The hidden cost of factory scrap: why protection economics start before assembly

Protective coating is quoted per device. Fractions of a cent to a few cents, depending on the technology, the geometry and the volume. That figure goes into a bill of materials, gets set against the alternatives, and the lowest number usually wins.

It is the wrong number to optimise.

What a device costs to protect is not what the coating costs. It is the coating plus everything the coating process does to the line around it: the units it scraps, the masking labour it demands, the floor space and capital it ties up, the repairs it rules out, and the field failures it does not prevent. Most of those costs sit in budgets other than the one the coating was quoted against. That is precisely why they stay invisible.

What a scrapped unit actually costs

Protection is applied late. By the time a device reaches the coating stage it is carrying nearly all the value it will ever carry: components, placement, reflow, sub-assembly, test. A unit scrapped at that point does not cost what the coating cost. It costs everything spent getting it there, plus the line capacity consumed producing something that will never ship.

The arithmetic gets uncomfortable at volume. Take a product with a bill of materials in the tens of dollars running at a million units a year. A 3% scrap rate at the protection stage is a seven-figure annual loss. None of it appears in the coating quote. It appears in the yield report, which is usually somebody else's problem.

Where protection processes create scrap

Scrap at the coating stage comes from a small number of recurring causes.

Masking is the first. Every keep-out area on a device (connectors, contacts, microphones, antennas, optical windows) has to be protected from the coating itself. A mask in the wrong position leaves an area exposed or puts coating where it must not go. Either outcome is a reject, and the failure rate scales with the number of masks per unit.

Process window is the second. Coatings applied as liquid films have thickness variation across a board and between boards. Too thin in one area and the protection fails qualification. Too thick in another and it interferes with fit, connectors or thermal behaviour. Processes requiring elevated cure temperatures add another exposure, since the components underneath have their own limits.

Handling is the third and the one most often overlooked. Every additional station is another opportunity for physical damage to a nearly finished device. Masking and de-masking are handling steps, which means a masking-heavy protection process carries a handling cost even when the masks themselves go on correctly.

There is also a category that never shows up as scrap at all, and costs more because of it: coverage gaps under densely packed components, where a dispensed liquid cannot reach. Those units pass inspection and fail in the field.

From a 3% plateau to zero

One production deployment had run parylene for years and optimised it as far as it would go. Scrap at volume settled at 3% and stopped improving. That is not a bad number for a permanent conformal process. It was simply the floor of what that process could do on that product.

Replacing parylene with P2i Barrier on the same deployment took scrap at volume to zero.

On a separate wireless earphone programme, also replacing parylene at scale, the P2i process runs at 99.995% first pass yield. The relevant comparison is not P2i against an unprotected device. It is P2i against the protection process a manufacturer is already paying for, measured on the line rather than in a datasheet.

Masking is a line cost quoted as a coating cost

Masking rarely appears in a protection comparison, because it is not part of the coating. It is part of the line: operators applying and removing masks, consumable materials, dedicated stations occupying floor space, and cycle time added to every unit.

Plasma processes change this. Because the coating is deposited at the molecular level rather than dispensed as a film, masking requirements for P2i Barrier are minimised or eliminated. Splash-Proof reduces masking, though device geometry determines how far. Removing masking from a process removes labour, consumables, floor space, cycle time and a failure mode simultaneously, which is why it usually outweighs the per-unit coating price difference that prompted the comparison.

The capital equipment question

A protection technology that requires a new line is not just a coating decision. It is a capital approval, a floor space negotiation, an installation window and a qualification cycle, and it commits the manufacturer to that technology for as long as the equipment is amortised.

Liquid Barrier Coating integrates with existing spray and dip-line infrastructure, which means the protection change happens without new capital equipment. For a factory manager, that is often the difference between a decision that can be made this quarter and a proposal that joins the capex queue.

Rework is the difference between a defect and a loss

Permanent protection processes make a hard commitment on behalf of the device: whatever happens next, this unit will not be opened again.

That commitment has a cost on the line. A unit that fails electrical test after a permanent coating process is scrapped, carrying its full accumulated value with it. The same unit protected with a reworkable coating goes back through, gets fixed and ships.

It has a larger cost afterwards. P2i coatings are 100% reworkable, which keeps battery replacement, component swap and board-level repair available for the life of the product. Potting protects a board effectively and rules all of that out. With right-to-repair requirements tightening across major markets, the reworkability of a protection process has moved from an operational preference to a regulatory consideration.

The cost that lands after shipping

Everything above happens inside the factory, where a failure is cheap. Outside it, the price goes up sharply.

In industrial electronics, a field failure can cost up to 100 times the original assembly cost once installation and replacement labour are counted. Consumer products carry it differently, through returns handling, warranty reserves and the retailer relationships that returns damage. The pattern holds across both: the further from the line a failure happens, the more it costs to resolve.

Two figures from P2i production deployments show what changes when protection is designed against the environment rather than specified late. A Splash-Proof consumer electronics programme saw field failures fall by more than 90%. Across the hearing aid sector, field failure returns fell by half.

The number worth optimising

Protection has never been a per-unit coating cost. It is a yield decision, a floor space decision, a repairability decision and a warranty decision, and it is usually made by comparing quotes that describe none of those things.

The number worth optimising is the cost per device that ships and survives its service life. Most protection comparisons never calculate it, which is why the cheapest coating on the quote so often turns out to be the expensive one on the line.