Offshore wind is expanding into regions where tropical cyclones are a defining feature of the climate — the Gulf of Mexico, the western Pacific, the South China Sea, the Philippines, and increasingly the U.S. Atlantic coast. But the engineering frameworks we use to design turbines in these areas were largely developed for the North Sea, where tropical cyclones simply don't occur. This disconnect is one of the most pressing challenges facing the offshore wind industry today.
In a recent preprint — "Grand Challenges in Designing Resilient Wind Energy Systems in Areas Prone to Tropical Cyclones" (Deskos et al., 2026), now under review in Wind Energy Science — my co-authors and I lay out the problem systematically. I want to unpack some of the key themes here, in plain language, for the developers, engineers, and investors who are confronting these challenges on real projects.
The problem in a sentence
Tropical cyclones produce wind and wave conditions that are fundamentally different from what the IEC design standards were built around — and the current standards don't adequately account for that difference.
The IEC 61400-3-1 standard does include a "T-Class" for tropical cyclone regions, with a reference wind speed of 57 m/s at a 50-year return period. That's a start. But the T-Class says very little about the character of the wind — the turbulence structure, the shear and veer profiles, the rapid directional shifts, or the coupling between wind and waves that makes tropical cyclones so destructive. And it says almost nothing about what happens beyond the 50-year event, which is exactly the scenario that lenders and insurers care about most.
What makes tropical cyclones different
If you're used to thinking about wind in North Sea terms, tropical cyclones break almost every assumption you take for granted:
Turbulence. The standard spectral models (Mann, Kaimal) were developed from measurements in flat terrain and extratropical boundary layers. In a tropical cyclone, the turbulence is driven by fundamentally different physics — convective cells in the eyewall, mesoscale vortices in the rainbands, and a boundary layer structure that is shaped by air-sea heat fluxes, not just mechanical shear. The turbulence intensities are higher, the coherent structures are larger, and the standard 10-minute averaging window can obscure the short-duration gusts that actually drive peak loads.
Wind-wave coupling. In a cyclone, wind and waves are not independent. The wave field is driven by the storm's wind field, but it also feeds back into the boundary layer through surface roughness and spray. You cannot treat the wind and wave loads as separate, uncorrelated inputs — yet that is effectively what many design workflows do.
Rapid direction changes. Tropical cyclones can produce directional shifts of 90 degrees or more as the storm passes over a site. If a turbine is idling and yaw-locked (as most are during extreme events), it can be exposed to sustained crosswind loading that the structure was never designed for.
Idling loads. During a major tropical cyclone, turbines are shut down — they're idling, not generating. But idling doesn't mean unloaded. Under certain conditions, parked rotors can experience stall-induced vibrations (SIVs) driven by negative aerodynamic damping, particularly near the first flapwise and edgewise natural frequencies. These oscillations can be divergent and are poorly captured by standard design load cases.
The standards gap
The paper reviews the major standards in use today — IEC 61400-1 and 61400-3, DNV-ST-0437, API RP 2MET, and others — and identifies a consistent pattern: they provide partial coverage at best.
The IEC T-Class specifies a reference wind speed but relies on turbulence models calibrated to extratropical conditions. The partial safety factor of 1.35 used in design load case DLC 6.1 (extreme wind, parked turbine) was derived from statistical distributions that reflect North Sea-type variability — not the heavier tails of tropical cyclone wind distributions, where the ratio between the 50-year and 500-year wind speed can be significantly higher than in extratropical climates.
What this means in practice: the safety margin that feels comfortable in the North Sea may not be sufficient in the Gulf of Mexico or offshore Japan. Our analysis shows that increasing the design storm return period from 50 to 100 years can increase maximum overturning moments by roughly 30% for jacket structures and up to 27% for monopiles. The standards do not require this sensitivity analysis. They should.
What we're missing: data
One of the most sobering findings is how little data we actually have. Despite decades of hurricane research, comprehensive wind measurements at turbine-relevant altitudes (20–300 m above the surface) during tropical cyclones remain sparse — especially offshore. Most of what we know about tropical cyclone wind profiles comes from dropsondes, aircraft reconnaissance, and satellite retrievals, none of which provide the continuous, high-frequency time series needed for turbine load calculations.
New observational capabilities are emerging — scanning lidars, Ka-band radars like the Texas Tech system deployed during Hurricane Laura, uncrewed surface vehicles like Saildrones — but they are not yet systematically integrated into the design workflow. The wind energy industry needs a concerted measurement campaign at turbine-relevant scales in tropical cyclone environments. Without it, we are designing for conditions we haven't properly characterized.
A path forward
The paper identifies five priority areas, and I think they're worth stating directly:
1. Expand measurements. We need sustained, standardized observations of tropical cyclone winds at hub height and rotor-plane altitudes, both onshore and offshore. This means deploying hardened instrumentation in exposed areas and coordinating across national labs, universities, and industry.
2. Validate hazard models. Synthetic tropical cyclone track models and climate-informed hazard assessments are powerful tools, but their application to offshore wind is still immature. They need to be validated against observed offshore conditions and integrated into site assessment workflows.
3. Develop reduced-order simulators. Full LES of a tropical cyclone impacting a wind farm is possible now (we've done it), but it's too computationally expensive for routine design. We need physics-informed reduced-order models that can capture cyclone-specific turbulence and inflow dynamics at a cost compatible with engineering practice.
4. Revise the standards. The IEC T-Class needs to evolve beyond a single reference wind speed. It should incorporate probabilistic safety factors calibrated to tropical cyclone hazard curves, explicit requirements for coupled wind-wave analysis, and guidance on directional load cases that reflect the rapid veer observed in cyclone events.
5. Build a risk framework. The earthquake engineering community has had a probabilistic risk assessment framework (hazard, fragility, consequence) for decades. Wind energy in cyclone regions needs the same — a transparent, quantitative framework for estimating the probability and cost of structural failure under extreme events.
Why this matters for your project
If you're developing, financing, or insuring an offshore wind project in a tropical-cyclone-exposed region, these aren't abstract research questions. They are the questions that determine whether your design is adequate, whether your risk assessment is credible, and ultimately whether your project is bankable.
The full preprint is available open-access at doi.org/10.5194/wes-2026-32. If you'd like to discuss how any of this applies to your specific site or project, get in touch.