Who this is for
Job shops with a changing mix, manufacturers scaling repeat work, teams short of setters, and businesses integrating cutting, forming, welding and finishing into one flow. It reads across our metal fabrication pages so you can find your row and go to the page that scopes it.
The conditions that apply to every fabrication floor
| Condition | What it means for the cell | What to have in hand |
|---|---|---|
| Power quality | Lasers, presses and drives on sites with poor power factor trip and age early; power-factor correction and variable-speed drives are specified from the start | Supply assessment; PFC and VFD scope in the electrical package; backup decision with the power guide |
| Setter skills | Skilled press and laser setters are scarce; recipe storage on the HMI and machine-taught set-ups replace the setter's memory | Controls specified for recipe storage; training plan for the operators you have |
| Wear parts and consumables | Punches, dies, gas lenses and laser optics are imported and expensive to wait for; local stock decides uptime | Consumables list priced with the cell; local stock holding agreed |
| Safety auditing | Light curtains, two-hand control, emergency stops and guarding are audited more often; a cell that fails an audit stops | Guarding and interlocks in the specification, not the retrofit |
| Gas and air | Fibre lasers need nitrogen or oxygen at pressure and volume; welding needs shielding gas; both need a supply contract and on-site storage | Gas supply contract and a compressed-air and nitrogen scope from a local dealer |
| Import duty and VAT | Most machine tools and presses under HS 84 enter at 0% duty-free; the exceptions to plan for are refrigeration equipment, some motors and large generating sets; import VAT is 15% on the customs value plus a 10% uplift plus duty, recoverable for a VAT-registered manufacturer | Eight-digit tariff codes; run the landed cost estimator |
The scenarios, one by one
Scenario 1: a first laser and press-brake cell
The most common step for a job shop: a fibre laser cutter and a CNC press-brake, imported as a matched pair, with sheet handling, extraction and gas supply sourced locally. Laser power and bed size follow the material and part mix; press-brake tonnage and length follow the longest bend. Press lines and CNC: local versus import sizes both machines and sets out the controls decision.
The matched pair is where job shops most often mis-specify. A laser sized for the thickest sheet you cut twice a year spends the rest of its life cutting thin material slowly; a press-brake chosen on tonnage alone will not take the longest part that laser can produce, so nested blanks queue in front of a machine that cannot bend them. Size the laser on the weekly mix, the brake on the longest bend, and check the two agree before either is ordered.
Assist gas is the running cost, not electricity. Nitrogen gives an oxide-free edge that needs no preparation before welding or powder coating and is consumed in volume; oxygen is cheaper and faster on mild steel but leaves a layer somebody has to remove. On a site with poor power factor the cell also needs correction and drives that ride through a dip — the alternative is nuisance trips and premature ageing on exactly the machines you cannot afford to lose.
Scenario 2: a press line for repeat parts
Mechanical or hydraulic presses with tooling for a repeat product, where the setter constraint bites hardest. Recipe storage, quick tool change and a tooling maintenance plan matter more than the press tonnage number. The press line case study shows a cell delivered in twenty weeks from order to first cutting hour, and which elements came from where.
Tonnage is the headline number and the least useful one. A press is sized by the force the part needs; a press line is limited by how fast a die set can be changed and how reliably the same set-up comes back after the change. In a market where skilled setters are scarce and mobile, recipe storage on the control and quick-change clamping are worth more than extra tonnage, because they move the set-up out of one person's memory and into the machine.
The tooling, not the press, is the asset that wears. Punches and dies are consumables on a repeat-part line, they are imported, and a die that is out of the country when it cracks stops the line as surely as a failed motor. Treat the tooling schedule as capital and stock the wear items deliberately — our spare parts strategy page sets out the same logic by category and cost band.
Scenario 3: welding cells, manual to robotic
Welding power sources come from a local dealer for service and consumables; positioners, jigs and fixtures are built locally; a robotic welding cell is imported when volume and repeatability justify it. The scope split is the decision, and the case study's element-by-element table is the template.
