A sand reclamation unit is a process system that strips spent binder, fines, and metal contamination from used foundry sand so it can re-enter the moulding loop at near-virgin AFS 50-70 fineness, typically hitting 80-92% recovery on chemically bonded feeds and 95%+ on green sand.
For a 10-20 t/h foundry running two shifts, the equipment class spans three architectures — mechanical attrition, thermal fluidised-bed, and combined mechanical-thermal — with installed footprints from 80 m² to 350 m² and power loads between 200 kW and 1.2 MW. The decision is not whether the technology works; it does. The decision is whether the binder chemistry, the tonnage, and the disposal arbitrage in your plant justify the capex.
Thermal vs Mechanical vs Combined: What Each Architecture Actually Does
Mechanical reclamation (vibratory attrition + air classification) relies on scrubbing the binder film off the grain surface through grain-to-grain impact, typically reclaiming 70-85% of silica at AFS 50-70 when the feed is bentonite-bonded green sand with low resin content [S1]. The units run at 0.6-0.9 kg/kWh specific energy, generate 8-15% fines that bleed to dust collection, and cannot remove phenolic or furan resin films above ~2% LOI.
Combined mechanical-thermal units front-end an attrition stage to knock off clay and fines, then calcine the coarse fraction, typically landing the best silica roundness recovery and the highest installed cost per tonne of throughput. The interface point where the resin chemistry is the deciding variable is covered in sand reclamation unit reference detail.
Cost and Energy Stack: What the P&L Sees
Capex for a 10 t/h thermal line lands in the USD 1.8-3.5 M band before building and dust-collection integration; mechanical-only for the same throughput runs USD 0.5-1.1 M [S1]. A combined plant of 15 t/h sits in the USD 3-5 M range with a 6-9 month erection schedule.
Operating economics split by cost driver: electrical energy 25-35%, thermal energy 20-40% on thermal units, wear parts (tube/cyclone/liner replacement) 12-18%, and labour 8-12%. The critical interaction is between the upstream sand mixer binder dose and the downstream reclamation LOI tolerance — every 0.1% drop in incoming resin cuts thermal energy ~3-5%, which is why modern plants meter binder to a tight 0.9-1.1% resin-on-sand rather than the historical 1.4-1.8% belt.
Where the Unit Earns: Feeds That Justify the Capex
Foundries running furan/p phenolic no-bake at 5-15 t/day closed-loop, or shell-core sand with 3-6% phenolic resin, hit the strongest payback — typically 2-4 years on avoided virgin silica and avoided landfill tax [S1]. The unit pays off hardest in regions with active landfill levies above USD 30/t or virgin silica above USD 25/t.
Iron and steel jobbing shops running mixed bonded sand, ductile iron high-volume automotive lines, and any plant co-located with a sand cooler and degassing unit downstream — these are the textbook fits.
Where It Doesn't: Feeds, Plants, and Operating Regimes to Avoid
Reclamation does not work for sand contaminated with heavy metals above trace thresholds, for zircon or chromite blends (density separation behaviour breaks), or for any feed where the binder is inorganic and cannot be thermally oxidised. Plants below 3 t/day of throughput rarely see payback inside 5 years on thermal units — mechanical attrition only is the economic ceiling.
Operations with poor housekeeping upstream — spillage, core sand mixed into the floor returns, tramp metal upstream of the attrition stage — push the unit into a dust-handling problem rather than a sand-recovery problem. The integration that consistently fails is forcing a reclamation line into a foundry without a dedicated sand mixer and muller upstream, because incoming uniformity is what the reclaimer's control loop is designed against.
Selection Criteria: A Side-by-Side for the Decision Matrix
Three architectures, four decision axes: (1) Mechanical attrition — low capex (USD 0.5-1.1 M @ 10 t/h), 70-85% recovery, handles green sand only, energy 0.6-0.9 kWh/t. (2) Thermal fluidised-bed — high capex (USD 1.8-3.5 M @ 10 t/h), 90-96% recovery, handles phenolic/furan/pUR, gas 7-12 Nm³/t. (3) Combined mechanical-thermal — premium capex (USD 3-5 M @ 15 t/h), 92-96% recovery, handles all organic-bonded sand including shell-core, gas 5-8 Nm³/t after attrition pre-cleaning.
Decision rule: if feed LOI is above 1.5% or binder is phenolic, furan, or pUR, only the thermal or combined path is technically viable. If feed is bentonite-only green sand, mechanical attrition returns the best IRR. The interface between reclamation and the upstream hydraulic power unit / FRL unit on the binder-delivery side is the second-most-overlooked integration point after dust collection.
Standards, Sourcing, and What the Procurement Audit Catches
European and US foundry audit protocols grade reclaimed sand on (a) AFS fineness 40-70, (b) LOI below 0.5% for chemically bonded reuse, (c) acid demand value below 6 mL, and (d) clay content below 0.5% for bentonite systems [S1]. Dust collectors on thermal units route to ATEX 2014/34/EU-certified filtration because the exhaust stream carries combustible fines at 150-220 °C.
Supplier short-listing should check three things: (1) reference list with at least three plants running the same binder chemistry for 5+ years, (2) wear-parts lead time under 4 weeks for cyclones and fluidising tubes, and (3) burner package compliance with local NOx rules — EU plants must hit below 100 mg/Nm³ at 3% O₂, which is the gating spec on any thermal unit installed since 2020. For context on the adjacent bag-filter access and dust-collection constraints that always ride alongside these units, the bag filter access spec map covers the maintenance-clearance side of the same problem.
Trackable signal: binder chemistry is moving — alkali-phenolic and inorganic geopolymer systems are now specified on 12-18% of new European foundry builds, and neither is reclaimable on a thermal unit, so any capex decision should include a 5-year binder-forecast clause. Watch for reclamation-vendor service contracts pivoting from per-tonne processing to LOI-guaranteed reuse contracts through 2027.