An Electric Furnace is an enclosed insulated chamber that uses electrical energy — not combustion — to heat a charge to a controlled temperature. The defining component is the heating element (resistance wire, induction coil, or graphite electrode), which converts electrical current into heat by I²R losses, electromagnetic induction, or arc plasma. We use electric furnaces wherever combustion atmospheres are unacceptable or precise temperature control matters: heat treating tool steel, melting aluminium, sintering ceramics, and producing steel from scrap. A modern 100-tonne electric arc furnace melts a full charge in about 40 minutes drawing 80–120 MW.
Electric Furnace Power Sizing Interactive Calculator
Vary charge mass, temperature rise, melt time, efficiency, and heat loss to see required furnace power and 250 kW melter loading.
Equation Used
This calculator sizes the electrical input power needed to heat and melt an aluminium charge. Sensible heat raises the metal temperature, latent heat melts it, efficiency converts thermal demand to electrical demand, and steady heat loss is added directly as power.
- A380 aluminium properties are fixed at cp = 963 J/kg-K and Lf = 389000 J/kg.
- Latent heat is included, so the calculator represents a melting case.
- Nameplate load is compared against the 250 kW induction melter from the worked example.
- Heat loss is treated as an average steady power loss during the melt.
How the Electric Furnace Actually Works
Three families of Electric Furnace exist, and they all turn electricity into heat by different physics. A resistance furnace pushes current through a high-resistivity element — Kanthal A1, Nichrome 80/20, silicon carbide, or molybdenum disilicide — and the I²R loss in that element radiates and convects into the chamber. An induction furnace wraps a water-cooled copper coil around a refractory crucible and uses an alternating magnetic field, typically 50 Hz to 10 kHz, to drive eddy currents inside the metal charge itself. An electric arc furnace strikes a plasma arc between graphite electrodes and the scrap charge, dumping enormous power densities — 1 MW per tonne is normal — directly into the metal.
The chamber itself is built around the heat path. You have a refractory lining (alumina, magnesia, or silica brick depending on chemistry), backing insulation (ceramic fibre or vermiculite), and a steel shell. A thermocouple — Type K to 1260°C, Type S or B above that — feeds a PID temperature controller that switches the elements via SSRs or thyristors. If the thermocouple drifts even 15°C high, you'll under-temper a batch of D2 tool steel and the customer will reject it on hardness. If the refractory cracks and lets heat reach the shell, you'll see shell temperatures climb past 80°C and element life drops because the chamber can't hold setpoint without overdriving the elements.
The most common failure modes are element burnout from local hot spots (usually caused by element-to-element contact after sag), thermocouple sheath degradation in oxidising atmospheres, and refractory spalling from thermal shock when an operator opens the door at 1100°C. Heat the chamber slowly — 150°C per hour on a cold start for a brick-lined furnace — and you double element life.
Key Components
- Heating Element: Converts electrical current into heat. Kanthal A1 wire works to 1400°C surface temperature, SiC rods to 1600°C, MoSi₂ elements to 1800°C. Element watt-loading must stay below the manufacturer's curve — typically 2.5 W/cm² for Kanthal at 1200°C — or the element sags, contacts a neighbour, and burns out.
- Refractory Lining: Insulates the hot face and contains the charge. Alumina-silica firebrick handles 1500°C, magnesia brick is required for basic-slag steelmaking. Lining thickness sits between 75 and 230 mm depending on furnace size; too thin and the shell overheats, too thick and you waste energy heating the lining mass each cycle.
- Thermocouple and Controller: A Type K, S, or B thermocouple feeds a PID controller that modulates element power through an SSR or thyristor stack. Calibration drift above ±5°C at 1000°C is the usual reason heat-treat batches fail spec — recalibrate against a reference probe every 6 months.
- Power Supply: Resistance furnaces run direct on 240 V or 480 V mains through a contactor or thyristor. Induction furnaces need a solid-state inverter producing 500 Hz to 10 kHz. Arc furnaces use a step-down transformer rated 30–150 MVA feeding three graphite electrodes.
- Shell and Door Seal: Carbon-steel shell carries the structural load and anchors the lining. Door seals — ceramic rope or fibre gasket — must compress evenly; a 3 mm gap at the door edge leaks enough heat to drop chamber temperature 40°C at the far wall and skews uniformity surveys.
Who Uses the Electric Furnace
Electric Furnaces show up wherever you need clean, controllable, atmosphere-free heat. They dominate heat treatment, precious-metal melting, sintering, and any process where combustion products would contaminate the work. They also dominate steel recycling — roughly 70% of US steel now comes from electric arc furnaces feeding on scrap, because the energy and capital math beats the integrated blast-furnace route at smaller scales.