The case for a robot is volume multiplied by repeatability, not volume alone. A cell running one weldment in long batches pays for itself in consistency and in relief from the artisan shortage; a shop welding a different assembly every week spends more time building fixtures and teaching paths than it saves, and the same money buys better positioners and a second competent welder. The honest test is whether the part family will still exist in three years.
Power sources stay with a local dealer for an unglamorous reason: wire, tips, nozzles, gas and warranty labour all come through that dealer, and a torch consumable on air freight is a stopped cell. Positioners, jigs and fixtures are built locally because they are steel work tied to your specific part — and a fixture that needs modifying is modified next door, not next quarter.
Scenario 4: tube and pipe processing
Tube lasers, benders and end-formers are specialist imports; the handling and the fixtures are local. The part family decides whether a tube laser replaces a saw and a drill or adds to them.
Tube work divides cleanly. A small, stable family of round or square sections with simple end conditions is well served by a saw, a drill and a good bender at a fraction of the capital. Parts carrying slots, holes, copes and intersections — handrails, frames, trailer and agricultural structures — are where a tube laser collapses several operations into one and removes the marking-out error that cascades into the weld fixture. Count operations per part; that is the decision.
The local half is larger than it looks. Long stock has to be stored, fed and taken away, and the floor has to allow a full length to swing. Handling and fixtures are built here for the same reason the structural scope is: heavy, low value per kilogram, and specific to your parts.
Scenario 5: structural steel and bulk fabrication
Frames, platforms, cranes and heavy structural scope are freight-uneconomic to import and are built in South Africa by a fabricator under an integration owner. This is the localise-the-bulk half of the rule, and it is also where local-content scoring is earned. Local and hybrid manufacturing sets out the three build paths and their timelines.
The freight arithmetic settles this one. A tonne of imported precision machinery carries a great deal of engineering per kilogram; a tonne of fabricated platform carries almost none, so paying sea freight and inland transport from Durban to move it is money spent moving steel that South African fabricators cut and weld every day. There is a schedule cost too: a frame that is wrong is re-cut locally in days and re-cut offshore in months.
This is also where local-content preferences are earned rather than claimed. The scope that counts is normally the fabricated, installed and engineered portion — precisely the portion that is uneconomic to import. One integration owner has to carry it, or the saving is spent on the gap between a local fabricator's drawing and the imported machine's real interface. Our B-BBEE and local content page sets out how the split moves the score.
Scenario 6: finishing, assembly and inspection
Surface treatment and powder coating lines are largely local scope; assembly automation and vision inspection are specified per product. Where a customer demands a hundred percent inspection record, machine vision inspection answers it at a published cost band.
Finishing is where a fabricated part becomes a sellable one, and it is almost entirely local scope: pre-treatment, powder coating or paint, and the oven. The variable that decides quality is the pre-treatment, not the powder — coating over poor preparation fails in the field and is blamed on the coating. Specify the process, then the line.
Inspection is specified per product and per customer. Where a buyer demands a hundred per cent record rather than a sample, machine vision becomes part of the cell rather than an addition to it, and the record it produces is often the contractual deliverable. Guarding belongs in the same conversation: light curtains, two-hand control and emergency-stop categories are audited here more often than elsewhere, and a cell that fails an audit stops earning while it is corrected.
Scenario 7: recovering an old cell, or a tender-driven expansion
An older press with a dead controller is often a control retrofit at US$2–8k per machine rather than a replacement, and a measurement layer shows whether the constraint is the machine, the set-up or the material flow. Where the buyer scores local content, the structural and handling scope is where the points are. See new, used or recover, when to upgrade your PLC, OEE on an old line and local content scoring.
A control retrofit is the highest-return move in this whole guide, and it is routinely skipped because it is unglamorous. An older press with a sound frame and a dead controller is mechanically fine; what has expired is the electronics and the documentation. Replacing the control at US$2–8k per machine returns a machine that can hold a recipe, which is the same benefit the new press was being bought for.
Measure before you spend. A measurement layer on the existing cell shows whether the constraint is the machine, the set-up or the material flow, and in tender-driven expansions it also produces the evidence a funder or a customer asks for. It is common for the answer to be set-up time rather than machine capacity, and that answer costs a fraction of a new cell to act on.
