Design prototype - draft copy, not for publication
Recyclers Guide Contact Build an RFQ

Technology reference: eddy current separation

Recovering fine non-ferrous metal

Fine non-ferrous metal here means aluminum, copper and other non-ferrous pieces under about 10 mm. Conventional eddy current separators lose effectiveness below about 5 mm, so plants that chase fines use dedicated fines separators, or other routes for metal too fine or too heavy for them.

By [author name, role and employer] · Technical review by [reviewer name, role and employer] · Updated

  • Aluminum and non-ferrous
  • Ferrous
  • Fiber
  • Glass
  • Film
  • Other and residue

1Why small pieces are hard to separate

An eddy current separator throws a conductive piece by inducing currents in it. For small pieces, that push shrinks much faster than the piece's weight. In the small-piece range the force grows with about the fifth power of size, while the mass grows with the cube. Below about 5 mm, friction and other forces take over on a conventional separator. Source: Settimo, Bevilacqua and Rem, 2004.

The same paper says fine non-ferrous metal is hard to recover: earlier processes had either low capacity or incomplete recovery below 5 mm. A 2017 paper notes that rotating drums struggle with pieces of about 1 cm and smaller. Sources: Settimo, Bevilacqua and Rem, 2004; Dholu et al., 2017.

Size changes how a piece moves, not only how hard it is pushed. A 2022 study found that, for any one set of machine settings, there is an optimum piece size. As pieces get smaller, their forward motion first rises, then drops fast, while their spin keeps rising. Source: Cao et al., 2022.

Air adds a second problem. Bunge reports that air drag on metal pieces can be ignored above about 4 mm but dominates below about 2 mm. At 2 mm a copper piece flies about as far as an aluminum piece of the same size, and at 1 mm copper can fly farther. Spin effects matter for pieces under about 8 mm. Source: Bunge, Recovery of metals from waste incinerator bottom ash, 2019.

Where air drag can be ignored, above about 4 mm, aluminum is thrown farther than copper because it conducts well for its weight. Sources: Bunge, 2019 (drag); Goudsmit, What determines the separation efficiency of eddy current systems (aluminum and copper). The eddy current reference explains why. At the finest sizes, throw no longer sorts the two the same way. That is this guide's reading of Bunge's figures.

Makers answer with a different field. Bunting says larger pieces need a deeper field with fewer poles, and smaller pieces a shallower field with more poles turning faster. Source: Bunting, Maximising non-ferrous metal separation, 2019.

A 2020 review of bottom-ash plants gives the numbers. Standard rotors at 2,000 to 3,000 rpm suit only pieces over about 5 mm. High recovery below 5 mm needs more than about 4,000 rpm, more poles, or both. Eccentric and high-frequency fines machines are now used in some full-scale ash plants. Source: Šyc et al., 2020.

2Deciding whether fines are worth chasing

Two questions decide it: how much metal sits in your fines, and how much separator capacity it takes to get it out. Published answers come mostly from incinerator bottom ash.

How much metal the fines hold

In three bottom-ash plants, most of the non-ferrous metal recovered on the separators was 2 to 30 mm. Bunge puts about half of the non-ferrous value worth recovering in pieces under 8 mm. Source: Bunge, 2019.

Which metal sits in the fines varies. In Amsterdam ash, aluminum dominated the 6 to 20 mm pieces, while copper made up most of the non-ferrous metal under 2 mm. In Czech plants, non-ferrous metal was spread evenly across the size fractions. Source: Šyc et al., 2020.

Some fine aluminum is out of reach. In two Italian plants, about 61 to 67% of the aluminum fed to the furnace ended up in the fine fraction of the bottom ash, where it could not be recovered with the technology then in use. Of the aluminum that could be recovered as metal, 76 to 87% came from pieces over 5 mm. Source: Biganzoli, 2013.

Thin packaging fares worst. In the same study, about 81% of the aluminum in cans was recoverable, against 27 to 47% for foil and 51% for trays. Source: Biganzoli, 2013.

