Foaming Slag Is a Financial Instrument

Three things worth your attention this week.

Injection carbon suppliers sell tonnes; the furnace buys CO. Ash content, sizing, and moisture decide how much of each tonne actually becomes foaming gas versus how much becomes slag mass and steam — which means two quotes at the same $/t can differ 20% in delivered value. When purchasing switches carbon grades to save a few dollars a tonne, the panel usually finds out before the buyer does. Insist the spec sheet travels with the price.

Water-cooled panel replacement remains one of those costs that hides in the maintenance budget instead of the energy story where it belongs. Panel life is substantially an arc-exposure history — every hour of thin foam is radiation the copper absorbed on your behalf. Shops that trend panel repairs against foam quality, rather than against calendar age, consistently find the correlation the budget was hiding.

Graphite India's June-quarter results, published August 4, put electrode capacity utilization at 97%, up from 82% a year earlier, with net sales up 27% on EAF demand. Supply is tightening while prices have so far moved only marginally. The operator read: the electrode-savings leg of the foam payout is about to be worth more — thin foam is arc instability, and arc instability is electrode consumption you'll buy at next year's price.

Strip the metallurgy away for a moment and foaming slag is a simple financial position: you pay a running premium — injection carbon plus oxygen plus the operator attention to run them — and the payout arrives as energy efficiency, panel life, electrode savings, and quieter electrics. The published physics is stark enough to carry the whole argument. Radiation studies going back to Bowman's arc work put a bare, exposed arc's useful transfer to the bath at roughly a third of its power, with the rest radiating to panels and roof. Bury that arc in foam and published figures put transfer up toward 90%. Nothing else on the furnace moves energy efficiency across a range that wide.

So the real question isn't whether to foam — everyone foams — it's whether your foam is actually in the money, heat after heat, and almost no shop measures that. The premium side of the ledger is easy to read off the consumption reports: kilograms of injection carbon and cubic meters of oxygen per tonne. The payout side is where discipline goes to die, because it's spread across three budgets — kWh/t on the energy report, panel repairs in maintenance, electrode consumption in consumables — and no one owns the sum.

The chemistry that prices the position: foam is CO gas held in a slag with the right viscosity and surface properties — the foaming-index work of Ito and Fruehan formalized what every good furnace operator reads by ear. You need FeO and carbon meeting in the slag to make gas, a basicity that holds the bubble structure, and MgO near saturation so the slag isn't dissolving your refractory while it works. Too fluid and the foam collapses; too stiff and gas channels through. The FeO band we argued two issues ago — broadly 15-25% in published practice — is also the foaming band. That is not a coincidence; it's the same balance seen from two sides.

A DRI charge changes the position fundamentally, and mostly in your favor. Carbon and iron oxide arrive together inside each pellet, generating CO continuously and exactly where the foam needs it — continuous feeding gives you a steady gas source scrap shops must synthesize with lances. In shops we've worked, well-run continuous DRI feeding produces the most stable foam you will ever see. The tax for it: gangue-heavy slag volumes mean basicity and MgO control take real flux discipline, and a continuously fed furnace can over-foam quietly until slag rolls over the sill — carrying that iron argument out the door with it.

The honest counter-case has two entries. First, foam hides the bath: temperature and sample confidence drop when you can't see or reach steel, and a shop that foams brilliantly but measures poorly gives part of the winnings back in re-blows. Second, the premium is real money — carbon beyond what converts to CO is just expensive slag conditioner, and oxygen beyond what your carbon answers becomes FeO, which is yield. The instrument pays when it's actively managed. Left on autopilot, it decays into a cost line with a good reputation.

Score each question 1 (yes) or 0 (no/poor), once per shift, from the panel and the platform. Under 7 means the position is leaking.

Trend the shift scores on the board next to kWh/t. In every shop we've seen do this, the two lines move together within a week — which is the whole argument of this issue, drawn by your own furnace.

Next week: the anatomy of tap-to-tap — where the minutes hide, and the wrong way to find them.

Written by active DRI-EAF operators. Anonymous by necessity, specific by design.