How to start an aluminium extrusion plant: building, machines, budget, timeline

Plenty of people can tell you that aluminium extrusion is a good business. Almost nobody will tell you what starting a plant actually involves: how big the hall has to be, what the utilities bill looks like, which people you cannot run the line without, and how long the road from signed contract to first invoiced tonne really is.

This guide is that missing part. It is written for an investor planning a greenfield plant — whether that is a window fabricator in Central Europe tired of buying profiles, or a Gulf investor building manufacturing capacity under a localisation programme. The numbers are indicative ranges, not quotations, but they are realistic enough to build a board-level investment case on. We use "aluminium" throughout; everything applies equally if you searched for how to start an aluminum extrusion plant.

Step zero: is there room for you?

Extrusion is a regional business. Profiles are long, light for their volume and easily damaged, so freight economics favour plants within roughly a day's truck ride of their customers. That is why the market keeps supporting new mid-size entrants despite the presence of giants.

Three demand drivers are worth naming honestly:

• In the EU, anti-dumping duties of roughly 22–32% on imported Chinese aluminium profiles have made local production structurally more attractive. Those duties apply to profiles, not to machinery — which is precisely why importing the machine and extruding locally works as a strategy. The European extrusion market is of the order of $20 billion a year and has been growing at mid-single digits.
• In the Gulf, construction pipelines under Vision 2030 and its regional equivalents are pulling in facade, fenestration and industrial profiles at a pace local capacity has not matched, and industrial localisation programmes actively favour in-country manufacturing over imports.
• Everywhere, solar mounting structures and PV frames have created a profile segment that barely existed fifteen years ago and suits smaller presses well.

The strategic question is not "should I extrude" but "what will my first 200 tonnes a month be". The generic architectural systems market is the most crowded corner; first-time extruders tend to do better anchored to captive or semi-captive demand — their own fabrication business, a fencing or furniture product line, PV structures, or a handful of industrial customers with drawings in hand. Decide this before you size anything, because the profile mix drives every choice that follows.

Choosing the press tonnage

The press determines your billet diameter, your maximum profile size and most of your budget. The practical shortlist for a first plant is short:

• 1000 tonnes (5″ billet) — covers the large majority of architectural profiles: window and door frames, insect screens, furniture, light PV frames. The most common first press for a reason; see the 1000-tonne press for full specifications. Realistic output around 250–350 t/month on two shifts.
• 1800 tonnes (7″ billet) — the workhorse of mid-size plants: full window, door and facade systems plus mid-section industrial profiles, at roughly 450–650 t/month on two shifts. Details on the 1800-tonne press page.
• 2500 tonnes and above — large industrial and structural sections. Rarely the right first machine unless your order book demands it; the full range from 530 to above 5,000 tonnes, including servo drives, is covered on our extrusion presses category page.

Two sizing rules save expensive regret. First, size for the widest profile you credibly expect to sell in three years, not for your launch product alone — the jump from 5″ to 7″ billet cannot be retrofitted. Second, do not oversize "to be safe": a 1800 t press running 40% loaded burns energy and capital a 1000 t press would not, and small-batch die changes on a big press hurt productivity. Send prospective suppliers your actual profile drawings and let the tonnage be calculated, not guessed.

The building: what an extrusion hall actually needs

This is where greenfield projects most often go wrong, because an extrusion line is long. Billets go in one end; heated, pressed, pulled, cooled, stretched, cut and aged profiles come out the other — and profiles up to 13 m long need run-out and handling length to match. Indicative requirements:

Requirement	1000 t line	1800 t line
Line length (billet furnace → press → run-out → stretcher → saw)	approx. 60–80 m	approx. 80–110 m
Practical minimum hall	approx. 70–90 × 15–18 m	approx. 90–120 × 18–24 m
Clear height under crane hook	5–6 m	6–7 m
Overhead crane	5–10 t	10–20 t
Press foundation	reinforced pit, approx. 1.5–2 m deep	reinforced pit, approx. 2–2.5 m deep
Connected electrical load (whole line)	approx. 0.4–0.6 MW	approx. 0.7–1.0 MW
Gas	medium-pressure connection for billet furnace and aging oven	same, higher draw

Treat these as planning figures — the supplier's layout drawing for your specific configuration is the document your architect should work from, and a serious supplier provides it free at quotation stage.

Beyond the headline dimensions, budget floor space for what beginners forget: the aging oven with its loading baskets (an 18-basket oven chamber alone is nearly 14 m long), a die shop with die oven and correction bench, billet storage, finished-goods and packing areas, and a logistics aisle wide enough for 13 m profiles to be moved without acrobatics. On utilities: the press wants a stable electrical supply (an 1800 t press alone carries three 110 kW main motors), the furnaces run on natural gas or LPG with diesel and electric as fallbacks, and you will need compressed air and a closed cooling-water circuit for the hydraulic oil cooler. Details of the furnace family — billet heating with hot log shear, aging, die ovens — are on the industrial furnaces category page.

One forward-looking note: if you intend to add surface finishing later, plan for it now. A vertical powder coating line needs an 8.5–9 m tall building — a decision much cheaper to make at the architectural drawing stage than after the roof is on.