A 2024 study found that fine metallic aluminum left in weathered bottom ash cannot be removed with eddy current separators, because the pieces are too small. Source: Chen et al., 2024.

What fines cost in capacity

A separator's belt must carry a single layer, which limits throughput (Bunge, 2019). Goudsmit sets that layer at about the size of the largest piece (Goudsmit, EddyFines). So a belt fed with small pieces carries less material per metre of width. That last step is this guide's reasoning, not a measured result.

One maker's published maximums show the scale. Bunting rates aluminum dross of 20 to 200 mm at up to 20 t/h per metre of belt width on a concentric model, and dross of 0 to 20 mm at up to 5 t/h per metre on its eccentric fines models. For granulated plastic of 3 to 10 mm on the same fines models, the figure is 2. These are the maker's maximums, not guarantees. Source: Bunting, Eddy current separators, undated.

An older research result points the same way. Rotors with many small poles recovered aluminum down to about 1 mm, but at about 1 t/h per metre of rotor width (a 1999 result, as reported in 2004). Source: Settimo, Bevilacqua and Rem, 2004.

What the buyer accepts

Check the product's grade as well. ReMA, the Recycled Materials Association, publishes the grade for shredded used beverage cans, Talcred, which allows at most 5% fines under 4 mesh (6.35 mm, as ReMA prints it). Source: ReMA, ISRI Specifications, July 2026. Aluminum scrap grades and what buyers accept covers the grades in full.

Where fine metal turns up

Fine non-ferrous metal by stream (what the sources state; most published figures are from bottom ash)
StreamWhat the sources say about its finesMore on this site
Incinerator bottom ashAbout 1 to 5% of the ash is non-ferrous metal, mostly aluminum. STEINERT says that, depending on how the ash is removed from the incinerator, up to 50% can be under 10 mm and up to 30% under 4 mm.Section 3
Shredder residueIn Argonne's vehicles-only shredding test, fines under 12 mm were about 2.5% of the shredder feed; residue processed in its pilot plant averaged 32% fines, partly because normal feed carries dirt and rock. Eriez sells a fines unit for non-ferrous metal and bare copper wire from shredder residue, down to 2 to 3 mm.Shredder residue and Zorba
Glass culletThe 12- and 16-pole separators of Industrial Magnetics (IMI) were first designed for glass cleanup, and IMI designs its 16-pole units for pieces under 2 in. ReMA's MRF glass grade allows up to 30% fines under 1/8 in, too small to sort optically.Non-ferrous metal out of glass cullet
Construction and demolition wasteIn one maker's case, a plant handling construction, demolition and commercial skip waste lets its 0 to 10 mm fines leave as a product without eddy current treatment; the separator works on the 10 to 100 mm heavy fraction.None yet
MRF container and glass lines[share of fine aluminum in MRF fines and glass-line material: no published figure found; a measured sample would fill it]MRF container lines

Sources: Šyc et al., 2020 (ash composition); STEINERT, Incineration bottom ash, undated (ash under 10 and 4 mm); Duranceau and Spangenberger, Argonne National Laboratory, 2011 (shredder fines); Eriez, Eddy current separators product page, undated, and Eriez, RB-780, 2026 (fines unit); IMI, TG-13A data sheet, 2022 to 2023 (glass cleanup); ReMA, ISRI Specifications, July 2026, MRF glass (glass fines); ReMA, MRF glass grade page, 2022 (the same values, and that fines this small cannot be sorted optically); Terex Recycling Systems, First plant install, undated (construction and demolition). All read 2026-10-06.

3What bottom-ash plants recover, by size

Bottom ash has the longest public record of fines recovery. The studies below use different plants, years and bases, so read each on its own terms rather than against the others.

