Explained, Technology
Onshore Wind
Onshore wind is the workhorse of the transition: roughly 1,200 gigawatts installed, more than every other renewable except solar, built quietly over four decades in fields, on ridgelines and across steppe. The machine is the same physics as its offshore sibling, a rotor converting moving air into torque, but the engineering problem is different. Offshore, the sea constrains you. Onshore, everything else does: the road the blade must travel, the crane that must reach the hub, the neighbour who can see it, and the grid connection that may take longer to secure than the wind farm takes to build.
The machine
A modern onshore turbine is four systems pretending to be one object: a rotor to capture energy, a drivetrain to convert it, a tower to put the rotor where the wind is, and a control system to keep the other three alive. Everything else is detail.
Three composite blades, pitch-controlled, on rotors now commonly 150 to 175 metres in diameter. Swept area is the whole game: power available scales with the square of rotor diameter, which is why blades have grown faster than ratings. Each blade is an aerofoil, a structure, a lightning conductor and a fatigue problem, simultaneously, for 25 years.
Two philosophies compete. Geared machines spin a fast, compact generator through a three-stage gearbox, lighter and cheaper, with the gearbox as the classic wear item. Direct drive deletes the gearbox and turns a large-diameter generator at rotor speed, fewer moving parts, more tonnes of nacelle. Both work. Neither has won.
Tubular steel in sections, increasingly hybrid concrete-steel as hub heights pass 160 metres chasing steadier wind. Below ground, a gravity foundation: several hundred cubic metres of reinforced concrete whose job is to make sure a 500-tonne lever in a storm stays a turbine and not a trebuchet.
Turbines feed a buried medium-voltage collector network, typically 33 kV, into an onsite substation that steps up to the transmission or distribution voltage. The grid code lives here: fault ride-through, reactive power capability and, increasingly, grid-forming behaviour are contractual obligations, not options.
What limits the machine
The interesting engineering question onshore is not how big a turbine can be, China has answered that, a 15 megawatt machine with a 270 metre rotor is running in Jilin province. It is how big a turbine can be delivered. A blade is the largest single-piece component moved on roads anywhere in industry, and every bridge, roundabout and overhead line between the factory and the site is a design constraint. Western markets top out around 7 megawatts largely because their road networks, crane fleets and planning-imposed tip height limits say so; China builds bigger partly by building blade factories near the wind and, increasingly, by segmenting blades for final assembly on site.
The second limit is social. An onshore turbine has neighbours, and noise limits, shadow flicker rules, visual impact assessments and aviation lighting requirements all shape layouts before the wind resource does. In much of Europe the binding constraint on onshore wind is not engineering at all: it is the permitting queue, followed by the grid connection queue. A machine that takes a year to build routinely waits five or more for its paperwork.
The third is the wind itself. Developers now deliberately fit large rotors to modest generators, lowering the specific power so the machine reaches full output in gentler winds. The result is capacity factors that would have seemed impossible onshore a decade ago, 40 percent and above on good sites, at the cost of leaving the rarest, strongest winds unharvested. That trade is almost always worth it, steady output is worth more to the system than occasional peaks.
What breaks
A wind turbine is a fatigue machine: every gust is a load cycle, and a 25-year design life is several hundred million rotations of accumulating damage. The maintenance economics of an onshore fleet are set by a short list of known offenders.
- Gearboxes. The historic reliability headline. Bearing wear and gear tooth damage, driven by transient loads the design models underestimated for years. Condition monitoring, vibration and oil particle analysis, exists mostly to see this coming, because an unplanned gearbox exchange means a large crane and a six-figure bill.
- Blades. Leading edge erosion, rain wearing the aerodynamics off the blade at 300 km/h tip speeds, lightning strikes, and manufacturing defects that reveal themselves years later. Drone inspection has largely replaced rope access for finding them.
- Pitch and yaw systems. Unglamorous, numerous, and the most frequent fault categories in most fleet statistics. The pitch system is also the primary aerodynamic brake, which makes it a safety system, not a convenience.
- Power electronics. Converters fail more often than generators. As machines have grown, the converter has quietly become one of the highest-failure-rate assemblies in the nacelle.
The second life of old sites
The first generation of wind farms is ageing out, and the best wind sites were, rationally, used first. Repowering, removing 20-year-old machines and replacing them with modern ones, is now a defining feature of mature markets: a typical project replaces ten small turbines with three or four large ones and doubles the site's output while reducing the machine count. It sounds like arithmetic and is actually politics and paperwork: the consented tip heights, grid connection capacity, and environmental baselines of the original project rarely fit the new machines, so a repowering is often a full new consent on a site that already proved the concept for two decades.
Control systems, and keeping them safe
Every turbine runs on a programmable controller managing pitch, yaw, and the converter in real time, reporting into a site SCADA system, which reports into a remote operations centre that may be watching thousands of machines across a continent. The fleet is operated over networks: vendor service links for warranty diagnostics, remote firmware updates, cloud analytics platforms fed from every nacelle. The industry standardised the data model for all of this years ago, IEC 61400-25, which is excellent for interoperability and means an intruder who learns one fleet has learned most of them.
This is the part of the technology that changed fastest and is discussed least. An onshore wind fleet is distributed critical infrastructure with a very large perimeter: hundreds of unmanned sites, physical access controlled by a padlock and a gate, network access mediated by whatever the operator and its half-dozen service contractors have configured. The 2022 KA-SAT attack, which severed remote monitoring of thousands of German turbines as collateral damage in the first hours of the Ukraine invasion, made the dependency visible: the machines kept turning, but the operator's eyes were gone. Securing this estate, segmented networks, controlled remote access, monitored OT traffic, an actual inventory of what is connected, is not an accessory to wind farm operation. It is wind farm operation. The discipline is explained at OT Cybersecurity.
Where it fits in the system
Onshore wind's role in the transition is being the cheap, proven, fast-to-build partner: paired with solar it smooths the daily profile, since wind in most climates blows more in winter and at night, exactly when solar does not. Its system problems are the shared ones, connection queues and curtailment where the grid lags the build, and its politics are the most local of any technology on this platform, decided planning committee by planning committee.
How it compares with the other five technologies, and where the 2025 record build actually went, is on the Markets page.