Aluminium extrusion plant setup cost: the realistic budget

We publish indicative ranges rather than "price on request", so here is the machinery side of the budget, EXW China, consistent with our full extrusion line cost breakdown:

Budget item	1000 t line	1800 t line
Extrusion press	€55–85k	€95–160k
Billet heating furnace + die oven	€53–115k	€60–130k
Handling: run-out and cooling tables, puller, stretcher, saw	€60–120k	€80–160k
Aging oven	€40–85k	€50–95k
Packing equipment	€10–40k	€15–45k
Equipment total, EXW	approx. €250–450k	approx. €350–650k

On top of the EXW figure, plan 15–30% for freight, import duty (typically 0–4.5% into the EU on machinery — the anti-dumping duties apply to profiles, not machines), EU conformity work, foundations and commissioning. VAT is a cash-flow item for a registered business, not a cost. Gulf buyers typically face a 5% common external tariff and 5–15% VAT depending on the country, with shorter sea routes cutting freight time by several weeks.

Then the positions that never appear in machinery quotations but decide whether you survive year one:

• Dies. A starting library of 20–50 dies is a five-figure position, and dies are a recurring cost, not capex.
• Working capital. Billet stock, energy, wages and three to six months of ramp-up losses while your recovery rate climbs. A mature plant turns roughly 80–85% of purchased billet into saleable profile; in the first months expect meaningfully worse. Under-capitalised ramp-ups kill more new extruders than machine problems do.
• Building. Purchase, construction or adaptation of the hall is on top of everything above and varies too much by market for a useful global range.

All told, a complete first extrusion plant — machines, landing, site works, dies and sensible working capital, before the building itself — typically lands in the €0.35–1.8 million span, depending on tonnage and how much finishing you add on day one.

Permits and conformity: EU and Gulf, briefly

In the EU, the building side means a construction or change-of-use permit, environmental notifications (the furnaces are combustion installations), grid connection agreements and a gas connection — start these at contract signature, not at delivery, because in most member states they comfortably consume the machine's production time. The machinery side means CE conformity: Chinese presses and furnaces do not normally arrive EU-compliant regardless of what the sticker says, guarding and safety-circuit retrofits are a real budgeted project, and from January 2027 the new Machinery Regulation puts explicit duties on whoever acts as importer. We keep the full, honest treatment — including what paperwork to demand and what the three realistic CE statuses are — in our CE marking guide.

In the Gulf, there is no CE requirement, but do not read that as "no requirements": Saudi Arabia runs its own product conformity system (SASO/SABER) for imports, the UAE has its equivalent schemes, industrial licences and civil-defence approvals apply to the facility, and utility connections for a megawatt-class load need early engagement with the provider. Local rules move quickly — verify the current state with a local consultant rather than a year-old forum post.

Staffing: the part nobody budgets honestly

The machine list is the easy half. A two-shift, one-press plant realistically needs around 20–35 people, and three of those hires matter more than the rest combined:

• A press foreman who has done this before. Die trials, temperature discipline, speed and butt-length decisions — this experience cannot be trained from zero in your first year. Hiring one experienced foreman, at whatever it costs, is the single best productivity investment in the whole project.
• A die corrector. Every new die needs trial-and-correction cycles before it runs clean. Without this skill in-house you will be shipping dies back and forth while the press stands idle.
• A maintenance lead comfortable with hydraulics and PLCs. Your supplier is seven time zones away; your first line of defence is in your own hall — supported by a proper spare-parts package ordered with the line and remote-diagnostics-ready controls.

The per-shift press crew — furnace operator, press operator, puller and saw, stretcher pair, packing — runs six to ten people depending on automation level. Add quality control, dispatch and administration. In markets with no extrusion tradition (common in greenfield Gulf projects), budget for an experienced expatriate core crew for the first year or two and structured training for local hires; it is expensive and cheaper than the alternative.

The 12–18 month timeline

Nobody sells a serious press off the shelf, and that is fine — the machine's production time is exactly the window you need for everything else. A realistic sequence:

• Months 0–2: product and customer decisions, tonnage calculation from real profile drawings, building secured, machinery contract signed with FAT terms in writing.
• Months 2–8: press in production (18–28 weeks; auxiliary machines are much faster). In parallel — and this parallelism is the whole trick: permits, utility connections, foundations from the supplier's drawings, die orders, billet supply contracts, key hires.
• Months 8–10: factory acceptance test (attend it — this is your last cheap moment to fix anything), shipment, 6–8 weeks of sea freight to Europe, less to the Gulf.
• Months 10–13: installation and commissioning, typically 4–8 weeks with the manufacturer's crew, first extrusion trials.
• Months 13–18: ramp-up — recovery rates climbing, die library growing, first customer approvals and audits done.

A disciplined project with an existing building can compress towards twelve months; a full greenfield with new construction and slow permitting realistically uses all eighteen.

Five mistakes first-time extruders make

1. Buying the press before designing the plant. The line layout, aging oven, die shop and future finishing must fit the building — sequence is layout first, hall second, purchase order third.
2. Budgeting capex only. Dies, spare parts, billet stock and ramp-up losses are where under-funded projects actually die. If the working-capital line in your plan is a round guess, the plan is not finished.
3. Believing the CE sticker. A logo on a nameplate with no traceable declaration, issuing body and technical file is decoration. Demand the paperwork per machine and serial number, and price the conformity retrofit from day one.
4. Getting tonnage wrong in either direction. Undersized means turning away the facade order that would have made your year; oversized means paying energy and capital for force you never use. Calculate from drawings, not ambition.
5. Buying the line as separate bargains. A press from one factory, furnace from another and handling from a third saves single-digit percent and costs you mismatched interfaces, three warranty counterparties and a commissioning nobody owns. One supplier, one FAT, one schedule.

Where to start

The cheapest way to test your project is to make it concrete: a profile drawing (or just the product family and target monthly tonnage), your hall dimensions if you have them, and your market. Send those through our contact form and we will come back with a calculated tonnage, a line configuration with layout drawing, indicative pricing on the same open basis as the ranges above, and an honest delivery schedule — including the cost lines other sellers leave out of the first offer.