Non-ferrous recovery from bottom ash by piece size, as published (each on its own basis; not a direct comparison)
PublishedFindingSource type and basis
2011: Berkhout, Oudenhoven and RemIn conventional dry processing, recovery falls from almost 100% above 20 mm to nearly zero somewhere between 5 and 12 mm, depending on the screening steps and separators. Wet and advanced dry processes reach about 2 mm.Peer-reviewed paper; conventional dry processing
2013: BiganzoliTwo Italian plants: 61 to 67% of the aluminum fed to the furnace ended in the fine fraction and could not be recovered with the technology then in use (2013). Of the recoverable aluminum, 76 to 87% came from pieces over 5 mm.PhD thesis; share of the aluminum fed to the furnace
2019: BungeTraditional separators work well above about 6 mm, and recovering 0.5 mm pieces was the state of the art. In three plants, most of the recovered non-ferrous metal was 2 to 30 mm.Technical monograph; expert review of plants
2019: Bunge, design sketchA sketched dry plant is expected to recover about 65% of the non-ferrous metal over 2 mm, about 75% with a closed crushing loop and about 85% with all options.Expert estimate, not a measured plant
2020: Šyc et al.Simple trains that split ash into two fractions recover 9 to 48% of the non-ferrous metal, and 86% of the fraction over 20 mm.Peer-reviewed review; metal fed to the furnace
2020: Šyc et al., citing one Copenhagen plantAbout 90% at 4 to 50 mm and 60 to 75% at 1 to 4 mm. Pieces under 0.5 mm are not treated.Peer-reviewed review; metal actually sold, not metal sorted out
2026: Mika, Mühl and LedererDry-discharged fluidized-bed ash gave 97% aluminum extraction into metal concentrates, against 78% for grate ash. The authors credit dry discharge and a lower bed temperature.Peer-reviewed paper (abstract read); aluminum extracted into concentrates

Sources: Berkhout, Oudenhoven and Rem, 2011; Biganzoli, 2013; Bunge, 2019 (sections 4.4.2 and 5.1); Šyc et al., 2020, the Copenhagen figures citing Kallesøe and Dyhr-Jensen, 2018; Mika, Mühl and Lederer, 2026. All read 2026-10-06.

One plant that treats each fraction

A Copenhagen plant handling 200,000 to 250,000 t of ash a year ages the ash for two to three months, removes ferrous metal, and screens the rest into seven fractions. Six fractions, from 0.5 to 50 mm, each get an eddy current separator, and the two coarsest also get induction sorters for stainless steel. Source: Šyc et al., 2020, citing Kallesøe and Dyhr-Jensen, 2018.

The same review says advanced plants screen into 6 to 9 fractions and tune a separator to each. Source: Šyc et al., 2020. Screening ahead of separation covers how plants split the feed.

Per ton of raw ash, European plants recover on average about 17 kg of non-ferrous metal. Advanced trains reach about 30 kg, and one train with 12 eddy current separators recovers over 20 kg. Source: Šyc et al., 2020.

Rules push recovery down in size in some countries. Switzerland requires non-ferrous metal to be recovered from ash pieces over 2 mm, so that landfilled residue holds under 1.5% non-ferrous metal, and the Netherlands has set a similar approach. Source: Šyc et al., 2020.

These are European results, and US practice differs: at US plants, fly ash is wetted and mixed with bottom ash before landfilling. Do not read the European rates as typical of US plants. Source: US EPA, Energy recovery from the combustion of municipal solid waste.

4Fines eddy current separators

Makers sell separate machines for fines. Bunge describes how they differ from standard separators: they spin faster and carry more poles, set closer together. Source: Bunge, 2019.

Several makers put the fines line near 10 mm. STEINERT builds its fines machine for 0.5 to 10 mm, Goudsmit gives its EddyFines a separation range of 0 to 10+ mm, and Bunting ties its eccentric rotors to pieces under 10 mm. Sources: STEINERT, EddyC FINES; Goudsmit, EddyFines 1500 mm; Bunting, Eddy current separators.

