Foundry energy audits consistently show coreless induction melting at roughly 6-17 GJ per tonne of metal against gas-fired crucible furnaces operating in the same duty band, with the widest gap appearing on cold-charge, high-temperature alloys [S6].
The U.S. DOE ITP Metal Casting study found that newer furnace designs with larger and more efficient power supplies reduce kWh consumption per ton, and dual-output power supplies let one unit feed a melt and a holding leg simultaneously, cutting idle losses in two-furnace cells [S1].
Energy Per Tonne: The Numbers That Actually Matter
For aluminum melting, peer-reviewed plant data places gas-fired crucible furnaces at the upper end of the 6-17 GJ/tonne range, while electric and induction units cluster near the lower end once charge preheat and bath stirring are accounted for [S6][S8].
The same Salonitis dataset shows that crucible-furnace energy efficiency depends heavily on combustion-air preheat, burner tuning, and lid-open time during charging, variables that induction simply does not expose because heat is generated inside the charge via eddy currents [S3][S6].
On steel and iron, the ITP study documents that induction coreless furnaces with dual-output power supplies can run one coil in melt mode and a second in boost mode on the same bus, effectively raising kilowatt-per-tonne delivery without enlarging the furnace shell [S1].
Where Induction Wins, Where Gas Still Holds the Floor
Induction coreless units win on cold-charge melt rate, electromagnetic bath stirring, and indoor air quality, with the trade-off being transformer capacity, harmonics mitigation, and a closed-loop cooling water plant sized to the coil and busbar load [S5].
Gas-fired crucible furnaces remain attractive where cheap natural gas, large charge doors, dirty scrap, and high turndown ratios dominate, especially in aluminum recycling and secondary smelting, because radiant and convective burners tolerate variable charge density better than a fixed-frequency induction coil [S2].
For a foundry deciding between electric and gas, the comparison has to start with metal type and production volume, then move into utilities, controls, maintenance, layout, and how molten metal leaves the furnace, because the same nominal capacity label can describe very different equipment [S2][S5].
Selection Criteria: Metal, Volume, Utility, Emissions
Selection logic for any casting furnace line should run in this order: alloy grade and charge condition, batch versus hourly output, furnace body and pouring method, then plant power and cooling, because each downstream decision constrains the one above it [S5].
Steel, cast iron, copper, brass, bronze, aluminum and zinc do not use the same lining, temperature window, or fume-control practice, and a 500 kg crucible does not automatically produce 500 kg per hour; melt rate also depends on power, starting temperature, charge density and the complete tapping cycle [S5].
Gas-fired systems must be reviewed around burner layout, exhaust, preheat potential, operator access, and maintenance, while induction systems need transformer sizing, harmonics strategy, power cabinet layout, and a cooling-water specification tied to furnace duty, not just nameplate kW [S2][S5].
Decision Matrix: Induction vs Gas-Fired vs Electric Resistance
On energy per tonne, induction and electric resistance sit at the lower end of the 6-17 GJ/tonne band, gas-fired crucible at the upper end, with channel-type holding furnaces trimming further losses only when they are kept hot on a steady schedule [S6].
On emissions, induction melting produces fewer direct greenhouse gases at the stack, but the upstream grid carbon factor dominates, so a coal-heavy grid erases roughly half the emissions advantage of going electric [S4].
On control and melt quality, induction wins on temperature precision, bath stirring, and repeatability; gas-fired wins on charge flexibility and tolerance of dirty, oily, or mixed scrap; electric resistance sits between, with quiet operation and clean indoor airflow [S3][S4][S5].
On capital and infrastructure, induction demands transformer, harmonics filters, and a cooling water plant; gas demands gas train, burners, exhaust, and combustion-air handling; electric resistance sits closest to a plug-and-play install where the supply feeder is already in place [S2][S5].
Operational Pitfalls That Skew the Energy Comparison
Choosing equipment around a single best-case kWh-per-tonne number is a common error, because real plants have shift changes, cleaning time, variable scrap density, maintenance windows, and downstream pauses, all of which should be included before the equipment is sized [S2].
Treating preparation, melting, holding, and casting as separate decisions is the second classic mistake, because each step controls the next: if material preparation cannot feed the furnace, or if casting cannot receive the metal, the plant will still stop, regardless of furnace efficiency [S2][S5].
For gas-fired cells, lid-open time and burner excess-air ratio are the two largest hidden energy drains; for induction cells, coil water temperature rise and busbar cooling are the equivalent leak paths, and both must be instrumented before the kWh-per-tonne number is trusted [S1][S5].
Matching Furnace Type to Foundry Reality
Low-volume flexible work, medium batch production, and high-volume recycling do not need the same equipment, and a furnace that is ideal for one volume range may be inefficient or awkward in another, so buyers should size from the metal, the normal and maximum batch, the kilograms per hour, the tapping schedule, and the available electrical infrastructure rather than from a catalog headline [S2][S5].
For a plant weighing induction furnace investment specifically, channel inductors deserve a separate look, since a channel unit kept on a stable molten heel and paired with continuous casting behaves very differently from a coreless unit used for batch melting, and the energy math shifts accordingly [S1][S5].
Buyers comparing all three paths should send vendors the material source, expected output, alloy range, current bottleneck, available utilities, floor space, operator access, and the equipment already installed around the proposed furnace location; photos and simple measurements are often more useful than a specification written too early [S2].
Standards, Sourcing, and the Next Decision Node
Energy claims in this comparison should be read against the Salonitis 2016 dataset (cited 82 times) and the 2005 ITP Metal Casting report, which remain the two most-cited public references for foundry furnace energy baselines and which both flag 6-17 GJ/tonne as the realistic operating envelope for gas-fired crucible melting [S1][S6].
Buyers should also confirm that any vendor kWh-per-tonne figure is tied to a specific alloy, charge condition, and start temperature, because the same furnace can post 540 kWh/tonne on hot-charge aluminum and over 620 kWh/tonne on cold-charge steel, and the gap is rarely disclosed on a single line of a quotation [S1][S5].
Track the next utility tariff revision and the next local grid emissions factor update, since both can flip the induction-versus-gas math inside one budget cycle, and request a harmonics survey alongside any induction quotation because the answer often pushes the project into a soft-starter or active-front-end power supply.