Are You Over-Specifying Your Coil Line? What Each Spec Line Really Costs
- Share
- publisher
- David Park - Senior Stamping Engineer, FANTY Machinery
- Issue Time
- Sep 20,2026
Summary
Over-specifying a coil line happens one line at a time, each for a defensible reason. This guide takes the spec lines a buyer actually writes down and asks one question about each: does it earn its keep on your production profile? It separates the lines you cannot afford to under-specify from the ones buyers over-buy, shows how an added line reaches landed cost through freight, duty and power, and gives a ten-minute payback test plus a wording guide for the purchase order.
Almost nobody sets out to over-specify a coil line. It happens one line at a time, usually for a defensible reason: the next coil might be thicker, the next die might be tighter, the customer might ask for a faster cycle. Each addition is small. The invoice is not.
This page takes the spec lines a buyer actually writes down and asks one question about each of them — does this line earn its keep on your production profile? For every line there are three numbers: what it adds to the machine price, what it adds to the cost of getting the machine to your floor, and what it costs you if you leave it out.
On this page
- 01 Start from the defects you already have
- 02 The lines you cannot afford to under-specify
- 03 Where buyers most often over-buy
- 04 Headroom is not free
- 05 How an extra line reaches your landed cost
- 06 A ten-minute payback test
- 07 What can wait, and what can never be added later
- 08 Writing a spec that cannot be read two ways
- 09 Where this framework misleads you
- 10 Checking the offer against your own numbers
- 11 Buyer's knowledge base
- 12 Questions buyers ask before signing
Start From the Defects You Already Have
The most reliable input to a coil line specification is not a wishlist. It is last quarter's reject data. Every parameter you are about to write down should be traceable to a defect you can name, count and cost.
Open the scrap report and sort by cost, not by count. A thousand cosmetic marks on a low-value part can matter less than forty scrapped stampings on a high-value one. What you are looking for is the specific failure mode that the new line is supposed to remove.
Then work backwards from that defect to the spec line that controls it. This is where most specifications go wrong: buyers specify the feature they can see advertised, not the one that governs their defect. A thicker frame is easy to photograph. Roll pitch is not.
| What the reject bin shows | Parameter that actually governs it | Parameter buyers usually specify instead | Where this misleads you |
|---|---|---|---|
| Length drift across a coil, worst at the tail | Decoiler brake behaviour and loop window, not feeder resolution | Encoder resolution and controller brand | Both are real numbers. Only one changes the defect |
| Hole-to-hole position outside tolerance | Feed repeatability plus pilot release timing | Nominal feed accuracy class | A ±0.05 mm class rating does not include the press interface |
| Edge wave that survives straightening | Roll count and roll diameter relative to strip width | Total machine tonnage and frame weight | A heavier machine with too few rolls straightens nothing extra |
| Coil set and crossbow that return after the die | Roll penetration and the yield-strength match | Number of straightening rolls advertised | Roll count without penetration range is half a specification |
| Surface marks on coated or polished strip | Roll material, finish and pressure control | Nothing. Buyers rarely specify roll surface at all | This is the line most often discovered after the machine lands |
| Cycle time short of plan at the coil tail | Loop size, ramp profile and decoiler acceleration | Peak SPM printed on the brochure | Peak speed is reached for part of one coil, not for a shift |
Three or four defects named honestly will usually define the machine. A specification built from fifteen wishes defines a machine nobody can quote against, because every supplier resolves the ambiguity in the direction that suits their price.
Write the defect next to the spec line. If a colleague can read your specification sheet and ask "which reject does this remove?", you have a document a supplier can price. If they cannot, you have a shopping list.
The Lines You Cannot Afford to Under-Specify
Some specification lines behave asymmetrically. Spending a little more costs you money once. Leaving them short costs you money every shift, and in most cases the machine has to be replaced rather than adjusted.