- Steelmaking: Nucor Steel runs 200-tonne EAFs at its Berkeley County mill, melting scrap in 45-minute heats drawing roughly 100 MW per furnace.
- Heat Treatment: A captive heat-treat shop running an L&L Special Furnace XLE box furnace tempers H13 die-cast tooling at 595°C ±5°C for 4-hour soaks.
- Investment Casting: Inductotherm VIP induction melters at a turbine-blade foundry vacuum-melt Inconel 718 in 50 kg crucibles at 1450°C before pouring into ceramic shell moulds.
- Laboratory and Materials Research: A Carbolite-Gero RHF tube furnace sinters zirconia dental ceramics at 1530°C with a 1°C/min ramp and a 2-hour dwell.
- Glass and Ceramics: Skutt and Paragon kilns fire stoneware at 1280°C cone 10 in studio potteries; SiC element kilns push to cone 12 for porcelain.
- Semiconductor: Diffusion furnaces from companies like Tempress and ASM oxidise silicon wafers at 1000–1200°C with ±0.5°C uniformity across a 200-wafer load.
The Formula Behind the Electric Furnace
The core sizing question for any Electric Furnace is: how much electrical power do you need to heat the charge from room temperature to setpoint in a target time, accounting for losses? At the low end of the typical operating range, with a small lab furnace and a slow ramp, you're dominated by chamber losses through the lining. At the high end — a foundry melter on a 30-minute cycle — you're dominated by the energy stored in the molten charge and the latent heat of fusion. The sweet spot for industrial heat-treat is usually a furnace sized so steady-state losses are 30–40% of nameplate power, leaving the rest as headroom for ramp.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| P | Required electrical input power | W | BTU/hr |
| m | Mass of the charge | kg | lb |
| cp | Specific heat capacity of the charge | J/(kg·K) | BTU/(lb·°F) |
| ΔT | Temperature rise from start to setpoint | K | °F |
| Lf | Latent heat of fusion (melting only) | J/kg | BTU/lb |
| η | Furnace thermal efficiency | dimensionless | dimensionless |
| t | Target heating time | s | hr |
| Qloss | Steady-state heat loss through walls, door, openings | W | BTU/hr |
Worked Example: Electric Furnace in a captive aluminium die-cast melter
A medium-volume die-cast shop in Windsor Ontario runs an Inductotherm VIP-I 250 kW induction melter holding 500 kg of A380 aluminium alloy. The shop wants to melt a cold 500 kg charge from 25°C to a 720°C pouring temperature in 60 minutes. Specific heat of A380 is roughly 963 J/(kg·K), latent heat of fusion is 389 kJ/kg, melting point 595°C, furnace thermal efficiency 72%, and steady-state shell losses are measured at 18 kW.
Given
- m = 500 kg
- cp = 963 J/(kg·K)
- ΔT = 695 K
- Lf = 389,000 J/kg
- η = 0.72 —
- t = 3600 s
- Qloss = 18,000 W
Solution
Step 1 — sensible heat to bring 500 kg of A380 from 25°C to 720°C:
Step 2 — latent heat of fusion to melt the charge:
Step 3 — total useful energy, divided by efficiency and time, plus standing loss, at the nominal 60-minute cycle:
That fits comfortably under the 250 kW nameplate. At the low end of the typical operating window — a lazy 90-minute melt — required power drops to roughly 154 kW. The melt is gentle, refractory life is excellent, but the shop loses 30 minutes of casting throughput per cycle. At the aggressive high end, a 40-minute melt, the math demands 326 kW — well above what the 250 kW unit can deliver, so in practice the operator either trims the charge to 380 kg or accepts a longer cycle. The sweet spot sits right around 55–65 minutes for this furnace.
Result
Required input power at nominal cycle is about 222 kW, which leaves roughly 11% headroom under the 250 kW Inductotherm rating. At a slow 90-minute cycle the unit only needs 154 kW and runs cool; at an aggressive 40-minute cycle the math calls for 326 kW which the inverter can't supply, so the cycle stretches. If the operator measures actual melt time 25% longer than 60 minutes, look at three things first: (1) coil-to-crucible coupling — a worn refractory crucible widens the air gap and drops electromagnetic efficiency below 65%, (2) capacitor bank tuning drift on the inverter, which detunes the resonant frequency and dumps real power into the coil cooling water rather than the charge, or (3) a charge contaminated with painted scrap, where surface oxides and organics absorb energy as smoke instead of melting metal.
