Windfall

01The windfall

A national stock of critical material, with a date on it.

£620m

of neodymium magnet material in offshore turbines installed 2025–2030, at £85/kg

NEPC Windfall · p.5

2038

first major offshore recovery — Gunfleet Sands lease expires, Siemens 6.0–120 turbines come ashore

NEPC Windfall · p.33

77%

of rare-earth refining projected to occur in China

NEPC Windfall · p.5

£25 / kg

remanufactured, against £85–100 / kg new — a 70% discount

NEPC Windfall · p.5

02Why it matters

Build the supply chain now, or lose it to decisions made today.

A · Supply

Demand outruns extraction

By 2030 the world needs 83 kt of rare earths a year for magnets and extracts 50 kt. One 12 MW turbine holds enough magnet for ~12,000 EV motors.

B · Carbon

Mining costs the planet

A new kilogram of magnet emits ~75 kg CO₂e and leaves toxic, radioactive tailings. Reusing one avoids nearly all of it.

C · Value

Remade, not shredded

Remanufactured magnets cost ~£25/kg against £85–100 new. Shredded for scrap, the same £620m stock is worth about £45m.

04Tracks

Four intervention points. Pick one.

Each track is one of the four points in the NEPC report where the windfall is won or lost. Every track has a way in for every discipline, and a Foundation and a Stretch brief, so a graduate and a Fellow can both find their problem.

P1Turbine design & assembly

Get the magnets out intact

If magnets come out intact, the UK keeps ~1 million kg a year of remanufacturable magnet from 2038 — worth £25/kg remade, rather than £45m in total as shredded scrap.

P2Planning, data & condition

Map every magnet at sea

With an open, trusted record of every magnet from installation to decommissioning, investors can size recovery capacity for the 1,400 t peak in 2044 rather than guess at it.

P3Decommissioning & logistics

Bring turbines ashore

A cheaper, faster, magnet-safe way to bring turbines ashore means ~1 million kg of magnet a year lands in the UK from 2038 — instead of being stranded at sea or shipped abroad.

P4Remanufacture & market

Keep magnets in the UK

One 1.85 kg turbine block becomes 108 EV rotor magnets. At ~1 million kg a year that could supply UK traction motors at £25/kg instead of £85–100/kg — and avoid 75 kg CO₂e for every kg not newly made.

P1Turbine design & assembly

Get the magnets out intact

Take apart

1Generator caseEncloses the rotor and stator.
2Coils28 stator coils around the rotor.
3Outer magnet ring28 NdFeB segments.
4Inner magnet ring28 NdFeB segments.
5Rotor coreCarries both magnet rings.
6Generator frameMounts it to the bedplate.

From the turbine CAD (GrabCAD / Onshape): a geared machine with a generator about 0.6 m across. Offshore direct-drive rotors are 6–10 m across, with magnets in 600 × 120 × 50 mm carriers (NEPC report p.34) — the same problem, much bigger.

A generator's magnets sit as a ring of segments on its rotor — and turbines going to sea today were never designed to give them back. Some segments are bolted on in carriers that can be unbolted from outside; others are bonded with epoxy. Every one is still strongly magnetised when the turbine comes ashore.

If you solve this

If magnets come out intact, the UK keeps ~1 million kg a year of remanufacturable magnet from 2038 — worth £25/kg remade, rather than £45m in total as shredded scrap.

How should a magnet segment fix to the rotor so it comes off intact — fixings, coatings, adhesives?
What would a whole rotor designed for disassembly look like, and in what order would it come apart?
What about the 16 million kg already going to sea that will never be redesigned?

Know before you start · 5Constraints — optional

Realities of the problem, not rules. Whether you take any on is your call — scope your own challenge, and if you do, against two at most.