The test is simple: can the shortfall be corrected on your floor, at reasonable cost, without taking the line out of service? If the answer is no, that line is not a place to negotiate.
| Spec line | What it buys you | What under-specifying costs you | Where this advice misleads |
|---|---|---|---|
| Frame and roll stiffness for your maximum strip thickness | Deflection that stays inside tolerance at full width and full thickness | Flatness that degrades only on the heaviest job — so it passes acceptance and fails in production | Stiffness is not linear in thickness. Margin that is adequate at 3 mm is not enough at 4.5 mm |
| Straightener roll count relative to strip width | Enough bending reversals to consume edge wave as well as coil set | Edge wave that reappears as soon as the coil changes and stays for the life of the machine | Extra rolls help only if roll diameter and pitch support them. Roll count alone proves nothing |
| Decoiler mandrel capacity and expansion range | Holding the full coil weight without slip, across your real ID range | Slip under load, coil damage at the inner wrap, and a hard limit on the coil sizes you can buy | The coil weight in the table is the maximum. The mandrel is happiest in the middle of its range |
| Feed drive torque at your required acceleration | Reaching cycle speed without the strip slipping in the nip | Slip — the single largest error term in most feed accuracy budgets | Torque and speed are quoted separately. Ask for both at the same time |
| Motor and control enclosure rating for your plant | A machine that survives your ambient temperature, dust and washdown practice | Faults that correlate with weather or with the shift that cleans the floor | A higher IP rating is not a substitute for a plant that is simply too hot |
| Tooling and coil changeover interface | Changeover time close to the figure you were promised | A machine that meets its specification only on a long run, and is unaffordable on high mix | A fast changeover demo uses a fresh, low-mass coil and a clean die |
Note what these six lines have in common. None of them is a control feature. They are all mechanical, and mechanics is what the controller has to live with.
That is not an accident. A controller is easy to replace and easy to upgrade. A frame, a roll set and a mandrel are the machine. When reading a specification sheet, the mechanical lines deserve more attention than the electronics block, even though the electronics block is longer.
Where Buyers Most Often Over-Buy
The opposite mistake is quieter, because nothing fails. The machine performs well, the acceptance test passes, and the extra money is simply gone. It shows up as capital that is no longer available for the second press, or as a line whose changeover takes longer than it needed to.
These are the lines where a margin that looked prudent turns out to be the most expensive thing on the sheet.
| Spec line | What buyers add, and why | What it actually costs | How it usually ends |
|---|---|---|---|
| Strip width headroom | One size up "in case the product range widens" | A wider frame, wider rolls and a heavier machine — plus freight by volume | The wider width is never used, and roll bending at narrow widths is worse than on the correct machine |
| Thickness ceiling | Capacity for the one thick job that comes twice a year | More stiffness, more torque, more power, and a machine that is clumsy on the thin work it does daily | Outsourcing those two jobs would have cost a fraction of the premium |
| Peak speed | The fastest number in the quotation, to protect against future demand | A longer loop, higher acceleration, more brake wear, and a machine that never runs at peak anyway | The line settles at the speed the die and the operator can actually sustain |
| Control tier above the task | Closed-loop strip sensing, remote diagnostics, plant-network integration | Licence and integration cost, plus longer commissioning and a new dependency at every fault | Two of the three functions are never switched on after handover |
| Full option list at order | Coil car, hydraulic expansion, oiler, extra sensors, spare mandrel | Options are cheapest at the factory, so buyers take them all at once and never in isolation | Some options are genuinely cheaper later as a kit. Others are impossible later. The list blurs the difference |
| Squeezing the last tolerance band | Buying the tightest accuracy class quoted | Cost rises out of proportion to the accuracy gain, and the line becomes less tolerant of ordinary variation | The die never needed it. See how to size a feeder for your press line for the arithmetic |
There is a pattern here worth naming. Over-buying clusters around three things: capacity, speed and control. Those are also the three things a supplier finds easiest to add to a quotation, because all three are purchased components rather than engineering.
The lines that are genuinely hard to get right — roll geometry, pass line, brake behaviour, changeover mechanics — are the ones buyers tend to specify loosely. That is exactly backwards.
