Single-Pass Corrugated Digital Printing vs Flexo Post-Print: Where the Crossover Sits

Every box plant that looks at a corrugated carton digital printer vs flexo post-print eventually asks the same question in the same words: at what run length does digital win? It is the right instinct and the wrong first question. Run length is only one of four variables that move the answer, and in most plants it is not the one that moves it furthest.

The other three are ink coverage on the artwork, the number of versions behind a single design, and — the one buyers skip — what your existing flexo line is actually costing you per job once you stop counting the capital you already spent. A plant with one design, heavy coverage and a half-idle flexo folder-gluer has a crossover somewhere down around a few hundred boxes. A plant with forty SKU versions of the same box and a flexo line running at 90% loading has a crossover an order of magnitude higher, in the other direction. Both plants are looking at the same two machines.

This is how to build the number for your own plant, and what to interrogate in the quotation before you believe anyone else's.

The two cost curves, and why only one of them has a step in it

Flexo post-print — printing directly on the corrugated board after it comes off the corrugator — carries a fixed cost that lands before the first good sheet:

• Plates. One photopolymer plate per colour, per design. A four-colour box needs four plates; a design change needs new plates; a small text change on one panel usually needs the whole plate for that colour remade.
• Plate mounting and proofing. Labour, mounting tape, and a mounter's time.
• Make-ready. Washing up the previous ink, changing anilox rolls if the screen count differs, setting impression, registering the decks.
• Waste board. Sheets consumed getting to colour and register. On a wide press this is the line that hurts.
• Ink matching. Spot colours drawn down and adjusted against a target.

Once you are past all of that, the flexo curve is nearly flat. The marginal cost of sheet number 5,001 is ink and board and a fraction of a second of press time.

Digital inverts the shape. There is no plate, no film and no screen, so the per-job step is essentially the time to load a file — Longrun's corrugated-carton press is described as printing "one carton or one thousand directly from artwork — no plate-making, no minimum, ink dry on exit," with a stated minimum run of one carton. What digital does have is a variable cost that stays stubbornly high: ink laid down per square metre, plus a per-hour cost of ownership that does not fall as the run gets longer.

So you have a curve with a step and a shallow slope, against a curve with no step and a steeper slope. They cross once.

That "board contact" row is worth more attention than it usually gets. Post-print flexo puts pressure on the board, and pressure on fluted board crushes flute tips. That is where washboarding comes from — the visible ribbing in solids and the register drift in halftones on B and E flute. It also costs you edge crush resistance, which is the property your customer's stacking spec is actually written against. A non-contact inkjet process does not apply that pressure. On light flutes and high-graphic work, that difference sometimes decides the job before any cost model does.

Finding your own crossover

The arithmetic is a single line. Let:

• S_f = total fixed cost of one flexo job (plates + mounting + make-ready labour + waste board + ink matching)
• S_d = fixed cost of one digital job (file prep and any RIP time)
• v_f = variable cost per box on flexo (ink + press time per box; exclude board, it is common to both)
• v_d = variable cost per box on digital (ink at your actual coverage + machine hour per box)

Then the crossover run length is:

N* = (S_f − S_d) ÷ (v_d − v_f)

Below N*, digital is cheaper. Above it, flexo is. Two habits make this number useful rather than decorative.

Build it per artwork family, not per plant. Ink coverage is the dominant term in v_d and it varies enormously across your own order book. A two-colour line-art shipper on kraft and a full-bleed, dark, four-colour e-commerce mailer are not the same job on a digital press even at identical dimensions — the second one can consume several times the ink of the first. On flexo the gap between those two jobs is much narrower. So a plant that computes one plant-wide crossover from an average will get an answer that is wrong for both ends of its mix.

Count only avoidable cost on the flexo side. If the flexo line is bought and paid for, its purchase price does not belong in this comparison. What belongs is what you would genuinely stop spending if the job moved: plates, mounting, make-ready labour, waste board, and the press hours themselves. Sunk capital in the comparison pushes the crossover artificially in digital's favour and gets the business case rejected the moment a CFO reads it.

Here is the shape of a worked example. Every figure in it is a placeholder — replace all of them with numbers off your own last three months of job tickets before you show it to anyone.

The two rows the supplier has to fill in are the two you cannot estimate: millilitres of ink per square metre at a stated coverage and quality mode, and rated sheets per hour at that same mode. Ask for both together. A speed quoted at a coverage nobody would ever print is not a number you can put in a model. This is the same discipline you would apply to any capital equipment quotation — the same one that belongs in a machinery specification sheet before you go out to three suppliers.

What your existing flexo line does to the calculation

If you already run flexo, the honest business case usually is not "digital is cheaper per thousand." It rarely is, once runs get long. The case is capacity.

Work out your flexo line's loading. If it is running below roughly 60–70% of available hours, short runs are not costing you anything except their own make-ready — you have press time to burn, and the crossover sits low. Digital has to beat flexo on cost alone, and on a lot of your mix it will not.