Energised magnets snap together
Magnets stay powerfully magnetised in service. Pulled out carelessly they 'rapidly come together' with ferrous parts or each other — extraction needs a lead-screw tool, nylon guides and a controlled environment.
NEPC Windfall report, p.34 ↗
Demagnetising has trade-offs
Heat (~200–220 °C, held for a long time) works, but the coating and fixings must survive it. A decreasing alternating field — a degaussing coil or bulk demagnetiser — scrambles the domains with no heat. A strong opposing field cancels the polarity. Mechanical shock also disorders the domains, but cracks brittle sintered magnets: poor for reuse.
DTU — direct reuse of magnets: coating integrity ↗
Some magnets are glued, not bolted
Bolted carriers can be unbolted from outside the rotor; epoxy-bonded magnets cannot. Recovery from bonded designs is slower, riskier and loses more material.
NEPC Windfall report, pp.6, 24 ↗
Sintered NdFeB is brittle
Magnets crack and chip at edges and corners. A thin cover and semi-cured polymer exist to catch fragments before they reach bearings and seals — any redesign must still do that.
NEPC Windfall report, p.24 ↗
It must still last 25 years at sea
Corrosion, 0–20 Hz magnetic fields, short-circuit faults and vibration for a whole service life. A design that comes apart easily cannot come apart early.
NEPC Windfall report, pp.24, 38 ↗

Foundation

Redesign how one magnet segment fixes to the rotor (model on this page) so a technician can remove it safely and intact. Sketch it and justify each change.

Stretch

Redesign the whole magnet rotor for disassembly — and the retrofit or extraction kit that gets magnets out of the rotors already at sea. Something a company could build and sell from 2038.

Ways in, by discipline · 5

Mechanical: Fixings, extraction tooling, fatigue and handling loads.
Materials: Adhesives that release on demand; coatings that survive 25 years at sea.
Electrical: Safe demagnetisation, airgap tolerances, generator topology.
Design: Design-for-disassembly principles applied to a 25 kg magnetic part.
Software / data: Simulate extraction forces; work from the generator CAD.

What good looks like

Not a finished solution — an interesting, credible idea, shown in a demo video (2 min max) or a single slide.

A specific mechanism or design change, sketched or modelled — not a principle.
A rough number: time per segment, force handled, or kg recovered intact.
One honest risk, and how you'd test it first.

Resources · P1 · 10

CAD

Guide

How are wind turbines made? — Union of Concerned Scientists Plain-English walk through foundations, towers, nacelles and blades.

Paper

Industry

Wind turbine module magnetic assembly — MPCO How a magnet manufacturer builds and protects generator magnet modules.

Standard

BS 8887 — design for disassembly and end-of-life The British Standard the report points to.

P2Planning, data & condition

Map every magnet at sea

Nobody holds granular data on how much magnet is in which turbine, who owns it, or what condition it will be in. The report's authors had to buy a commercial database and reconstruct magnet mass from spare-parts enquiries. Without that data there is no business case for recovery, and no way to tell whether a magnet can be remanufactured or must be recycled.

If you solve this

With an open, trusted record of every magnet from installation to decommissioning, investors can size recovery capacity for the 1,400 t peak in 2044 rather than guess at it.

How far can you get with open data alone — and where exactly does it run out?
What is the minimum record a magnet 'passport' must carry, and who keeps it up to date?
Which operational signals (temperature, faults, short-circuits) could indicate magnet condition?

Know before you start · 6Constraints — optional

Realities of the problem, not rules. Whether you take any on is your call — scope your own challenge, and if you do, against two at most.

There is no public magnet register
The UK planning database lists projects, turbine counts and turbine ratings — but not turbine models or magnet mass. The report's authors had to buy a commercial database.
DESNZ — Renewable Energy Planning Database ↗
The model changes everything
Magnet per MW ranges from 0 kg (geared DFIG turbines) to 800–900 kg (direct-drive). Without the turbine model, a capacity figure tells you almost nothing.
NEPC Windfall report, Annex A, p.43 ↗
Owners change over 25 years
Arrays are sold and refinanced; decommissioning is subcontracted. Ownership of the magnets at each handover is unclear.
NEPC Windfall report, p.38 ↗
Stock data is commercially sensitive
Operators won't publish fleet details freely. Sharing needs a trusted data architecture and industry collaboration on what is shared, with whom.
NEPC Windfall report, pp.10–11 ↗
Nobody monitors the magnets
Turbines log plenty of operational data, but nothing magnet-specific. Whether a magnet can be reused or must be recycled is unknown until it's out.
NEPC Windfall report, p.11 ↗
Decommissioning dates move
The report assumes commissioning + 25 years, but life extension, repowering and lease terms shift real dates by years — and the magnet forecast with them.
NEPC Windfall report, pp.12, 27–28, 42 ↗

Foundation

Design the passport for one magnet carrier: fields, owner, update points, and how it's tagged. Use the Asset Administration Shell or DecomBlades passport as a template.

