Does a spectrometer for checking molten metal save power in an induction foundry?
Yes, a little. The BEE and SAMEEEKSHA project report for the Batala, Jalandhar and Ludhiana foundry cluster puts it at about 10 kWh per tonne of molten metal, from shorter holding while the sample is tested. At 10,000 tonnes a year that is 100,000 kWh.
Where the saving comes from
In a large number of plants in the cluster the molten metal is sampled and tested in a laboratory. That takes about 30 minutes, during which the furnace holds the metal. A spectrometer reads the sample in about 5 to 10 minutes, which cuts the analysis time by 60 to 70%. Less holding means less power.
The report's baseline
- Furnace: 4 to 5 tonne dual track induction.
- 250 operating days, 1 hour batch time, 10,000 tonnes a year.
- Specific power: 590 to 600 kWh per tonne, so 10 kWh is under 2% of melting energy.
- Process downtime: nil. Suitable for large furnaces.
- Recommended: at least one sample per batch.
Be honest about the size
This is a small saving next to the melt itself. The report also credits better product quality. If your lab already tests fast, or you run small furnaces, the saving shrinks.
Check it on your own meter
Log holding minutes per heat before and after, and the furnace kWh per tonne for a month each. The report's figure is an estimate, not a measured result.
Worked CO2 from the report's numbers
- Saving: 10 kWh per tonne x 10,000 tonnes = 100,000 kWh a year (100 MWh).
- The report shows 81 t CO2 a year, which implies about 0.81 t per MWh.
- At the FY 2025-26 grid factor of 0.675 kg CO2 per kWh: 100,000 x 0.675 = 67,500 kg, or 67.5 t.
- Use the factor for the year of your own bills, not the report's older one.
The report dates from 2011.
What Senseible does here
Senseible reads grid electricity bills into a monthly kWh record per site, so a before and after comparison for a change like this comes from the bills, not memory.
Last reviewed 2026-11-03.