Hot Metal
Three things worth your attention this week.
Every efficiency pitch you'll hear this quarter quotes a kWh/t saving; almost none quotes the boundary. Furnace-only or shop-wide, cold charge or hot, with or without the chemical side counted — the same physical furnace can honestly report four different kWh/t figures depending on where you draw the line. Before signing anything off a slide, ask which line the vendor drew. The saving usually lives in the definition.
Industrial power prices remain the biggest external swing factor on EAF conversion cost, and the direction isn't friendly: the IEA's mid-year update puts EU and Japanese spot wholesale electricity up more than 30% year-on-year in the second quarter, with European industrial power still running about twice US levels. The operator read: a shop that knows its energy balance can move load intelligently when the contract rewards it; a shop that only knows its meter total can't tell which load is movable.
Algoma Steel's results on July 29 put its second EAF at "nearing completion," first steel expected this quarter; the first unit is already running a full 24-hour schedule on Ontario grid power. An integrated shop finishing its EAF conversion is about to relearn its energy accounting from scratch. The coke rate retires; the meter — and everything below — takes its place.
The 40% that never reaches the steel
Start from the physics, because it's the only number nobody can argue with: heating iron from ambient to liquid steel at tapping temperature takes roughly 380 kWh per tonne. That's the floor. Published EAF energy balances — Pfeifer and Kirschen's survey is the standard reference — put total energy input for real furnaces broadly between 510 and 880 kWh/t, clustering in the mid-600s. Somewhere around 40% of what you buy, in electricity and in oxygen, carbon and gas, never reaches the steel.
The balance says where it goes, and published balances agree on the ranking:
Off-gas, 15-20% of input. The biggest sink, and the one most inflated by bad habits: air pulled in through the slag door and roof gaps burns nothing useful and hauls heat up the elbow. Door discipline is an energy program disguised as housekeeping.
Cooling water, roughly 10%. The panels are doing exactly their job — intercepting arc radiation that missed the bath. The fix is not colder water; it's covering the arc so less radiation escapes in the first place. Hold that thought for next week.
Slag, 8-10%. Every kilogram of slag leaves at bath temperature. This is last week's argument wearing an energy hat: slag mass you didn't need is enthalpy you paid to heat and iron you paid to lose.
Electrical losses and radiation, mid single digits. Real, mostly fixed, and the last place to hunt.
A DRI charge moves both ends of this ledger. On the debit side, every point of metallization below nameplate is oxide the furnace must finish reducing — published rules of thumb cluster around 12-15 kWh/t per point — and gangue is slag mass with all the enthalpy that implies. On the credit side sits the biggest single energy lever published anywhere: hot charging. Published figures from the major DRI process suppliers put the saving from charging at around 600°C in the neighborhood of 120-150 kWh/t against cold DRI. No arc-regulation upgrade on the market gets close to that number.
One boundary trap deserves its own sentence before anyone benchmarks: those published totals mix electrical and chemical energy, and shops quoting "kWh/t" often mean electrical only — with oxygen, carbon, and gas doing unacknowledged work off the books. Comparing your electrical-only number against a neighbor's total-energy number flatters exactly one of you, and never usefully.
Now the part the meter actively hides. Power-off time looks free — the meter isn't spinning — but the bath radiates whether or not you're paying attention, and every power-off minute is heat you'll buy back at the start of the next power-on, plus the productivity bill on top. Shops we've seen chase kWh/t as a standalone number have a way of "improving" it while total cost per tonne gets worse, because kWh/t doesn't see the delay minutes. The honest metric pair is kWh/t and power-off minutes, reviewed together, per crew.
The counter-case, so nobody over-corrects: not every loss on the balance is worth attacking. Off-gas heat recovery is capital-heavy and earns its own skeptical treatment another week. The first-order wins are nearly free — door discipline, honest delay logging, arc coverage, and charging DRI as hot as your material handling allows. Do the cheap 80% before anyone shows you a heat-exchanger brochure.
Operator's Notebook — six signals to read in the energy log first
Skip the averages. Open the log and look for these, in this order:
Spread between crews, same charge mix. If kWh/t varies by more than a few percent shift-to-shift on comparable heats, your cheapest energy project is a conversation, not equipment.
kWh/t drift at constant mix. Slow upward creep with no charge explanation usually means secondary-circuit condition or foam quality decaying — both measurable, both fixable.
Power-off minutes vs logged delays. Subtract logged delays from total power-off. The unexplained remainder is your most expensive unowned line item.
First-heat-after-idle penalty. Compare heats after a long power-off with steady-sequence heats. The gap tells you what thermal discipline is worth in your shop, in your numbers.
Tap temperature distribution. A wide distribution means somebody is buying insurance superheat with your kWh. Every unneeded degree was paid for twice — once to make it, once downstream.
Oxygen and carbon totals vs standard. Chemical energy that deviates from standard without a logged reason is the same problem as unexplained kWh — it just hides on a different invoice.
Any one of these takes fifteen minutes at handover. All six take an hour and typically pay for the year.
Next week: foaming slag as a financial instrument — what arc cover is actually worth, and what you pay for it.
Written by active DRI-EAF operators. Anonymous by necessity, specific by design.