The map on one table
| Scenario | Import | Build or buy locally | Floor must have | Scopes it |
|---|---|---|---|---|
| First laser and press-brake cell | Fibre laser, CNC press-brake | Sheet handling, extraction, gas and air supply, MCC | PFC and drives, nitrogen supply, recipe-storing controls | Press and CNC guide |
| Press line for repeat parts | Presses, tooling | Feeders, jigs, guarding integration | Quick tool change, tooling spares, safety interlocks | Press line case study |
| Welding cells | Robotic cell when volume justifies | Power sources from a local dealer, positioners, jigs | Shielding gas supply, extraction | Press line case study |
| Tube and pipe processing | Tube laser, benders, end-formers | Handling, fixtures | Part family defined before the machine | Press and CNC guide |
| Structural steel and bulk scope | Nothing that is heavy and low value per kilogram | Frames, platforms, cranes, conveyors | An integration owner across local and imported scope | Local and hybrid manufacturing |
| Finishing, assembly, inspection | Vision systems, automation modules per product | Powder coating and surface treatment lines | Inspection record the customer will audit | Machine vision cost guide |
| Recover an old cell / tender-driven | Controls and drives for a retrofit | Structural and handling scope for local-content points | Drawings, PLC access; scoring rules | New, used or recover |
| Wire products: nails, mesh, brick force, razor wire | Drawing line plus product machines | Galvanising, usually subcontracted to a licensed galvaniser | Wire-rod supply, floor space for coil handling | Wire machine prices |
| Purlins, light steel frame, window frames | Roll formers, LGSF machine with software | Decoilers, tables, coating partner | Coil supply and the current coil duties | Roll forming prices |
Rows added October 2026 price new lines in their own guides; comparing across industries? See which production line to start in South Africa.
Precision machines are quoted per specification in US dollars and the rand cost moves with the exchange rate; a full-import cell runs 30–36 weeks from design to commissioning, a hybrid cell 22–26 weeks and a fully local build 20–22 weeks, on the paths set out in our local manufacturing page.
What CISH does with your row
The scoping call starts from the part mix, splits the cell into import and local scope element by element, sizes the precision machines, and hands you a written scope with one integration owner. Delivery is the same for every row: turnkey delivery commissioned on your parts, operators and setters trained during run-up, drawings, PLC code and consumables lists handed over.
Frequently asked questions
Why do you not publish prices for metal fabrication cells?
Because a cell is a laser or press specified to your material and part mix plus a local scope that is quoted by South African fabricators and dealers; a published band would mislead in both directions. We publish the scope split and the machine sizing rules instead, and quote the pair for your parts.
Can a local fabricator build the whole cell?
The bulk, yes: frames, handling, cranes, extraction, MCCs and integration. The precision machines, fibre lasers, CNC press-brakes, machining centres and robotic cells, are mature imported product classes that local fabrication cannot replicate at the price or the performance. The rule is import the precision, localise the bulk.
What laser power and press-brake tonnage do I need?
Laser power follows the thickest material you cut regularly and the bed size follows your sheet; press-brake tonnage follows the longest and thickest bend. Oversizing both is the common mistake and costs power and floor space for years. The press and CNC guide gives the sizing rules.
How long does a fabrication cell take to deliver?
A full-import cell runs about 30–36 weeks from design to commissioning, a hybrid cell 22–26 weeks and a fully local build 20–22 weeks; the press line case study delivered in twenty weeks from order to first cutting hour with a hybrid scope.
What do I do about scarce setters?
Specify controls with recipe storage so the machine remembers the set-up, teach set-ups once with the setter you have, and train operators during run-up. It is a controls decision made at purchase, not a hiring problem solved later.
Should I refurbish my old presses first?
Often, yes: a control retrofit at US$2–8k per machine and a measurement layer show whether the constraint is the press or the set-up around it. Refurbishment fails when the frame, the ram guides or the safety system are past saving; the case study explains why that line chose new precision machines.