Stretch

Build the open alternative to the data the report had to buy: join REPD, EMODnet and GEM, apply the report's kg/MW table, and forecast magnet tonnes landing by year and port.

Ways in, by discipline · 5

Software / data: Join open datasets; schemas, registries, APIs.
Electrical: Condition signals in SCADA that hint at magnet health.
Systems: Fleet stock-and-flow, 2030–2050, from kg/MW and decommissioning dates.
Chemical / materials: Grade, coating and composition fields that decide the recycling route.
Design / UX: Who reads the passport — technician, investor, regulator — and what do they need?

What good looks like

Not a finished solution — an interesting, credible idea, shown in a demo video (2 min max) or a single slide.

A working join, map or schema built from real open data — even if partial.
A clear statement of the gap: which fields are missing and who holds them.
One decision your output would change, and for whom.

Resources · P2 · 10

Data

Framework

Paper

NEPC Windfall report — Annex A, pp.41–43 Methodology and the kg-of-magnet-per-MW table by turbine model.

P3Decommissioning & logistics

Bring turbines ashore

Sequence

01Rotor offBlades, hub and spinner lifted as one.
02Nacelle off — wholeGenerator and magnets stay sealed inside.
03Tower, top downBolted sections, one lift each.
04Transition piecePlatform and grouted connection.
05Monopile cut~1.5 m below the seabed; the stub stays.

The startupYou are founding the company that brings turbines home. Decommissioning costs ~£285/kW and vessels are 60–80% of that. Design the service, tool or vessel concept that wins the contract — and lands the magnets intact.

From 2038, hundreds of offshore turbines a year reach end of life. Each must be taken apart at sea, in weather windows, with expensive jack-up vessels — and the generator, full of energised magnets, has to reach shore undamaged. Today nobody is set up to do this at scale, and the 2042–2047 peak will compete with oil and gas decommissioning for the same ports and people.

If you solve this

A cheaper, faster, magnet-safe way to bring turbines ashore means ~1 million kg of magnet a year lands in the UK from 2038 — instead of being stranded at sea or shipped abroad.

Where in the sequence below is the cost, the risk, or the lost magnet — and what would remove it?
What could be done on land instead of at sea, or in parallel instead of in series?
What would you sell, to whom, and what would they pay per turbine?

Know before you start · 6Constraints — optional

Realities of the problem, not rules. Whether you take any on is your call — scope your own challenge, and if you do, against two at most.

It's installation in reverse
Blades (or the whole rotor) come off first, then nacelle, tower, transition piece and foundation; array cables are removed before the foundation. Each lift depends on the one before — you can't reach the nacelle with blades attached overhead, or the foundation with the tower on it.
ORE Catapult — End-of-life planning in offshore wind ↗
Don't open the generator at sea
Energised magnet carriers must be extracted in a controlled environment with a lead-screw tool and nylon guides: magnets 'rapidly coming together' is a serious hazard. So the nacelle and generator come ashore whole, and must not be damaged on the way (allowance: ~5% magnet loss).
NEPC Windfall report, pp.34, 43 ↗
Vessel time is the cost
Vessels are 60–80% of decommissioning cost. Modelled for a 400 MW farm: ~36 h of large jack-up per turbine and ~24 h per foundation; ~£285/kW in total, of which 40% is offshore preparation and 35% foundation removal.
ORE Catapult (Table 1) and BEIS/Arup cost study, 2018 ↗
Weather sets the calendar
Wave-height limits drive project length: in a North Sea study, raising the operable wave height by 1 m cut average duration by ~28%. Jack-ups have lower weather thresholds than larger vessels, and starting late in summer makes things worse.
Seasonality of OWF decommissioning — Renewable & Sustainable Energy Reviews, 2025 ↗
Regulation decides what stays in the seabed
A decommissioning programme is required under the Energy Act 2004, and full removal is the default — but no proven method fully removes a monopile, so piles are cut ~1–2 m below the seabed and the stub stays.
GOV.UK — Decommissioning offshore renewable energy installations ↗
Ports and people are finite
Recovery peaks 2042–2047 (1,400 t of magnet in 2044) — at the same time as oil and gas decommissioning. Port space, storage yards and skills are named bottlenecks.
NEPC Windfall report, pp.12–13, 31 ↗