Headroom Is Not Free
Headroom is the margin between what your production needs today and what the machine can do. It is genuinely valuable in the right place and genuinely expensive in the wrong one. The difference is whether the margin is bought with engineering or with purchased capacity.
Buy headroom in mechanics if you must. A frame that can take the next thickness without deflection costs more steel and some design effort. It does not add a second motor, a larger control cabinet, a heavier shipping crate or a bigger crane on your floor.
Be far more suspicious of headroom in driven capacity. A larger servo, a longer loop, a higher brake rating and a wider frame all travel together. Buying 25 per cent more speed typically means buying 25 per cent more of everything, including the parts that wear.
- Thickness headroom is usually the best-value margin, because the cost is stiffness rather than power.
- Width headroom is usually the worst, because roll bending and flatness get worse at narrow widths on a wide machine.
- Speed headroom costs you in four places at once: motor, loop, brake and wear.
- Coil weight headroom is worth paying for, because coil supply is the least controllable part of your input.
- Accuracy headroom beyond the die requirement buys nothing at all — the die, not the feeder, is the limit.
The arithmetic that keeps this honest is to state your requirement as a range, then check the machine's capability against the range's extreme rather than its middle. A line specified for 0.8 to 3.0 mm that spends 90 per cent of its life between 1.2 and 1.8 mm should be judged on the band it lives in — provided the extremes are rare enough to be scheduled.
If the extremes are not rare, you are not over-specifying. You are specifying correctly, and the earlier section on lines you cannot afford to under-specify applies instead.
How an Extra Line Reaches Your Landed Cost
A specification line never appears on your invoice once. It appears at every stage between the factory and your production floor, and some of those stages multiply rather than add.
This is why a nine per cent increase on the machine price can become a fourteen per cent increase on the project. The line adds weight and volume to the crate, value to the customs declaration, load to your power supply and complexity to your spares list.
| Cost stage | What triggers it | Why an extra spec line moves it | Where it is negotiable |
|---|---|---|---|
| Machine price | The specification as written | Directly. This is the only stage buyers usually model | Almost always — but on scope, not on the total |
| Packing and crate | Machine envelope and centre of gravity | A wider or taller machine needs a stronger crate and more bracing | Rarely worth negotiating. A weak crate costs more than it saves |
| Freight by weight and volume | Gross weight and cubic measure | Both grow with added capacity, and they are charged together | Incoterm choice moves who pays, not how much. See how shipping cost is built up |
| Duty and import charges | Declared value and HS classification | A higher-value machine on the same classification carries proportionally more duty | The classification is settled before ordering, not after. Import cost lines explains why |
| Rigging and installation | Weight, footprint and crane limits | An extra metre of frame can decide whether your existing crane is adequate | Not negotiable, and often the largest surprise |
| Power and services | Connected load and control cabinet size | Larger drives and more options raise the connected load and the supply cable rating | Depends on your plant. Power supply requirements covers the calculation |
| Spares and consumables | Component count and variety | More drives, more sensors and more optional kit mean a longer, more expensive shelf | Partly — see the spares you actually need |
| Running cost | Power draw, roll wear, brake wear | Every added capacity is paid for again on every shift, for years | Not negotiable, and rarely modelled at all |
Notice which stage has the widest negotiating range and which has almost none. The machine price is where the discussion always happens, and it is also the stage where a supplier has the most room to move scope without moving the headline number.
The stages that are effectively fixed — crate, freight, rigging, duty — are the ones that reward you before the order, not during the negotiation. Every kilogram you decline to specify is a kilogram you never ship, never clear through customs and never lift into place.
That is the practical reason to write a tight specification. It is not frugality. It is that the cost of a spec line is not linear in the price of the spec line.
A Ten-Minute Payback Test for Any Spec Line
Most spec lines can be tested without a model. You need three numbers: what the line adds to the price, what it saves or earns per year, and how you would solve the problem without it.
