When a duct shop asks me how to cut cost, most expect an answer about machine speed. Then I ask one question: what is your sheet utilization rate? Very few can answer with a number. The ones who track it sit anywhere between 80 and 95 percent, and that spread is worth more money per year than any upgrade to the line. On a shop consuming 300 tons of galvanized coil a year, moving utilization from 84 to 92 percent saves about 24 tons of steel. At 750 dollars a ton landed, that is 18,000 dollars a year, found in the scrap bin. This article is about where that material disappears and how to get it back, whether you blank with shear or plasma.
Every cut leaves a kerf or a sheared strip. On a shear line the loss between two parts is nearly zero if they share a cut line, which is why shear blanking with smart sequencing can hit very high numbers. On a plasma line each cut eats 1.5 to 3 millimeters of kerf depending on nozzle size and consumable condition, and the lead-in and lead-in overcut add more. Nest twenty parts across a 1219 millimeter coil with individual outlines and you can easily lose 60 to 80 millimeters of width to kerf and clearance. That is five to six percent of the coil before anything else goes wrong.
The fix is common-line nesting: let adjacent parts share one cut. Two duct blanks of the same width separated by a single plasma cut lose one kerf instead of two, and the parts stay square because the shared cut is one continuous motion. Shops that adopt common-line nesting on rectangular blanks typically recover three to four percent of sheet immediately. It requires a programmer who thinks about part order, not just geometry, but it costs nothing in hardware.
The second leak is quieter. Coils arrive at 1219 or 1250 millimeters, and your most common blank development is, say, 1090 millimeters. Every sheet leaves a 129 millimeter strip that is too narrow for that product but too wide to scrap without guilt. Multiply that strip by every meter of coil and it becomes a warehouse corner full of odd widths nobody reaches for. The classic answer is a remnant logbook, physical or just a spreadsheet, listing every usable offcut with width, gauge, and length. The better answer is to plan the week so those widths get consumed: corner pieces, stiffener blanks, small fittings, and test coupons all eat remnants happily.
Consumable condition deserves its own warning. A plasma electrode and nozzle past their service life do not just cut ugly edges; the kerf widens. A fresh 60 amp setup on 1.0 millimeter galvanized cuts a kerf around 1.5 millimeters, and a worn one drifts toward 2.5 or 3. Your nesting program assumed the narrow kerf, so every shared cut loses half a millimeter more than planned and every pierce overcuts the part edge. Track pierce counts per electrode, change them on a schedule rather than when the cut starts to hang dross, and your programmed yield and your actual yield stay the same number. Shops that run consumables to failure routinely lose two percent of sheet to this alone, and the loss hides inside the machine, not in the bin, so nobody sees it.
The choice of cutting method changes your nesting ceiling. A shear on an auto duct line cuts a full-width slit in one stroke, and the offcut is a clean rectangle you can reuse. Plasma cuts any shape, which is freedom for fittings, but every shaped cut costs kerf and edge quality attention. Neither is universally better; they ceiling out differently.
| Factor | Shear blanking | Plasma cutting |
|---|---|---|
| Best nesting pattern | longitudinal strips, shared cut lines | free shapes, common-line pairs |
| Kerf loss per cut | near zero, offcut reusable | 1.5 to 3 mm per cut |
| Typical utilization, disciplined shop | 92 to 96 percent | 88 to 94 percent |
| Fitting flexibility | limited to straight blanks | any shape including reducers and taps |
| Edge condition | clean, no dross | needs dross control on bottom edge |
| Consumable cost | blade wear only | electrodes, nozzles, swirl rings |
| Changeover to a new size | seconds, program length | seconds, program path |
Read the table honestly for your product mix. If eighty percent of your volume is straight duct, shear blanking with strip nesting is hard to beat on yield. If you fabricate a lot of transitions, plenums, and custom fittings, plasma pays for its kerf in flexibility. Many shops run both and route parts accordingly, which is exactly what I recommend when the budget allows.
