Most people who work with ferro silicon know it as a steelmaking deoxidiser — the 70% or 75% grade that goes into the ladle to strip dissolved oxygen out of liquid steel. Ferro Silicon 15% does almost none of that. It is a different material with a different job, and the job has nothing to do with chemistry at all. FeSi 15% is bought for its density and its magnetism, and it spends its working life suspended in water, being recirculated thousands of times through a separation circuit that never actually consumes it.
That single fact reframes everything about how the grade should be specified, purchased, and judged. A steel plant buying FeSi 75% cares about silicon recovery and aluminium content. A diamond recovery plant or a coal washery buying FeSi 15% cares about particle shape, rheology, corrosion behaviour and magnetic recoverability — because every kilogramme that does not come back on the magnetic separator is a kilogramme it has to buy again next month. Understanding that difference is the whole point of this article.
1. What Ferro Silicon 15% actually is
Ferro silicon is produced by the carbothermic reduction of quartz (SiO2) with a carbon reductant — typically coke, char and woodchips — in a submerged arc furnace, with iron units added as scrap or mill scale. The silicon content of the tapped alloy is set by the charge balance. Standard commercial grades sit at 45%, 65%, 70% and 75% silicon. Ferro Silicon 15% sits at the far low-silicon end of that family, typically 14–16% Si with the balance essentially iron.
Metallurgically, that composition puts the alloy in a region of the iron–silicon system where it behaves much more like an iron alloy than a silicon alloy. It is strongly ferromagnetic, it has a true specific gravity in the region of 6.7–7.1 g/cm³, it resists rusting far better than plain iron powder, and it is hard and abrasion-resistant enough to survive being pumped around a plant continuously. Those four properties, taken together, are why nothing else has displaced it as the workhorse dense medium for the last seventy years.
Compare that to FeSi 45%, which has a lower density and much weaker magnetic response, or to FeSi 75%, which is barely magnetic at all and is brittle enough to degrade rapidly under recirculation. Neither can do this job. Magnetite (Fe3O4) can, but its lower density caps the achievable medium at roughly 2.5 g/cm³, which is fine for coal and useless for diamonds, chromite or base metal ores.
2. Why 15% silicon — and not 20% or 45%
The 14–16% window is not a marketing convention. It is the narrow band where three competing requirements overlap.
Density
You want the highest possible particle density, because the density of the medium you can stably maintain is limited by how much solid you can carry in suspension before the slurry becomes too viscous to separate cleanly. Adding silicon lowers the alloy density — silicon has a specific gravity of about 2.33 against iron's 7.87. So from a purely density point of view, you would want as little silicon as possible.
Corrosion resistance
You cannot use plain iron or steel powder, because it rusts. In a circuit that is a wet, aerated, abrasive environment running continuously, an iron powder medium would oxidise, generate fines, thicken the slurry, and progressively lose density. Silicon is what confers the corrosion resistance. Below roughly 12% Si, that protection degrades noticeably. So from a corrosion point of view, you want more silicon.
Magnetic recoverability
Every particle has to be pulled back out of the water by low-intensity magnetic separators — usually several stages of them — and returned to the mixing box. Recovery rates in a well-run circuit are above 99%. As silicon content rises, the saturation magnetisation of the alloy falls. By the time you reach 45% Si the material is a poor magnetic performer; at 75% it is essentially non-magnetic. So from a recovery point of view, you again want less silicon.
The net result: two of the three requirements push silicon down, one pushes it up, and the resolution lands at 14–16%. That is the entire reason Ferro Silicon 15% exists as a distinct commercial grade rather than as a point on a continuum.
3. Atomized vs. milled: the decision that drives your medium cost
Once the chemistry is fixed, the next decision is how the alloy is converted into powder — and this affects plant performance far more than most buyers expect.
Milled ferro silicon is produced by crushing and grinding solidified lump. The particles are angular and irregular, with a broad size distribution. Atomized ferro silicon is produced by breaking up a stream of molten alloy with high-pressure water jets, which freezes the droplets into rounded, near-spheroidal particles with a tightly controlled size distribution.
Shape matters because it governs rheology. Angular particles interlock and resist flow, so an angular medium reaches an unworkable viscosity at a lower solids loading. Rounded particles roll past one another, so an atomized medium stays fluid at a higher solids content — which means a higher achievable operating density and a sharper separation.
| Property | Atomized | Milled |
|---|---|---|
| Particle shape | Spheroidal / rounded | Angular, irregular |
| Medium viscosity at a given SG | Lower | Higher |
| Practical maximum medium density | Higher — suits diamond and heavy-mineral duties | Lower — ample for coal and most iron ore duties |
| Medium stability / settling | Better suspension stability at fine sizes | Settles faster; needs more agitation |
| Abrasion on pumps and cyclone liners | Lower | Higher |
| Drainage and rinsing on screens | Drains more freely; lower carry-over | Higher adhesion to product, more carry-over loss |
| Purchase price per tonne | Higher | Lower |
| Typical duty | Diamonds, chromite, base metals, fine cuts, high-density work | Coal washing, iron ore, coarse robust circuits |
The commercial lesson is that comparing atomized and milled on price per tonne is a category error. The number that matters is medium consumption in kilogrammes of FeSi per tonne of feed processed. A plant that switches to atomized and reduces carry-over loss on the drain-and-rinse screens can pay for the price premium several times over, while also gaining a sharper cut and lower liner wear. Equally, a high-throughput coal washery running a coarse cut at moderate density may find milled entirely adequate and the premium unjustified. Both answers are correct in their own context; what is not correct is choosing without measuring your loss rate.
