Coal prep plants are squeezing more water out of fine tailings with high‑rate thickeners, belt filter presses, and solid‑bowl centrifuges—and they only hit spec with specialty flocculants and coagulants. The payoff: stackable solids and clear supernatant water ready for recycle.
Coal preparation generates large volumes of fine, water‑laden tailings—often more than 20% of plant feed is smaller than 0.5 mm (scielo.org.za). After settling in ponds, those slurries typically still carry 65–75% water by weight (scielo.org.za). The industry answer: concentrate tailings and recover water via high‑rate thickeners, belt filter presses (vacuum belt filters), and solid‑bowl (decanter) centrifuges—paired with specialty flocculants and coagulants—to deliver high underflow solids and a clear supernatant suitable for recycling (scielo.org.za; mdpi.com).
High‑rate thickening performance
High‑rate thickeners—large, raked, sloped‑floor clarifiers—routinely produce 30–50% solids (w/w) underflow; in paste configurations they push to 60–70% solids (westechwater.com; mclanahan.com). Overflow clarity is typically strong—often below 10–50 NTU (nephelometric turbidity units) with polymer; lab tests on coal tailings have hit 7.4 NTU (scielo.org.za).
These units scale to thousands of cubic meters per day with high liquid recovery (90%+ of feed volume), though they carry a large footprint (mclanahan.com; westechwater.com). In one iron‑ore analogue, a high‑rate thickener delivered roughly 45–50% solids versus 65–70% from a paste thickener equipped with steeper floors and pickets (westechwater.com). Plants often first thicken coal tailings to ~30–40% solids to reduce downstream loads. In practice, this gravity step aligns closely with a flocculant‑aided clarifier; process trains commonly deploy equipment comparable to a clarifier as the primary stage.
Belt filter press dryness and clarity
Vacuum belt filters (belt filter presses) produce the driest cakes among the three, commonly 75–85% solids (i.e., 15–25% moisture) in mining service (mdpi.com). Filtrate is extremely clear: pilot data report less than 0.6% solids in filtrate, with typical turbidity very low (often below 5 NTU) after polymer conditioning (researchgate.net).
Throughput spans roughly 0.3–1 m³/m²·h of suspension, with modular units offering 30–200 m² of filtration area (mdpi.com). Stable, well‑flocculated feed is critical: raw tailings are typically coagulated/flocculated in a pre‑thickener or conditioning tank, then fed slowly for gravity drainage, low‑pressure nips, and high‑pressure rollers. Cake is stackable and conveyable, though at 15–25% moisture it remains only partially dewatered and may need further drying or washing.
Solid‑bowl centrifuge trade‑offs
Solid‑bowl (decanter) centrifuges use high G (gravitational acceleration) to accelerate separation, yielding 65–76% solids products (24–36% moisture) in coal‑refuse tests (researchgate.net). Centrate quality is moderate—0.5–10% solids depending on settings—and generally less clear than belt‑filter filtrate (researchgate.net).
Capacity is continuous and compact (fully enclosed units, small footprint), scaling to 300+ kg/h solids in published baselines, but energy demand and sensitivity to feed conditions are higher (researchgate.net). Cake dryness tends to improve with higher feed solids or throughput, often at the expense of solids recovery; studies report centrate solids of 0.5–11% with overall solids recovery between 65% and 98% (researchgate.net). When operated at high G and low feed rate, recoveries of ~90%+ are typical, with product solids ~65–75%; any undebrained water and fine clays exit as dilute centrate requiring further treatment or recycle.
Side‑by‑side performance snapshot
For coal tailings, the belt filter press delivers the driest cake (75–85% solids) and the clearest water—filtrate can be approximately 1,000× cleaner than feed in pilots (mdpi.com; researchgate.net). Solid‑bowl centrifuges can approach that cake dryness (often ~70% solids) but typically lose more solids to the effluent (researchgate.net). High‑rate thickeners yield coarser underflow—typically 30–50% solids—while maximizing liquid recovery with minimal energy; paste thickeners push to ~65% solids (westechwater.com; mclanahan.com; mclanahan.com). In flowsheets, thickener plus centrifuge is often paired for bulk water recovery followed by final dewatering, while belt presses are chosen where dry‑stack tailings are targeted.
Flocculants and coagulants in control
Fine coal and clays carry net surface charges that resist settling; the industry relies on polymeric flocculants to aggregate these fines. The standard class is polyacrylamide‑based copolymers—non‑ionic, anionic, cationic, or blends—sold under labels such as Nalco AN934, FLOCAN 6815, and Fisher FLOCAN FO‑4700 (mdpi.com). These high‑molecular‑weight polymers “bridge” particles into large flocs for rapid gravity or high‑G separation. Plants source these as flocculants for thickening and filtration performance.