What makers state about their fines machines (each claim is the named maker's own; not a ranking)
MachineRotor and belt, as statedFeed, as stated
STEINERT EddyC FINESEccentric pole system, up to 4,000 rpm, reversing the field up to 1.3 kHz; the point where the field acts can be varied. Working widths 1,000, 1,500 and 2,000 mm.Pieces of 0.5 to 10 mm, with the largest no more than three times the smallest. For wet fine incinerator ash, an alternative splitter that reduces caking, and a choice of belt scrapers.
Goudsmit EddyFines38-pole eccentric rotor at 3,000 rpm. A 2.4 mm polyurethane belt runs on stainless steel sliding plates instead of rollers, to limit bouncing.One layer, about as deep as the largest piece. Its vibratory feeder module does not suit incinerator slag, which sticks; its conveyor module has a belt scraper and needs a spreader belt.
Eriez UHF32-pole eccentric rotor, for fractions under 3/8 in.Ships with an Eriez Brute Force feeder, which spreads material across the belt.
Bunting HIFE30-pole eccentric rotor, 500 mm in diameter.Bunting sizes a separator on the size range, the metal and its shape, feed moisture, recovery and purity targets, and burden depth. Its options include belt brushes for sticky or wet feed.
IFE NANOSEccentric rotor, 400 mm, up to 4,000 rpm. IFE says it can be fitted to an IFE centric machine by swapping the rotor.IFE's trials used 1 to 8 mm and 4 to 8 mm feed, and its first customer unit runs on 2 to 5 mm.

Sources: STEINERT, EddyC FINES, undated; Goudsmit, EddyFines 1500 mm, undated, and EddyFines eddy current separators, undated; Eriez, Eddy current separators product page, undated, and RB-780 brochure, 2026; Bunting, Eddy current separators, undated, and 5 key eddy current separator operating parameters, 2020; IFE, Eddy current separators brochure, 2026, and The new NANOS eddy current separator, 2024. All read 2026-10-06. The smallest piece each maker states, model by model, is in the size table on the eddy current reference.

The finest limits makers publish are their own claims, for their own test material. STEINERT's range starts at 0.5 mm, and Goudsmit says EddyFines ejects pieces as small as about 0.5 mm. Eriez states 2 to 3 mm for its UHF unit, and SGM lists one model, BVIS, down to 0.2 mm. Sources: STEINERT, EddyC FINES; Goudsmit, EddyFines; Eriez, RB-780, 2026; SGM, Eddy current separators.

Rotor layout alone may not set the limit. SGM says it builds concentric rotors, and it lists a concentric model, the BVIS, down to 0.2 mm. The other fines models in the size table are eccentric. Sources: SGM, Eddy current separators; SGM, BVIS model.

SGM also lists a high-frequency model, VIS, at up to 4,800 rpm, for e-scrap, shredder residue and fine aluminum. Source: SGM, Eddy current separators.

IFE lists a design with a static eddy current bar instead of a rotor, for pieces from 0.5 mm, with the belt at 2 to 4 m/s. Source: IFE, Eddy current separators brochure, 2026.

Which rotor layout for fines

Makers disagree. Bunting ties eccentric rotors to pieces under 10 mm. IFE suggests eccentric rotors below roughly 20 mm, where a precise release point matters. Dings markets its eccentric model for fractional-size non-ferrous metal in recycling. Sources: Bunting, Eddy current separators; IFE, NANOS, 2024; Dings, Eddy current separator.

SGM argues the other way. It says its concentric rotors keep pieces in the field longer, which gives a gradual separation suited to fine and ultra-fine material. Source: SGM, Eddy current separators. Concentric and eccentric rotors sets each maker's claims side by side.

Remcon Equipment, the publisher's affiliate, supplies both layouts and describes the eccentric rotor as the usual choice for fine aluminum; it also says its separators serve glass lines. That is the publisher's affiliate's practice, not a finding of this guide: how the publisher's affiliate builds one (source 34). [which rotor diameter and layout the affiliate fits for foil and for aluminum in glass cullet: to come from the shop]

5Preparing the feed for a fines separator

Feed preparation decides as much as the machine. These are the steps the sources name, in the order a plant meets them.

