Cost & Quoting

Cost Estimation and Quoting for Heat Exchanger, Coil, and Radiator Manufacturing

What actually drives cost per coil and how to build a quote that survives brazing yield and leak-test rejects.

On a fin-and-tube coil, raw material is usually 55 to 70 percent of cost, so a quote lives or dies on metal. Copper tube and aluminum fin stock move with the LME daily, and at, say, 4,700 ft of 3/8 in copper tube at roughly 0.14 lb/ft and 4.50 USD/lb, tube alone runs about 2,960 USD before fin. Estimators who quote off last quarter's metal price surrender margin on the first delivery. Pull purchased footage from the Tube Length Usage calculator, not drawing length, because the 706 ft scrap allowance on a 4,000 ft core is real cash you must recover in the price.

Scrap is the quietest margin leak because it multiplies across three processes. A tube lost at hairpin bending carries only cut and bend cost; a core lost at leak test after brazing carries tube, fin, braze energy, and labor, often 40 to 60 USD of embedded value gone in one reject. The Scrap Fin/Tube Cost calculator separates these tiers so you price the weighted scrap correctly. At a 3 percent bend-scrap rate plus a 3 percent post-braze reject rate, effective yield is 0.97 x 0.97 = 94.1 percent, meaning you must build and cost about 6 percent more cores than you ship.

Brazing is the second cost driver after metal, and it is energy plus yield, not just labor. A controlled-atmosphere braze furnace pulling 250 to 400 kW at an industrial rate near 0.12 USD/kWh adds real energy per shift, and every point of post-braze yield below 97 percent turns brazed cores into scrap at full embedded cost. Estimators should load furnace cost per core from throughput, not from a flat burden. Divide shift energy and depreciation by the good cores per shift from the Brazing Furnace Load calculator; at 84 good cores versus a 96-core gross, the per-core furnace burden is 14 percent higher than a naive gross-count estimate.

Leak test is where labor and cycle time hide. Pressure-decay and helium cycles cannot be rushed without losing sensitivity, so each core absorbs tens of seconds to minutes of fixed booth time plus operator attention. If the booth passes 168 good assemblies per shift and a staffed shift costs 320 USD in direct labor plus 180 USD burden, that is roughly 3.00 USD of test cost per shipped core before any reject rework. Size this from Leak Test Capacity so the quote reflects the real bottleneck rate, not an optimistic nameplate cycle count that understates cost per unit.

Direct labor beyond testing, finning, bending, expanding, fixturing, cleaning, and packaging, typically runs 12 to 20 percent of a coil's cost. Pace it against Assembly Takt so you cost the true staffed minutes per core, not an idealized line-balance. A cell at 5.0 minute takt across six manned stations is 30 labor-minutes per core; at a loaded 42 USD/hr that is 21 USD of assembly labor. Estimators who quote off a theoretical cycle time and ignore changeover, breaks, and rework consistently underbid labor by 15 to 25 percent and erode the whole margin on the floor.

Overhead and refrigerant complete the build-up. Apply plant burden as a rate on labor or machine hours, commonly 60 to 120 percent of direct labor, and never bury furnace or leak-booth depreciation in a flat number when those assets dominate the cell. For charged coils, price refrigerant from the Refrigerant Charge Estimate volume: 6 liters of R-410A at prevailing per-kg cost is a line item that swings with the same market volatility as copper. Undercharge and overcharge both create field failures, so a small charge-accuracy reserve belongs in the quote as risk, not as padding.

Assemble the total with the Cost Per Coil calculator: material at real purchased footage, weighted scrap, throughput-based furnace and leak-booth burden, staffed assembly labor, cleaning, packaging, refrigerant, and plant overhead. Then apply target margin. On a coil costing 180 USD to build, a 22 percent target margin quotes at 231 USD, but only if scrap and reject cost are already inside the 180, not absorbed afterward. Quoting the theoretical build cost and hoping yield holds is the single most common way coil shops lose money on jobs they won.

The three estimates that go wrong most often are metal, yield, and test time. Metal because the quote is stale; yield because bend and braze scrap are averaged away instead of tiered by embedded value; and test time because the booth's real bottleneck rate is slower than the cycle sheet. Reprice metal at current LME on every quote, drive scrap from Scrap Fin/Tube Cost with separate bend and post-braze rates, and set labor and burden per unit from the actual throughput of the Brazing Furnace Load and Leak Test Capacity constraints. A quote built that way holds margin instead of eroding it after the first production run.

Published 2026-07-01.