Ann: AUZ enters research partnership to develop scandium alloy, page-78

I struggled with this, but think it goes like this:

Assume the Al-alloy ingot is 5 kilos; now imagine a user needs 0.5% Sc for the extrusion, sheet or plate.
The Cambridge process allows the standard 2% or 100 gms of Sc, to rise to 1.5 kilos of Sc per ingot. The Al content falls to 3.5 kilos from 4.9 kilos.

The application may be say a ship builder who needs 500 t0ns of Al plate to build a luxury yacht or a sea going ferry, or perhaps Daimler which wants to manufacture 10,000 electric delivery vans or 30,000 buses that weigh say 15% less.

Previously they would order alloy from a Al smelter group like Constellium, and it would order "master alloy" ingots which would include Al that the smelter doesn't necessarily want to buy. Now under this process Constellium buys more Sc and less volume of Al per ingot and significantly fewer actual ingots to achieve the same 0.5% across the 500 tons.

As these ingot concentrates are made by a simpler, lower energy process there is a second order saving as well, but there is also a third saving since AUZ is likely to sell the Sc to the ship-builder at lower price than has been prevailing. This may be 30% less, perhaps 50% less, perhaps more. It will be price that preserves margin, but also a price that offers a strong an incentive for the transport, power and additive manufacturing industries to exploit the advantages of stronger, more formable less corrosive, less stretchable, more easily welded Al. It seems to be a very important announcement as Al smelters are looking to extend their markets and of course in the background is reduced fossil fuel consumption and the German auto industry reeling as diesel sales fall and EVs take off.