4. Inside a dense medium separation circuit
Dense medium separation exploits the oldest principle in mineral processing: things denser than the fluid sink, things less dense float. The innovation is that the fluid is not water — it is a suspension of fine FeSi in water, whose apparent density can be tuned by changing the solids concentration. Coal circuits typically operate between about 1.3 and 1.8 g/cm³. Iron ore and manganese circuits run higher. Diamond and heavy-mineral plants operate at the top of the range, and it is there that atomized FeSi earns its premium.
A typical circuit runs as follows. Sized, de-slimed feed is combined with the medium in a mixing box. In a static bath the separation happens under gravity; in a cyclone circuit the slurry is pumped tangentially into a dense medium cyclone, where the centrifugal field — many times gravity — produces a much sharper cut and allows finer particles to be treated. Sinks and floats report separately to drain-and-rinse screens, where the bulk of the medium drains off for direct re-use and the residual film is washed off the product. That dilute wash medium then goes to magnetic separators, which recover the FeSi; the recovered medium is thickened back to working density in a densifier and returned to the circuit.
Two operational details deserve attention because they are where medium is actually lost. First, de-sliming the feed: clays and fines that report to the medium raise its viscosity and degrade the cut, and the correction — bleeding medium off — is a direct loss. Second, magnetic separator condition: worn drums, incorrect feed presentation and inadequate demagnetising coils all quietly increase losses. Circuits that appear to have a "poor quality FeSi" problem very often have a magnetic separation problem instead.
5. Specifying FeSi 15%: chemistry, sieve analysis, and the numbers that matter
A complete purchase specification for FeSi 15% should cover four blocks.
Chemistry
Silicon in the 14–16% band is the primary control. Beyond that, the elements worth capping are aluminium (which raises the risk of hydrogen generation in the presence of moisture and alkalinity), carbon, sulphur and phosphorus. For a DMS duty these are secondary to the physical properties, but a chemistry that drifts is usually a symptom of a process that is not under control.
Particle size distribution
This is the most important physical parameter and should be specified as a full sieve analysis, not a single number. Grades are commonly named by their fineness — coarser cyclone grades for coarse feed and lower densities, finer grades where high medium densities and fine cuts are required. Ask for the full cumulative distribution including the sub-45 µm fraction, because that fraction dominates both medium stability and viscosity.
Physical properties
Apparent (bulk) density, true specific gravity, and moisture content. Moisture matters both commercially — you are buying water — and technically, since a caked or partly oxidised powder disperses poorly on first make-up.
Magnetics
The proportion of the material that is recoverable on a low-intensity magnetic separator. This is the parameter most closely tied to your operating cost and the one most often omitted from purchase specifications. If you are qualifying a new supplier, ask for it explicitly.
A note on comparing quotations: two suppliers quoting "FeSi 15%, DMS grade" can be offering materially different products. Always compare on the full sieve analysis, the atomized/milled route, the magnetics figure and the moisture basis — then normalise the price to a per-tonne-of-feed-processed basis using your own measured medium loss.
6. Beyond DMS: other uses for low-silicon ferro silicon
Dense medium separation dominates FeSi 15% demand, but it is not the only outlet. The same combination of high density and reasonable corrosion resistance makes the material useful as a heavy aggregate — in high-density concrete for radiation shielding and for counterweights and ballast, where mass in a confined volume is the design constraint. It also appears in some oil well cementing formulations for the same reason, as a weighting agent, and in specialised heavy fillers.
These are genuinely different specifications. A shielding-concrete customer cares about bulk density, gradation and freedom from deleterious material, and is largely indifferent to magnetics. It is worth stating the end use when you enquire, because the optimal grade is not the same one.
7. A practical buying and handling checklist
- State the duty, not just the grade. "FeSi 15% for a chromite DMS cyclone circuit operating at 3.0 g/cm³" gets you a far better recommendation than "FeSi 15%, 100 MT".
- Ask for the full sieve analysis and the sub-45 µm fraction, and hold the supplier to it batch to batch. Consistency between lots matters more than any single lot's numbers.
- Measure your own medium loss in kg/t of feed before and after any change of supplier or grade. This is the only honest basis for comparing offers.
- Audit the magnetic separators before blaming the powder. Drum wear, feed presentation, and demagnetising coil function account for a large share of unexplained losses.
- Store it dry and sealed. Moisture ingress causes caking and surface oxidation, and in the presence of moisture and alkaline conditions, ferro silicon powders can evolve hydrogen and phosphine — which is why sealed packaging, ventilated storage and the relevant transport classifications are not optional formalities.
- Match packaging to your make-up practice. 25 kg and 50 kg bags suit manual charging and small top-ups; 1 MT jumbo bags suit bulk make-up with a hoist and a slitting station. AIM ALLOYS supplies 25 kg bags, 50 kg bags, 100 kg drums and 1 MT jumbo bags.
- Qualify on a trial campaign, not a sample. A 200 kg drum tells you the chemistry; only a full campaign tells you the rheology and the loss rate.
The bottom line
Ferro Silicon 15% is one of the few industrial materials where the purchase price is a poor guide to the cost. It is bought once and used continuously, so everything that determines how much of it survives each pass through the circuit — particle shape, size distribution, corrosion resistance, magnetic recoverability — is worth more attention than the number on the quotation. Specify it as a physical product rather than a chemical one, measure your losses, and the grade will repay the attention.