The third number is the one buyers skip, and it is the one that decides the case. If the alternative to the feature is scrapping parts, the saving is easy to justify. If the alternative is a phone call to an outside processor twice a year, the case is much weaker.
| Spec line | Typical added cost | What you save or earn | Payback |
|---|---|---|---|
| Roll count sufficient for edge wave | Modest premium within the same machine class | Removes a defect class outright, including the labour spent chasing it | Usually inside the first year, and occasionally inside the first month |
| Roll surface suited to coated strip | Small. Sometimes it is a specification choice rather than an upgrade | Removes a scrap mode that no setting can fix | Effectively immediate, because there is no workaround |
| Second die recipe storage | Very small, often included | Changeover time on every product change | Weeks, if changeovers are frequent |
| Higher speed class | Large, and it drags extra components with it | Only what the die and the operator allow you to realise | Often never, unless a demonstrable bottleneck exists |
| Wider machine | Large, and it carries freight and floor-space consequences | Nothing until the wider product is genuinely on the order book | Speculative. Treat as an investment decision, not a specification one |
| Coil car | Moderate | Loading time and the risk of coil damage at every change | Fast at high changeover frequency, slow on long runs. Running-cost analysis puts it in context |
A worked example makes the method concrete. Suppose the roll surface upgrade for a polished-strip job costs a low four-figure sum in the machine price. That job produces scrap averaging a few hundred currency units a month from surface marks that no pressure setting resolves.
Under three years, the upgrade is a rounding error against the scrap. Under that calculation almost every surface-quality option pays. But run the same test on a speed class that costs several times as much and only helps if the die can keep up, and the answer inverts immediately.
Watch for the trap of comparing an upgrade against zero. The correct comparison is against the cheapest configuration that still solves the problem, including the cost of solving it a different way. An upgrade that looks expensive beside the base machine can look cheap beside a second shift.
What Can Wait for Phase Two, and What Can Never Be Added Later
The single most useful question to ask a supplier is not "how much is this option?" It is "can this be fitted after the machine is in my plant?"
Options that can be retrofitted should be judged on their payback, because the decision is reversible. Options that cannot are judged on a different basis: getting them wrong means living with the mistake for the life of the machine.
| Spec line | Factory-fit only? | Cost if added later | How to treat it |
|---|---|---|---|
| Roll count and roll pitch | Yes | Not available. This is a new straightener head, not an option | Decide before order, on defect data. This is the least reversible line on the sheet |
| Roll surface and material | Yes in practice | Roll set replacement. Possible, but only as a full exchange | Decide from the strip finish you will run, not the strip you run today |
| Mandrel expansion range | Yes | New mandrel assembly | Decide from the coil ID range you can actually buy |
| Frame width and stiffness | Yes | Not available at any price | Treat as permanent. Buy the smallest machine that covers your real range |
| Coil car | Usually yes, but easy | A modest premium over the factory price | Reversible. Judge purely on payback |
| Oiler and lubrication options | Rarely | Straightforward, and sometimes better specified after you know the die | Defer if the die is not yet fixed |
| Recipe storage and interface options | Usually software | Licence plus commissioning time | Reversible. But check the licence terms — and check what payment and contract terms they sit under |
| Plant-network and data integration | Software, but depends on the control platform | Depends entirely on whether the controller has the port and the protocol | Ask now, decide later. The answer is worth knowing even if the feature is not |
Two rules follow. Never defer a decision that is factory-fit only, and never accelerate a decision that is cheaply reversible. Most specifications break the second rule, because options are cheapest at the factory and buyers rationally buy them there — without asking whether the need exists yet.
Ask the retrofittability question in writing. Have the supplier state, per option, whether it can be fitted on site and at what premium. That single column turns an option list into a decision tool, and it is a fair question that a serious manufacturer will answer without hesitation.
Writing a Specification So It Cannot Be Read Two Ways
A specification is not a description of a machine. It is a set of obligations, and every sentence that can be read two ways will be read in the direction that costs the supplier less.
This is not dishonesty. It is the ordinary consequence of writing a commercial document in the language of a brochure. Brochures sell a capability. Specifications buy a machine, and the two use the same words with different meanings.
