There has been a decade of electric furnace announcements. What makes this one worth a long article is that it has been running, continuously, at a scale where the economics can be checked.
The numbers
At full production the furnace delivers roughly 400 tonnes of glass a day against a comparable hybrid line of 380. Energy per tonne is lower, and materially more stable across the month — the stability is worth as much as the headline figure, because a stable melt means a stable ribbon.
What was harder than expected
Electrode wear is the operational story. The wear rate is well understood, but the consequence for campaign planning is not: a plant that plans 18 months between changes is planning on the wrong number.
The second surprise was refractory. A furnace without a regenerator still has a hot face, and the wear pattern upstream of the electrode line is unlike anything the maintenance team had seen before.
Would we build another one
Asked directly, the plant management said yes — with two changes. They would specify a larger transformer for headroom, and they would buy the maintenance relationship rather than the parts.