Electric Furnace vs Alternatives
Picking between an Electric Furnace and a fuel-fired alternative comes down to atmosphere control, energy cost, and process temperature. Electric wins on cleanliness and uniformity. Gas wins on operating cost in regions with cheap natural gas and on raw thermal power per dollar of capital.
| Property | Electric Furnace | Gas-Fired Furnace | Oil-Fired Furnace |
|---|---|---|---|
| Maximum practical temperature | 1800°C (MoSi₂), 3000°C+ (arc) | 1650°C (regenerative burner) | 1500°C |
| Temperature uniformity (typical) | ±3°C to ±5°C | ±10°C to ±15°C | ±15°C to ±25°C |
| Thermal efficiency at setpoint | 65–80% | 30–55% | 25–45% |
| Energy cost per kWh delivered (USA avg) | $0.10–$0.15 | $0.04–$0.07 | $0.08–$0.12 |
| Atmosphere control | Excellent — vacuum, inert, reducing all viable | Limited — combustion products always present | Poor — sulphur and soot contamination |
| Capital cost (per kW installed) | $300–$800 | $150–$400 | $200–$500 |
| Element/burner service life | 2,000–10,000 hr (resistance) | 20,000+ hr (burner) | 15,000 hr (burner) |
| Best application fit | Heat treat, sintering, vacuum melt, EAF steel | Forging, reheating, large continuous lines | Remote sites without gas supply |
Frequently Asked Questions About Electric Furnace
You're seeing thermal stratification. The control thermocouple sits in one location — usually mid-chamber on the back wall — and the PID is happy when that single point reads setpoint. But the back corners, especially behind a dense load of fixtures, can run 30–60°C colder if convection and radiation paths are blocked.
Run a 9-point uniformity survey per AMS 2750 with thermocouples in the load. If the cold spot is more than ±15°C from setpoint at 600°C, you need to either add a recirculating fan, redistribute the load with airflow gaps, or move the control thermocouple closer to the cold zone and bias the recipe.
Throughput and metal quality drive the call. Induction stirs the bath electromagnetically, which homogenises alloy composition and pulls dross to the surface — useful when you're remelting mixed scrap. Resistance crucible furnaces are quieter, cheaper to buy (roughly half the capital), and tolerate dirty charges better, but melt cycles run 1.5–2× longer for the same kW input.
For a jobbing foundry doing 4+ heats per shift, induction pays back the capital inside 18 months on throughput alone. For a sculpture studio doing one heat per day, a resistance tilting furnace like a Morgan F-series is the smarter buy.
MoSi₂ — Kanthal Super or similar — fails fast under three specific conditions, none of which appear in the catalogue lifetime curve. First, cycling below 600°C in a humid atmosphere causes pest oxidation, where the silica protective layer breaks down and the element disintegrates. Hold above 1000°C or fully cold; don't park at 400°C overnight.
Second, contact with refractory dust containing iron or aluminium contaminates the silica skin and creates low-melting eutectics. Third, watt-loading above 8 W/cm² shortens life dramatically — check that the controller isn't over-driving the elements during ramp. A current clamp on the element leads during ramp will tell you in 30 seconds.
Technically yes for low-melt alloys like aluminium or zinc, but the failure modes bite hard. Heat-treat furnaces are sized for radiative heat transfer to solid loads, not the convective demand of a liquid bath. Element watt-loading was specced assuming the chamber atmosphere, not a 50 kg thermal sink that pulls power continuously.
You'll also have a real problem the first time the crucible cracks — molten aluminium against Kanthal elements is a $4000 element-pack replacement and a chamber rebuild. Use a purpose-built crucible furnace with bottom-pour or tilting geometry and a sacrificial spill tray.
Current at rated value with slow ramp means the elements are converting power to heat, but that heat isn't reaching the load efficiently. The most common cause on a furnace 2+ years old is degraded backing insulation — ceramic fibre shrinks and pulls away from the hot face, creating a convective short between the lining and the shell that dumps 20–40% of element output into shell losses.
Check shell temperature with a thermal camera. If the shell exceeds 80°C anywhere, the lining is compromised. Second-most-common cause: a door seal gap. Run a dollar-bill test along the seal at the cold furnace and replace the ceramic rope if the bill pulls free anywhere.
Medium-frequency wins on small charges, mains-frequency wins on large ones. At 1 kHz the skin depth in molten steel is roughly 15 mm, so a 100 kg crucible couples efficiently — total electrical-to-thermal efficiency hits 75–80%. At 60 Hz the skin depth is closer to 70 mm, which means a small charge barely couples and efficiency falls below 50%.
The crossover is around 500 kg charge mass. Above that, the lower capital cost and simpler power supply of mains-frequency starts to win. Below it, medium-frequency is the right call every time, which is why nearly all modern jobbing foundries under 1 tonne run medium-frequency.
References & Further Reading
- Wikipedia contributors. Electric furnace. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.