Wind blade manufacturing: mould-bound capacity and the transport limit

Tooling, not floor space, sets the output ceiling

Ask a blade plant its capacity and the honest answer is a mould count multiplied by cycles. Cure and demould time is fairly fixed, so the only levers are adding mould sets, trimming the finishing tail, or running more shifts against a labour pool that may not exist locally. Each mould set is a serious capital item with a finite number of pulls before refurbishment, and it fits exactly one blade geometry. When a turbine maker retires a platform, the tooling has no second life. That asymmetry explains why blade operators push hard for volume commitments before they cut a new mould, and why an idle plant can still be capacity-constrained on the model a customer actually wants.

The layup bay is skilled manual work that resists automation

Fabric plies, root inserts, core kits and spar elements are positioned by crews working inside a mould that is longer than most buildings. Automated cutting and kitting have taken hold, and robots handle grinding and painting, but placement itself still depends on people reading a curved surface. A wrinkle, a dry patch or a misplaced core block will not announce itself until infusion has already committed the whole shell. Training a new crew takes far longer than hiring one, which is why blade plants in new regions ramp slowly and why scrap rates in the first year of a site are the number nobody quotes in the investment case.

A thin input base with qualification attached

Reinforcement fabric, epoxy systems, structural core and adhesives all come from a short list of qualified suppliers, and the qualification belongs to the turbine maker's design rather than to the blade factory. Swapping a resin is not a purchasing decision; it reopens coupon testing and, potentially, certification. Core material and carbon pultrusion have both had periods where allocation, not price, decided who could build. Operators who buy on the customer's approved list carry the supply risk while the customer holds the specification, so contract language on pass-through and on force majeure matters more here than the unit price negotiated at award.

You sell to a programme, not to a market

Blade orders arrive as platform supply agreements: a named turbine model, a named region, a volume band and an annual price step-down. There is no spot market to absorb a gap. If the customer loses an auction round or delays a project, the factory finds out through a schedule revision, not a cancellation, and the working capital already sits in resin drums and half-built shells. Plants serving a single turbine maker behave commercially like a subcontracted department with an inventory obligation. Diversifying across two turbine makers is the obvious answer and the hardest one, because tooling, specifications and confidentiality all cut against sharing a building.

Freight geometry chooses the site

A blade is light for its size and impossible to fold. Road transport needs escorts, junction surveys and sometimes physical modification of roundabouts; rail is rarely an option; a quayside solves the problem and creates a different one, since the plant then depends on vessel availability. Site selection therefore starts with the delivery corridor and works backwards to labour and power, which is the reverse of most greenfield exercises. As blade lengths grow, the catchment a given plant can serve shrinks, and the logistics saving from local production begins to outweigh differences in hourly labour cost between regions.

Frequently asked questions

Why do blade factories open and shut so frequently?

Because the asset is tied to a customer platform rather than to a product the plant could sell elsewhere. When a turbine maker moves to a new blade design, or shifts volume to a region with local content expectations, the moulds in the old site become scrap value. There is no alternative buyer for that tooling and no second product the building can make, so the rational move for the owner is closure rather than retooling on speculation.

How far has blade production actually been automated?

Automation has taken the edges of the process. Ply cutting and kitting are machine-driven, adhesive dispensing is increasingly robotic, and grinding, painting and inspection have moved to machines partly for quality and partly because the manual versions are unpleasant work. Layup and shell bonding remain human because each mould has geometric variation and each ply must be worked onto a compound curve. Full automation keeps being announced and keeps proving hard to hold at production rates.

What is the quality defect that actually hurts a blade maker?

Bond line integrity. A shell can pass visual inspection with a bonding gap that only becomes visible as a field failure years later, once fatigue loading has worked on it. That timing is the commercial problem: warranty exposure runs for the operating life of the turbine, and remediation happens offshore or at height. Blade plants therefore invest in process data capture and in bond gap control far beyond what the finished part inspection alone would justify.