Coagulants—multivalent salts or cationic polymers—often precede polymer dosing to neutralize charge and form micro‑flocs. Typical choices include ferric chloride (FeCl₃), aluminum chlorohydrate (referred to as PAC in the cited review), ferrous sulfate, or polymeric coagulants such as Polyaluminium Chloride and polyDADMAC (mdpi.com). A small addition—on the order of 5–50 mg/L—primes the slurry, lowers polymer demand, and improves overflow clarity. Many sites procure these as coagulants; aluminum chlorohydrate is available as ACH. Plants also deploy polyaluminium chloride as PAC in similar roles.
Quantitatively, the chemistry is decisive. Bench tests on dilute coal tailings (~8% solids) reported 32.5 g polymer per tonne yielding settling at 178 mm/min and a filtrate turbidity of 7.4 NTU (scielo.org.za). As a regulatory marker, Indonesian limits often require less than 50 mg/L suspended solids in discharges. In belt‑filter operations across mineral tailings, flocculant usage of about 20–40 g/t solids supports continuous filtration at 15–25% cake moisture (mdpi.com).
Selecting the dewatering chemical program
Laboratory screening sets the baseline. Characterize tailings (pH, particle size, zeta potential, conductivity), then run jar tests to sweep coagulant dose and pH, followed by polymer selection and mixing to track settling velocity and final turbidity. Key metrics include settling rates in mm/min (or cm/s) and supernatant clarity in NTU; for coal slurries, targets often exceed 100 mm/min initial settling and less than 10 NTU final turbidity. Polymer molecular weight and charge (anionic vs cationic) are tuned to maximize floc size, at the lowest feasible dose (often dosed per tonne of solids).
Bench‑scale dewatering validates equipment performance: column tests for thickener underflow density, or lab vacuum/belt cells for cake moisture and filtrate clarity. Modern tools include the LUM (Lumby) analyzer for separation dynamics and the Capillary Suction Timer (CST) as a dewatering index. Tailings guidelines highlight bench “SuperFlo” settling tests to pinpoint “optimum flocculant type and dosing rate” before pilots (tailings.info).
Pilot and field trials then translate lab picks into plant settings: mobile thickeners or sample‑return tests to confirm feedwell addition points, rake speed, underflow percent solids, and overflow clarity over time; belt‑press or centrifuge rigs to optimize feed rate, polymer mixing points, and cake dryness. Outputs are quantified—cake moisture and compressibility, filtrate/centrate turbidity and TSS (total suspended solids), and any recovery shifts—while confirming polymer solution stability across pH and temperature. Automated metering frequently hinges on accurate dosing via a dosing pump.
Optimization is data‑driven. For each coagulant/polymer combination, track dose (g/t or ppm), overflow turbidity (NTU or mg/L TSS), and final cake solids; compare chemical cost versus dewatering benefit. A frequent outcome is a two‑stage scheme—for example, a cationic coagulant at 10–50 g/t solids followed by an anionic high‑MW polymer at roughly 20–50 g/t (mdpi.com; scielo.org.za). Dosing points matter: coagulant first into the slurry feed or thickener feedwell, polymer immediately before the thickener or dewatering device.
Monitoring continues online. Plants track underflow density and overflow clarity with NTU meters or solids probes, adjusting to seam changes, pH or hardness swings, and seasonal water temperatures that shift settling behavior. The commissioning record should capture results in plain data statements—e.g., “Dose = 30 g/t polymer yielded overflow = 5 NTU and cake = 75% solids”—and follow the stepwise path: lab/jar tests to shortlist reagents (scielo.org.za), bench/pilot to validate capture and clarity (tailings.info), then on‑site refinement with real‑time monitoring.
Why equipment and chemistry are inseparable
Under‑optimized chemistry leaves tailings dispersed and throttles throughput, clarity, or both. With the right polymer/coagulant pairing, high‑rate thickeners recover over 90% of water at moderate underflow solids, paste thickeners push to ~65–70%, belt filters deliver 75–85% solids cakes with filtrate often under 5 NTU, and solid‑bowl centrifuges reach ~65–76% solids while balancing recovery and centrate quality (mclanahan.com; westechwater.com; mdpi.com; researchgate.net). That’s the operating window where recycled water is clear enough for reuse and tailings are dense enough to stack.
