Case 05 · · 6 min read
Induction skull melting of red mud and ash-slag
Red mud from alumina plants and ash-slag from power stations are oxide mixes that are hard to melt cleanly. Induction skull melting melts this material in a water-cooled copper crucible, so the melt never touches the wall. The method is developed by Dr. Sc. A. V. Shkulkov. The experimental unit belongs to Era Group. The project is in progress. Below: the principle, the parameters, the checked results and the open questions.
01Red mud and ash-slag: why they are hard to process
Red mud is the residue of bauxite leaching in alumina production. Russia has accumulated about 600 million tonnes, and the volume is still growing. The mud contains iron, aluminium, titanium and calcium, plus scandium, gallium and rare earths. Rusal already produces scandium oxide from red mud on an industrial unit at the Urals aluminium plant.
Ash-slag from coal power stations is also oxides: silica, alumina, iron and calcium oxides, with unburnt carbon and rare metals. We wrote separately about turning Cherepetskaya GRES ash-slag into expanded clay.
To take rare metals out of such a mix, it often has to be melted first and put into a form an acid can reach. In a lined furnace the melt dissolves the wall: fireclay, graphite or refractory enter the material and change it. If the acid is then meant to pick out rare metals, contamination from the lining gets in the way.
02The principle: a cold crucible and a skull
The crucible is built from water-cooled copper segments. Gaps between the segments let the inductor’s electromagnetic field through. The field heats the material itself, and it melts from the inside.
At the cold wall the melt freezes into a crust of the same material, the skull. The crust keeps the bath off the copper. There is no lining, so nothing is there to contaminate the melt.
Working temperatures are 2000–2500 °C and higher. A lined furnace does not reach them. The cold crucible is protected by its own skull.
03Cold-crucible melting in practice
In English the method is Cold Crucible Induction Melting (CCIM). The best-known industrial case is the Orano plant at La Hague, France. In April 2010 the R7 vitrification line replaced a hot metal crucible with a cold crucible developed by CEA and AREVA. More than a thousand containers of vitrified waste have been produced on it since.
The developer dates the Russian line of the method to the late 1970s. A cold crucible is used in Russia to vitrify liquid radioactive waste, and also in Japan and India.
04The unit and its parameters
The experimental skull-melting unit was assembled by and belongs to Era Group. In October 2026 the stand held a stable melt: the bath was fully molten, and a skull had formed around the crucible.
- Laboratory stand
- 12–15 kW
- Frequency
- 5.28 MHz
- Batch
- ~2 kg
- Rate
- 2–3 kg/h
Pilot melts are sized for batches of 2 to 50 kg. Industrial units from 500 kVA at 300–500 kg/h are claimed by the developer and are not yet backed by a reference unit; there are no industrial drawings yet. For a sound design the developer estimates specific energy use at about 1 kWh per kilogram of melt.
05What has already been checked
Power-station ash-slag. A reducing melt produced an iron-silicon alloy (ferrosilicon).
Gabbro with soda. Karelian gabbro was melted with soda, quenched to an amorphous state and treated with 30 percent sulphuric acid. The solid residue was 92.3 percent SiO₂ at a solids yield of 53 percent; alkalis, aluminium, iron and rare earths went into solution. In the dried solution, europium and yttrium exceeded 600 ppm. The developer treats this as a direct analogue for ash-slag.
Electronic scrap. A 2004 study reported 96–98 percent recovery of precious metals.
There has been no direct melt of red mud yet. There are also no recovery figures for rare earths from the ash of a specific site: the analogue is gabbro. Use on red mud is a working hypothesis. The pilot will test it.
06What to do with the melt
After melting there are two paths. The first is a product at once: ferrosilicon, mineral wool, glass-ceramics, feedstock for geopolymers. The second is amorphisation: the melt is quenched into a glass-like matrix, and rare metals pass into solution in acid leaching.
In the Era Group deep-processing scheme each stage has its partner. Feedstock composition and beneficiation: IPKON RAS. Hydrometallurgy of rare metals: IPCE RAS and CRIRS. Gasification of unburnt carbon: the Institute of Fossil Fuels. Melting: the skull-melting unit, run by Era Group.
07Which feedstock fits skull melting
The method is for “dirty” oxide materials that a lined furnace either contaminates or cannot melt. The developer names power-station ash and slag, alumina-plant mud and electrolysis sludge, gabbro and eudialyte. All of them have a high melting point and a complex composition.
Melts have been run on complex oxides, electrolysis sludge and ash-slag. The regime for red mud still has to be found: it carries a lot of iron oxides and residual alkali, and their behaviour in the melt sets the matrix and the later leach.
A pilot melt answers three questions: whether the skull is stable on that feedstock, what matrix quenching produces, and how much of the valuable metals enters solution.
08Who developed the method
Anatoly V. Shkulkov holds a doctor of technical sciences degree. In 2006 at Saint Petersburg Electrotechnical University LETI he defended the thesis “Theory and practice of induction skull melting of inorganic dielectric materials”, specialty 05.09.10, electrotechnology. The work was done at the Vologdin Institute of High-Frequency Currents. He handles the theory himself: skull stability and melt regimes.
LETI was founded in 1886 and was the first civilian electrical-engineering school in Russia. A separate laboratory there works on induction heating.
09Status and the next step
The project is in progress. The next step is a pilot melt of 25–50 kg of real feedstock: red mud, or ash-slag from a specific site. That melt will show the regime, the matrix composition and the metal recovery in leaching.
Project economics on request: my@ostoya.ru.
FAQ
- How does skull melting differ from a lined furnace?
- The melt is held by a crust of its own material against cold copper walls. There is no lining, so the furnace material does not enter the melt.
- Has red mud already been processed this way?
- No. There are results on ash-slag, gabbro and electronic scrap. Red mud will be checked in a 25–50 kg pilot melt.
- What throughput does an industrial unit have?
- The developer claims 300–500 kg/h for units from 500 kVA. That figure is not yet backed by a reference unit.
- Where are the project economics?
- Project economics on request. Write to my@ostoya.ru.
Read also
Sources
- A. V. Shkulkov, Theory and practice of induction skull melting of inorganic dielectric materials, doctoral thesis, LETI, 2006
- Thesis abstract (the work was done at the Vologdin Institute of High-Frequency Currents)
- WM Symposia 2015: Cold crucible deployment in the La Hague facility
- CEA: cold crucible for HLW vitrification (more than 1,100 containers)
- RIA Novosti: about 600 million tonnes of red mud accumulated in Russia
- Rusal: scandium oxide from red mud