Foundation

Take the sequence above for one 6 MW turbine. Pick the single step that costs the most or risks the magnets most, and design a better way to do it.

Stretch

Design the company: the service or kit, the unit economics per turbine against ~£285/kW, the first customer, and what you'd need from a port.

Ways in, by discipline · 5

Civil / marine: Vessels, lifts, cutting, partial vs full removal.
Mechanical: Rigging, nacelle separation, generator transport frames.
Materials: Blades, tower steel and everything that isn't magnet — where does it go?
Operations: Scheduling against weather and the 2042–2047 peak; port logistics.
Business: Who pays, per what unit, and why they'd switch to you.

What good looks like

Not a finished solution — an interesting, credible idea, shown in a demo video (2 min max) or a single slide.

Clearly names which constraint above it attacks.
A rough before/after number: hours, vessel days, £ or kg of magnet saved.
Who the customer is and why they'd pay.

Resources · P3 · 8

CAD

Guide

End-of-life planning in offshore wind — ORE Catapult Costs, vessel types, hours per turbine, and the end-of-life options compared.

Paper

Policy

Decommissioning guidance (Energy Act 2004) — GOV.UK What a decommissioning programme must contain.

P4Remanufacture & market

Keep magnets in the UK

Remanufactured magnets would be a new product. There is no accepted quality-assurance process to give buyers in automotive or aerospace confidence, and the UK has almost no magnet-making capacity. Without both, recovered magnets will simply be sold abroad instead of supplying UK manufacturing.

If you solve this

One 1.85 kg turbine block becomes 108 EV rotor magnets. At ~1 million kg a year that could supply UK traction motors at £25/kg instead of £85–100/kg — and avoid 75 kg CO₂e for every kg not newly made.

What would a UK buyer need to see before putting a reused magnet in a car, train or aircraft?
Which UK product should be first to use them — and why that one?
Reuse, short-loop (hydrogen) or long-loop (chemical) recycling — which for which magnet?

Know before you start · 6Constraints — optional

Realities of the problem, not rules. Whether you take any on is your call — scope your own challenge, and if you do, against two at most.

No quality standard exists yet
A remanufactured magnet is a new product. There is no accepted QA process that would give a buyer confidence in its performance.
NEPC Windfall report, pp.12–13 ↗
Safety-critical buyers need traceability
Automotive and aerospace require evidence of compliance — serialisation from the turbine's serial number through every process step.
NEPC Windfall report, p.37 ↗
Only high grades are useful
Traction and aerospace motors need high-energy magnets — 'certainly N38 and higher'. Degraded or low-grade blocks go to recycling instead.
NEPC Windfall report, p.37 ↗
The geometry is wrong
A 100 × 100 × 25 mm turbine block must be cut into ~108 small EV magnets, leaving ~0.3 kg of waste per block for recycling.
NEPC Windfall report, p.36 ↗
It has to beat new on price
New blocks cost £85–100/kg. Remanufacturing, restoration and testing add ~£10–15/kg, for a sale price of £20–30/kg.
NEPC Windfall report, p.38 ↗
The world wants these magnets too
Sold on global markets, recovered magnets will likely leave the UK. The EU's Critical Raw Materials Act and possible export restrictions pull the other way.
NEPC Windfall report, pp.13, 16 ↗

PrimerHow a sintered NdFeB magnet is made — and where reuse cuts in

01Alloy
Nd, Fe, B (+ Dy/Tb) melted and strip-cast into ~0.3 mm flakes.
02Powder
Hydrogen breaks the flakes up; jet-milled to 2–5 µm.
03Press
Powder aligned in a magnetic field and pressed to shape.
04Sinter
Vacuum-sintered at ~1,000–1,100 °C to full density, then annealed.
05Finish
Machined or wire-cut, coated (Ni, epoxy), tested and magnetised.