Three wordings cause most of the disputes. The first is a bare number with no test condition: "feed accuracy ±0.05 mm" says nothing about at what speed, with what strip, over what proportion of the coil, or measured how. See how a specification sheet should be read for the parameters that surround a rating.
The second is a capability stated for the base machine and assumed for the optioned one. Adding a wider frame or a heavier coil capacity changes the numbers the original rating was measured against, and the quotation rarely restates them.
The third is a performance figure with no acceptance method. If nobody agreed how it will be measured, the machine passes acceptance by default and the shortfall appears in the third month. What an acceptance test should cover sets out the tests worth naming before the order.
| Weak wording | Why it fails | What to write instead | What it costs you if you leave it |
|---|---|---|---|
| "Feed accuracy ±0.05 mm" | No speed, strip or measurement basis | Accuracy at a stated speed, strip and material, measured over a stated proportion of coil length, by a stated method | The acceptance argument, and every argument after it |
| "Speed up to 60 SPM" | "Up to" describes a peak reachable for part of one coil | Sustained speed over a full coil at the specified strip, with the defined ramp | A line that meets the paper number and misses the production plan |
| "Straightener with 9 rolls" | Roll count without diameter, pitch or penetration range | Roll count, diameter, pitch and the penetration range available at each strip thickness | The defect returns and the roll count turns out to have been decorative |
| "Suitable for coil weight up to 5 t" | Silent on coil ID range and on slip at rated load | Coil weight, ID and OD range, and the condition at which the mandrel is specified to hold | Slip on the inner wrap and a limit on the coils you can buy |
| "Includes standard tooling" | "Standard" is defined by the supplier | An itemised scope of supply, with anything excluded written down | The changeover time you were promised, discovered on your floor |
An itemised exclusion list is worth as much as the inclusion list. The disputes that cost real money are almost never about what was promised. They are about what everyone assumed was included and nobody wrote down.
One further point: the specification is also the document that makes quotations comparable. Two suppliers quoting against a loose specification are not quoting the same machine, which is why the comparison exercise described in comparing supplier quotes only works on a tight one.
And before the enquiry goes out at all, what an RFQ should contain determines whether the answers come back comparable.
Where This Framework Misleads You
Treating every spec line as a payback calculation is a good default and a bad religion. There are real situations where buying more than you can justify on reject data is the correct decision, and it is worth knowing what they are before you argue with a colleague who is applying this page too literally.
When a customer specifies it. If your machine is dedicated to one customer's programme and their engineering standard names a specification, the payback test is irrelevant. You are not buying capability, you are buying the right to quote. Argue about cost recovery, not about the specification.
When the product range is genuinely uncertain. A shop that has just won a contract with a stated expansion path is not over-specifying when it buys width headroom. It is buying an option on future work. The honest version of that decision names the future work and puts a probability on it — not "we might need it someday".
When the reject data is too thin to read. A new line, a new part or a very low-volume shop may not have enough history to build a specification from defects. In that case skilled judgement, a reference installation and a supplier who will explain their reasoning are worth more than a spreadsheet with three data points.
When flexibility is cheaper than skill. A line with a wide tolerance band, generous recipes and forgiving mechanics costs more to buy and less to operate in a shop where the operators change frequently. This is a legitimate trade, and it is often made badly — buyers take the flexibility without costing the changeover time it adds.
When safety, statutory or site rules apply. Guarding, isolation, noise and electrical standards are not options to be optimised. They are conditions of operating the machine at all, and a payback test does not apply to them.
There is one further case, and it is the most expensive: when the specification is being used as a substitute for a decision. If nobody in the organisation can say which product the line is for, the option list becomes a way of postponing that question. Every line added is a small bet on an unnamed future, and the total is the price of not deciding.
Checking the Offer Against Your Own Numbers
The point of writing a tight specification is that it gives you something to check. A quotation read against a loose specification is a price. A quotation read against a tight one is a set of answers you can verify line by line.
