Dense medium separation simultaneously raised phosphate grade and rejected silica — and proved the tailings recyclable.
| Product | P₂O₅ % | P₂O₅ rec. % | SiO₂ % | SiO₂ rec. % |
|---|---|---|---|---|
| Concentrate | 29.92 | 96.26 | 8.34 | 65.72 |
| Rejects | 8.48 | 3.74 | 11.29 | 34.28 |
| Feed | 27.34 | 100.00 | 11.34 | 100.00 |
For phosphate, the payoff is twofold: grade above the commercial threshold and silica below the acid-plant penalty line. Density separation achieves both in a single pass, without reagents.
Guizhou hosts China's largest phosphate province, dominated by sedimentary collophane ores where silica is the persistent enemy: every point of SiO₂ in concentrate costs acid-plant efficiency and incurs commercial penalties. The task was therefore two-dimensional from the start — raise P₂O₅ and depress SiO₂ in the same operation.
The single density cut delivered both objectives: P₂O₅ up 2.58 points to 29.92% at 96.26% recovery, SiO₂ down 3.0 points to 8.34% with 34.28% of the silica rejected. A 29.9% concentrate sits comfortably above common commercial-concentrate thresholds (≥28–30% P₂O₅), achieved without reagents — leaving the full flotation toolbox in reserve if a premium product is later required. The rejects at 8.48% P₂O₅ additionally proved lean enough to consider for construction-materials use.
| Item | Value |
|---|---|
| Feed size fraction | 0.5–12mm |
| Separator | Pressurized two-product dense medium cyclone |
| Medium | Ferrosilicon + magnetite powder, closed-circuit recovery |
| Density control | In-line density gauge with automatic water/medium make-up loop |
| Value mineral density | Collophane / apatite: 3.0–3.2 g/cm³ |
| Gangue density | Quartz: 2.65 g/cm³ |
| Process water | 0.1 m³/t, fully recirculated |
| Medium drainage | Sieve bends + drain-and-rinse screens |
The ore was first wet-screened to remove the −0.5 mm slimes (evaluated separately for flotation). The 0.5–12mm fraction was fed to a ferrosilicon/magnetite suspension; a density-gradient series located the optimum cut density, after which the sample was passed once through a pressurized two-product cyclone. Products were drained on sieve bends, rinsed on drain-and-rinse screens, dewatered, sampled and assayed to compute yield, grade and recovery. Dilute medium was recovered by magnetic separation and densified back to the correct-medium sump under closed-loop density control.
For integrated phosphate-chemical producers the arithmetic runs through the acid plant: less silica per tonne of P₂O₅ means less energy, less wear and fewer penalties. And because the test also characterized the tailings, it doubled as a feasibility screen for the companion question — what value is still sitting in Guizhou's legacy tailings piles.
Q: What recovery did dense medium separation achieve on this phosphate ore?
A: 96.26% P₂O₅ recovery into a 29.92% concentrate, with SiO₂ cut from 11.34% to 8.34% (0.5–12 mm).
Q: Will this work on my phosphate deposit?
A: Density separation depends on liberation size and density contrast, not on locality. A one-tonne representative sample answers the question definitively — washability results within 24 hours of sample receipt.
Send ~1 tonne of representative ore — washability results within 24 hours of sample receipt.
Contact us