Free the metal

Small metal is often locked in mineral clumps. Shredding or crushing frees it and raises recovery. Source: Šyc et al., 2020. Bunge describes ash processing in three stages: free the metal, classify it by size, then separate it. Source: Bunge, 2019.

Take out steel and magnetic oxides

Separator makers' manuals ask for magnets ahead of the separator. IMI advises at least one, and notes that several lower the risk, because no magnet removes all ferrous metal. Eriez designs its eccentric separator to run with a minimum of ferrous metal in the feed. Sources: Industrial Magnetics, manual 900505; Eriez, RevX-E manual RM-99, 2015.

Ash plants add a step. A 2020 review describes a drum magnet after the overhead magnet, to remove iron oxides and agglomerates, because that magnetic fraction lowers separation. Source: Šyc et al., 2020.

Bunge notes a trade-off. Weakly magnetic particles can spoil fine non-ferrous separation, which argues for a strong magnet ahead of a fines separator. But nearly half of bottom-ash feed can be weakly magnetic, and stripping it all loses the copper and precious metal attached to it. Source: Bunge, 2019.

This guide follows the separator makers' manuals: magnets go ahead of the separator, and the separator is not relied on to remove steel. Removing ferrous metal before the separator covers the magnets.

Screen to a narrow size range

STEINERT asks that the largest piece fed to its fines machine be no more than three times the smallest. Source: STEINERT, EddyC FINES. A 2 to 6 mm cut meets that rule; a 1 to 10 mm cut does not. That is this guide's arithmetic.

The splitter is one reason. Bunge reports that coarse pieces separate better with the splitter set low and small pieces with it set high, and that the right setting is found by test. Source: Bunge, 2019. So one setting cannot suit both ends of a wide range. That last step is this guide's reading.

Screening fines is hard in itself. Bunge puts the lowest effective split of a star screen at about 8 mm, and of a flip-flow screen, used on wet, sticky ash, at about 3 mm. He calls screening ash below 8 mm hard. Source: Bunge, 2019. Screening ahead of separation covers the screens.

Spread it one layer deep

Bunge says the separator belt must never carry more than a single layer, and that modern plants spread material once and keep it on wide flat belts through the whole separator train. Goudsmit says a single layer matters most with smaller fractions. Sources: Bunge, 2019; Goudsmit, EddyFines. The belt feed-rate calculator shows how belt speed, width and bulk density set the depth.

Keep it dry

Moisture makes material stick to itself and to the belt. Source: Goudsmit, What determines the separation efficiency of eddy current systems.

Bottom ash shows the problem at its worst. Wet-discharged ash holds about 18 to 25% moisture, mostly in the fines, and compounds formed in quenching harden and coat the metal. The 2020 review calls these sticky fines the central problem for dry metal recovery. Source: Šyc et al., 2020.

In plant and lab tests on 1 to 6 mm ash, very wet feed lowered recovery. Source: Rahman and Bakker, 2013.

Dry handling helps. In dry-handled ash, metal is not clumped by sticky fines and is easier for the separator to reach. One plant in the same review strips light, sticky fines under 10 mm ahead of its separators, using a process the review calls Advanced Dry Recovery (ADR), and runs its fine-fraction separator at 4,000 rpm. Source: Šyc et al., 2020. The makers' own answers to wet feed are in the table in section 4.

Moisture limit: [moisture level at which a dry fines separator starts to lose recovery: no published figure found; a maker's test on your material would fill it]

Set the splitter by test

In the same 1 to 6 mm ash tests, moving the splitter by 10 mm could change metal recovery by about 10% and grade by about 18%. Source: Rahman and Bakker, 2013. Setting the splitter covers how to set it.

6Other routes for fine metal

A fines separator is one route. The sources name others, each for a different metal or size. The table sets them side by side on what the sources state.