If the line is running at 85% and above, the arithmetic changes completely. Every make-ready for a 500-box order consumes press hours that could have run a 50,000-box order at a much better contribution per hour. The relevant comparison is no longer digital cost versus flexo cost. It is digital cost versus the margin on the long-run work you turned away because the press was occupied. On a constrained line, moving short runs to digital does not save print cost — it sells press hours you did not have.

Three practical consequences:

1. Digital replaces one step, not a line. Board still has to be die-cut, folded and glued. A digital press that can outrun your folder-gluer buys you nothing downstream. Check the constraint before you buy the printer.
2. Kraft versus white-top matters more on digital. Four-colour ink on unbleached kraft liner shifts warm and loses gamut. The fixes are a white underbase (ink cost, and a question to ask, because the fact basis does not state whether the carton press carries a white channel) or buying white-top liner (board cost). Either way it belongs in v_d.
3. Register to the die is a separate problem. Ask how printed sheets are registered into your existing die-cutter, and what the tolerance is. This is the integration question that quietly decides whether the machine earns out.

Versioning and obsolescence — the lines that never appear on the quote

Two costs sit outside the print cost per thousand and often outweigh it.

Versioning multiplies the flexo step and leaves digital untouched. Twelve regional language variants of one carton is twelve plate sets and twelve make-readies on flexo. On a plate-free press it is twelve files, and the changeover between them is the time to load the next one. Run the crossover formula with S_f multiplied by the number of versions and N* per version collapses. If your customers are moving toward promotional, seasonal or channel-specific packaging, this term grows every year.

Obsolescence is a real write-off line. Long flexo runs are printed to forecast, and forecast-printed boxes carry the risk of an artwork change, a recipe change, a regulatory label change or a discontinued SKU. Printed board that has to be scrapped is a cost that never appears in anyone's cost-per-thousand comparison. Printing to order converts that finished-goods risk into plain board inventory, which is fungible across every SKU you run. Pull the last two years of packaging write-offs from your finance system before you build the model — for some plants that single number carries the case on its own.

Where digital loses

An honest model has a losing side, and it is worth being blunt about it.

• Long, low-coverage, single-version runs. A one-colour line-art shipper at 80,000 pieces is flexo work, and no crossover model will say otherwise.
• Very heavy solids. Deep flood coverage is where digital ink cost peaks and flexo's per-piece cost barely moves.
• Board-cost-dominated jobs. When board is 70% of the box cost, print process differences move the total less than you expect. Model total delivered box cost, not print cost.
• Throughput ceilings. Digital speed is fixed. It does not improve with run length, so on the long tail of your mix it simply runs out of hours.

The realistic outcome in most box plants is not replacement. It is a second, differently shaped machine that takes the jobs the flexo line was losing money on and hands the flexo line back its capacity.

What to ask before you specify a corrugated digital press

Take this to any builder, including the one you were already talking to, and require the answers in writing on the quotation:

1. Maximum board size in mm, and which flute profiles the press is rated for (B, C, E, EB, BC). Confirm against your own blank sizes, not your finished box sizes.
2. Feed type — sheet-fed or continuous — and rated sheets per hour at a stated coverage and quality mode. Both numbers, together, or neither is usable.
3. Head architecture: fixed single-pass array or scanning carriage, and how many rows. This drives speed, redundancy and the cost of a head failure.
4. Ink: chemistry family, cost per litre, packaging size, laydown in ml/m² at a stated coverage, and whether the system is open or chipped. Ask whether the ink has been assessed for indirect food contact if your customers pack food.
5. Curing: what technology cures the ink, and what the power draw of the cure system is.
6. White channel availability, and what it does to speed and ink cost on kraft liner.
7. Register into converting: how printed sheets index into your die-cutter, and the achievable tolerance.
8. Test print on your own board. Send two different lots of your actual liner — not a clean sample the builder chose — and require print samples back before the balance payment. Longrun's stated build sequence includes a test-print step where the buyer's own files and substrates are run before shipment, and standard models can be sampled the same day; treat that step as your acceptance opportunity, and write the criteria for it into the contract rather than leaving it as goodwill. Structuring that properly is the whole subject of a factory acceptance test for machinery.
9. Payment structure tied to the test. Terms for this class of equipment commonly run T/T with a deposit and the balance before shipment; make the balance trigger a passed run-off on your board, not a calendar date.

Longrun Printing Machinery in Shenzhen builds corrugated-carton digital presses alongside UV flatbed and cylindrical machines, sells from a single unit, and lists carton factories among the customer types it already supplies — the profile is on the Longrun Printing Machinery factory page, and the wider category context sits in the UV flatbed printer buyer's guide and the packaging manufacturer sourcing guide. If sheet handling and usable print area on rigid stock is the part of the specification you are least sure about, the companion piece on what a 2500 x 1300 mm flatbed table actually holds works through that arithmetic in the same way this one works through the crossover.