Do this before the order, not at acceptance. Most of what you find is a misunderstanding rather than a shortfall, and misunderstandings are cheap to fix in a document and expensive to fix in steel.
- Take your own defect list and ask, per defect, which line of the offer removes it. Unanswered defects are the gaps.
- Check every rating for a stated test condition. A number without conditions is a marketing number.
- Confirm the mechanical lines against the extreme of your real range, not the middle.
- Read the exclusion list before the inclusion list. Exclusions are where projects slip.
- Ask which lines are factory-fit only, and which can be retrofitted and at what premium.
- Cost the differential in freight, duty and rigging, not only the machine price. See how total cost of ownership is built for the full frame.
- Name the acceptance method for every performance figure, and agree it in the purchase order.
- Check the commercial frame as well: warranty scope, spares, and what happens in the first year. See what a warranty typically covers.
None of this requires special knowledge. It requires reading the offer against a document you wrote yourself, with a specific defect in mind for every line. That is the whole method.
The same discipline applies on the supplier side of the transaction. A buyer who asks precise questions and shows their own reject data is easier to quote for, because the ambiguity that would otherwise be resolved by priced assumptions simply disappears.
Buyer's Knowledge Base
This page is the selection and procurement cluster of the coil line knowledge base. The questions below go one level deeper into a single stage of the buying process.
1. What the machine costs, and why quotes differ
Price ranges by class, the components that move a quotation, and the gap between a bought price and a delivered one.
2. Choosing the configuration
Which machine class, how many units, and which options earn their place on your production profile.
3. Supplier, quality and verification
Reading a supplier, auditing a factory, checking the machine and avoiding the classic buying errors.
4. Contract, delivery and plant readiness
Documents, payment and finance, lead times, installation, spares and what the machine is worth later.
Part of the 3-in-1 decoiler straightener feeder cluster. The other three clusters feed into this one:
Questions Buyers Ask Before Signing
At what strip thickness does frame stiffness stop being adequate?
There is no fixed answer, because stiffness matters relative to strip width as well as thickness. A practical rule is that if you are within 15 per cent of the machine's rated maximum thickness at full rated width, ask the supplier for the deflection figure at that combination rather than the nominal capacity. If they cannot produce it, that is itself the answer.
How much does widening the machine add to landed cost?
More than the machine price suggests. A one-step increase in width typically adds a meaningful premium to the machine, increases crate volume, raises freight by both weight and measure, and can add enough mass to change your rigging plan. Budget the increase as a project number, not a machine number, before deciding.
Is it worth specifying 9 straightening rolls instead of 7?
Only if the extra rolls come with a roll diameter and pitch that can actually reach the defects you have. Nine small, closely pitched rolls and seven larger ones are different machines, not the same machine with two more rolls. Decide on the defect — edge wave in particular needs roll count — and then check that the geometry supports it.
How much speed headroom should a specification include?
Enough to cover the die and the operator, and not much more. If your die and handling realistically sustain 40 strokes per minute, a machine rated at 60 is not buying you production — it is buying loop length, brake wear and a higher connected load. Specify sustained speed over a full coil, not a peak figure.
Which options can genuinely be added after installation?
Coil cars, oilers, many software and interface functions and some sensing packages can usually be retrofitted, at a premium that varies from trivial to substantial. Roll count, roll geometry, mandrel range and frame width cannot. Get that split in writing per option before the order, because it is what separates a reversible decision from a permanent one.
How do I check a quotation without a specification?
You cannot compare it. Two offers read against a loose requirement are not offers for the same machine, and the difference surfaces as line items that appear in one and are priced as extras in the other. Write the specification first, then ask both suppliers to respond to it line by line — that is the only comparison that means anything.
Where to go next
This is the last of the four coil line clusters. The machine side of the decision — decoiler specification, straightening and feed accuracy — is covered in the other three guides, and the overview that ties them together is the 3-in-1 decoiler straightener feeder guide.
Read the 3-in-1 decoiler straightener feeder guide