The single highest-leverage decision is standard duct lengths. If your market accepts 1150 or 1200 millimeter modules, design the week's cutting so blanks fit coil widths with minimal trim. A 1219 millimeter coil minus 8 millimeters of edge trim gives 1211 usable; two blanks of 605 fit perfectly, or one of 1150 with a 61 millimeter remnant destined for corner inserts. Shops that let sales promise any length the customer dreams up pay for that freedom in the scrap rate, and they deserve to know it.
The second lever is buying the right coil widths. List your top ten products by annual tonnage, compute each blank development width, and buy coils that split into those widths with the least remainder. Sometimes the answer is buying 1000 millimeter coil for one high-volume product instead of forcing it out of 1219. Steel mills and service centers will slit to width for a small premium, and that premium is almost always cheaper than the strip you would otherwise scrap. A power slitting machine in-house makes even more sense once remnant volume justifies the floor space. One customer of ours moved from two coil widths to four and lifted utilization eight points in a quarter, with zero change to the machines.
Third, nest by material, not by customer. Cutting an order-sequenced week means every gauge change wastes setup and every job leaves its own private remnant. Grouping the week by gauge and width lets one coil run to near exhaustion and one remnant serve many jobs. It requires a scheduler willing to push back on delivery dates by a day occasionally, and a supervisor who keeps the remnant logbook alive. Both are habits, not purchases.
You cannot manage a number you only estimate. The working definition I use is simple: square meters of finished parts divided by square meters of coil consumed, per batch. Weigh nothing, calculate from drawings, and record it weekly. Track it by product family and by machine, because the plasma cell and the shear line will show different numbers and both will show different trends. When the weekly number dips, there is always a reason: a new operator, a customer's odd lengths, a fitting-heavy batch, or a coil width bought by habit. Find the reason within the week while the sheets are still traceable, and the number climbs back on its own.
Set the target realistically. Straight duct on a shear line should hold 93 or better. Mixed fabrication with plasma should hold 90. If you are at 85, the road from 85 to 90 is mostly discipline. The road above 93 gets expensive, and I would not chase it at the cost of deliveries.
Concrete numbers help, so here is a week from a shop I worked with, running a Line 5 shear line and a plasma table, consuming about 12 tons of 1219 millimeter galvanized weekly. Straight duct was 68 percent of tonnage, blanked on the shear with strip nesting at 94 percent utilization. Fittings and custom pieces were 22 percent, plasma cut with individual outlines at 89 percent. The remaining 10 percent of tonnage came from remnants consumed on small parts at 75 percent, because remnants never nest perfectly. The blended rate works out to roughly 91 percent. After the crew moved fitting blanks to common-line pairs and cleaned up the plasma kerf assumptions, the fittings number rose to 92 and the blended rate reached about 92.5. One and a half points on 12 tons a week is around 600 dollars a month at the coil prices of that period, from two changes that cost nothing but programming hours and attention.
The same shop had tried to jump straight to 95 by buying custom slit widths for everything. It backfired: minimum order quantities on odd widths tied up capital, slow-moving coils sat in the rack, and the remnant corner grew instead of shrinking. Utilization optimization has a sweet spot, and for most mixed shops it sits in the low nineties. Beyond that you are buying inventory problems with your yield points, and the warehouse will eventually send you the bill.
For straight duct blanked on a shear line, 92 to 96 percent is achievable with strip nesting and standard lengths. Mixed fabrication with plasma cutting typically lands 88 to 94 percent. Below 85 percent means money is leaking somewhere structural, usually coil width selection or uncontrolled custom lengths.
Each plasma cut loses 1.5 to 3 millimeters of kerf, while a shear cut leaves a reusable rectangle. On pure rectangular nesting that difference can cost three to five points of yield, but plasma lets you nest fittings and odd shapes that a shear cannot produce at all. Run both and route parts by shape if volume justifies it.
Keep a remnant logbook with width, gauge, and length, and schedule products that consume those widths: corner pieces, stiffener blanks, small fittings, and coupons. Better yet, ask your supplier to slit coils to the widths your top products actually need, and pay the small slitting premium instead of scrapping the strip.
Yes. Standard modules such as 1150 or 1200 millimeters let blanks split a coil width with minimal trim, and they also speed assembly because connectors and hangers repeat. If sales insists on every custom length, at least track the scrap cost of those orders so the pricing reflects it.