMO controller achieved the lowest rotor speed oscillations across that entire range, outperforming all four alternatives on generator speed regulation.
The cycle continues, amplifying rather than resisting motion. Engineers call this “negative damping,” and it traces mathematically to right-half-plane zeros in the transfer function from blade pitch to generator speed.
The sensor problem that has long limited control design
The engineering community has known about this problem for nearly two decades. Most proposed fixes share one requirement: extra sensors — typically an accelerometer on the nacelle or hull — measuring fore-aft motion and feeding that signal back into the controller. Salt, spray, and constant mechanical stress degrade sensors over time, and a malfunctioning sensor can actively hurt performance or push the system toward instability.
The simpler alternative, known as “detuning,” deliberately slows the controller so it never excites the platform — but power quality suffers and the underlying mathematical constraint stays firmly in place.
A new controller that works with generator speed alone
Researchers at TU Delft have published a design that sidesteps the sensor dependency entirely. It uses only the generator speed signal — a measurement already present on every turbine — and adds a second actuator channel, generator torque, to handle the frequency band where blade pitch is blocked by the right-half-plane zeros.
The key element is an inverted-notch filter placed on the generator torque channel, activating only within that narrow problematic band. When blade pitch is mathematically constrained, generator torque steps in. The researchers also reduced this two-channel structure to a single-loop form compatible with standard industrial proportional-integral tuning, requiring no exotic hardware or additional instrumentation.
Head-to-head: how the new design fared in simulation
The controller was tested in OpenFAST, a high-fidelity nonlinear simulation tool, against four competing strategies: a detuned PI controller, a robustly tuned PI controller, a blade-pitch damper configuration (MISO), and a generator-torque feedback configuration (MIMO). Simulations ran under turbulent wind and irregular waves across the full above-rated range of 12 to 24 meters per second.
The new SIMO controller achieved the lowest rotor speed oscillations across that entire range, outperforming all four alternatives on generator speed regulation. It also reduced platform pitch oscillations and tower base fatigue loads relative to the detuned baseline — a structural benefit the design wasn’t explicitly optimized for. The main trade-off is increased generator torque activity, raising drivetrain loads and introducing greater power variability.
Trade-offs, limits, and what comes next
No single controller wins on every metric. The SIMO design leads on speed regulation; the blade-pitch damper configuration leads on platform motion suppression; the standard robust PI produces the cleanest power output. Rather than picking one approach, the research team recommends an adaptive strategy — switching between controllers depending on whether the priority is power quality, structural fatigue, or rotor speed stability.
There’s a practical resilience argument for the SIMO controller too: because it requires no platform measurements, it can serve as a reliable fallback when sensors fail. Looking ahead, the authors point to feedforward control — via LiDAR wind measurements or wave state data — as a promising path to relaxing the bandwidth demands that make negative damping so difficult to manage at scale.
Reference: Hegazy, A., Naaijen, P., and van Wingerden, J.-W.: Control design for floating wind turbines: a novel feedback control structure, Wind Energ. Sci., 11, 3455–3474, https://doi.org/10.5194/wes-11-3455-2026, 2026.
Daniel García is an Editor-in-Chief with strong expertise in structural work and engineering principles. He combines this technical foundation with deep knowledge of energy, spatial design, and emerging technologies, bringing a forward-thinking and analytical approach to editorial leadership.
