Why a substitute that exists does not relieve a shortage

Somebody announces an alternative material. The story says the shortage is solved. Three years later the shortage is still there and the alternative is still being tested. This happens so reliably that it is worth understanding why.

The answer is a single unglamorous word: qualification.

What qualification actually is

Before a factory will use a material, it has to prove that the material works in that factory, for that product, at the yields it needs. Not that it works in principle. Not that it works at a competitor. There.

That proof means running the material through the process, measuring what comes out, running it again at volume, checking that nothing drifts over months, and documenting all of it. For a semiconductor process the cycle is routinely one to three years. For an aerospace or medical part it can be longer.

And it does not transfer. A resist qualified at one fab is not qualified at the fab next door, because the two have different tools, different process histories and different defect signatures.

Why this changes what a shortage means

Once you know about qualification, a lot of confident commentary stops making sense.

"There is an alternative" is usually true and usually irrelevant. A substitute that exists in a laboratory and a substitute that is qualified are different facts, and only the second one relieves anything on the horizon that matters.

Capacity announcements are not supply. A new plant that has not been qualified by any buyer is a building. The gap between commissioning and shipping qualified material is where most optimistic forecasts quietly die.

Second sources are worth more than spare capacity. A buyer with one qualified supplier and a shortage cannot simply buy elsewhere. A buyer who qualified a second source three years ago can. Which is why the sensible time to qualify a second source is when you do not need one.

What it means for the numbers on this site

The severity score has a question in it — how long from decision to new capacity — that is mostly measuring qualification, not construction. That is deliberate. Building a plant is a known problem with a known cost. Getting its output accepted is the part that takes the years.

It is also why the Capital section says money is not the constraint for most rows. You cannot buy your way through a qualification cycle. You can start it earlier, and that is all.

The one thing that would change this

If qualification cycles compressed — if second sources reached production in quarters rather than years — a great deal of this map would stop mattering. That is written down as a falsifier against one of the published claims, because it is the single change that would most undermine this research.

It is worth watching for. It has not happened yet.