Reuse (the report's route): demagnetise at ~220 °C → remove from carrier → wire-cut → re-coat → test → remagnetise — skipping every step above except Finish. Short-loop recycling (HPMS, Birmingham): hydrogen turns old magnets to powder → re-press and re-sinter, skipping Alloy. Long-loop: dissolve and separate the rare earths chemically (Ionic, Belfast) → back to Alloy.

Foundation

Lay out the remanufacturing line, from carrier to certified EV magnet: steps, tests, labels and data at each stage. Use the primer above as your starting point.

Stretch

Design the product and the deal: the first UK application, the test evidence a buyer needs, and the offtake that keeps the magnets here rather than on the export market.

Ways in, by discipline · 5

Materials: Degradation, cracking, re-coating, grade verification.
Chemical: Recycling routes for blocks that fail remanufacture.
Manufacturing: Wire-cutting 108 parts per block: yield, kerf, automation.
Electrical machines: Design a motor around a known, reused magnet geometry.
Business: Offtake, pricing, and why a UK buyer beats an export bid.

What good looks like

Not a finished solution — an interesting, credible idea, shown in a demo video (2 min max) or a single slide.

Picks a route (reuse, short-loop or long-loop) and says why for this magnet.
Names a real first buyer or product, and what they'd need to see.
A cost or carbon number against new magnets.

Resources · P4 · 12

CAD

Guide

Sintered NdFeB magnet manufacturing process flow — Advanced Magnets Step-by-step from alloy to finished magnet.

Paper

Industry

Policy

UK Critical Minerals Strategy — Vision 2035 Target: 20% of critical mineral demand met by recycling by 2035.

05Schedule

Two hours, end to end.

T+0:00 – T+0:10

Brief

The windfall, the four tracks, the rules. Teams lock a track.

T+0:10 – T+1:30

Build

80 minutes. Design, model, prototype or argue. Submit through the form before T+1:30.

T+1:30 – T+1:40

Judging

Judges review every submission against the criteria.

T+1:40 – T+1:50

Close

What we learned, and the top teams announced.

T+1:50 – T+2:00

Four teams present

One team per track. Two minutes each.

06How it works

What we are looking for: not a finished product. An interesting idea.

We are not looking for finished solutions. We want interesting, credible ideas — ideally shown in a demo video or a single slide. The best team from each track pitches live for two minutes.

Rules

01: Teams of 3–5. Mix disciplines and seniority.
02: Any tools are allowed, including AI.
03: Submit through the form by T+1:30. Late entries are not judged.

What to submit

01: Ideally a demo video (2 minutes max) or a single slide — repos, sketches and sheets also welcome
02: Which track, and the problem in one line
03: Your idea, and why it would work

07Judging

What we are judging on.

01

Impact

If this worked, how much of the windfall would it keep in the UK — and how soon?

60%

02

Evidence

Is it backed by data, a calculation, a test or a credible source?

40%

Judge · lead author of the NEPC Windfall reportDr Stuart BradleyWMG, University of Warwick

The winning team is recognised by the Royal Academy of Engineering and Engineering Excellence.

09Resources

Start here.

Reading & data

National Engineering Policy Centre · May 2026

Hosts

Keyne WalkerRoyal Academy of Engineering

Sebastiano ZuddasEngineering Excellence

Shaurya SinghEngineering Excellence

A national stock of critical material, with a date on it.

Build the supply chain now, or lose it to decisions made today.

Demand outruns extraction

Mining costs the planet

Remade, not shredded

Never designed to come apart.

Four intervention points. Pick one.

Get the magnets out intact

Map every magnet at sea

Bring turbines ashore

Keep magnets in the UK

Get the magnets out intact

Map every magnet at sea

Bring turbines ashore

Keep magnets in the UK

Two hours, end to end.

Brief

Build

Judging

Close

Four teams present

What we are looking for: not a finished product. An interesting idea.

Rules

What to submit

What we are judging on.

Impact

Evidence

Submit by T+1:30.

Start here.

Reading & data

Hosts