Routes for fine non-ferrous metal (stated criteria, not a ranking)
RouteTargetsSizes the source statesWhere it stands
Fines eddy current separatorAluminum and other conductive non-ferrous metalMakers' ranges reach 0.5 mm, and one model 0.2 mm (section 4)Used in some full-scale ash plants
Electrostatic separationConductive metalAbout 4 mm down to 45 µmNamed by an eddy current separator maker as the route for metal too fine for its separators
Density separationFine heavy non-ferrous metal, such as copper and brassNot stated in the sources readUsed on ash lines; cannot recover aluminum from ash
Induction sensor sortingAll metals, including stainless steelOver about 4 mmUsed on ash, on a thin layer; recovery and grade can exceed 90%
Wet eddy current separationNon-ferrous metal in slightly damp feedTested on 2 to 6 mmResearch; many such designs were never built at full scale
Leaving the finest fractionNothing: the fraction leaves untreatedUnder 0.5 mm at one ash plant; 0 to 10 mm at one plant handling construction, demolition and commercial skip wasteWhat those plants do

Sources: Šyc et al., 2020 (fines separators, density separation, induction sorting, full-scale status, the ash plant); Bunting, Eddy current versus electrostatic separation, 2020 (electrostatic); Settimo, Bevilacqua and Rem, 2004 (wet eddy current); Terex Recycling Systems, First plant install, undated (construction and demolition). All read 2026-10-06.

Who publishes this. Recyclers Guide is published by [LLC name], an affiliate of Remcon Equipment, Inc. Remcon designs and builds conveyors and eddy current separator systems around rotors or complete separators made by others, builds overhead magnet units, magnetic head pulleys and trommels, resells screens and feeders, and integrates equipment discussed on this site.

Electrostatic separation

Bunting, which makes eddy current separators, says they suit non-ferrous pieces above about 3 mm, and that electrostatic separators handle conductive metal from about 4 mm down to 45 µm. Its own model table lists one eccentric model from 1 mm, so its two pages draw the line in different places. Sources: Bunting, Eddy current versus electrostatic separation, 2020; Bunting, Eddy current separators, undated.

[electrostatic separators for recycling fines: makers, feed conditions and capacity; not yet researched for this guide]

Density separation

In bottom ash, density separation recovers fine heavy non-ferrous metal such as copper and brass. It cannot recover aluminum from ash, because aluminum's density is close to that of the ash. Source: Šyc et al., 2020.

Induction sensor sorting

Induction sorters detect all metals, stainless steel included. The 2020 review gives them pieces over about 4 mm in ash, spread in a thin layer, with recovery and grade that can exceed 90%. Source: Šyc et al., 2020. Sensor-based sorting for metals covers the sensors.

Research designs

Researchers have tried other ways to reach fines. In a wet-feed process, slightly damp feed of 10 to 15% moisture sticks to the belt, and the rotor frees only the metal. Tested on 2 to 6 mm waste at 4 t/h per metre, it recovered all the copper, zinc and lead and about 79% of the aluminum. On e-scrap it worked above 1 mm but not below. Source: Settimo, Bevilacqua and Rem, 2004.

A Magnus separator took up to about 6 t/h per metre without single-layer feeding, but recovered only about 70 to 80% of the aluminum, and less of the heavy metals. Source: Settimo, Bevilacqua and Rem, 2004.

A 2020 review lists wet, Magnus and backward-running separators among the designs proposed in the literature, and says many were never built at full scale. Source: Šyc et al., 2020.

7What to tell the manufacturer

Ask each maker the same questions, in the same order. Each item below changes the result, according to the sources on this page:

  • the size range after screening: the largest and the smallest piece
  • which metals are in the fines, and their shapes: angular, wire, round or foil
  • how wet the feed is, and whether it sticks
  • how steel and magnetic oxides are removed upstream, and how many magnets do it
  • the feed rate of the fine fraction on its own, and the layer depth you can accept
  • what your buyer accepts, including any limit on fines in the product
  • whether the maker will test a sample of your fines, and the feed size range behind any figure it quotes

Sources: Bunting, 5 key eddy current separator operating parameters, 2020 (sizing inputs); Bunting, Maximising non-ferrous metal separation, 2019 (testing); STEINERT, EddyC FINES (size ratio); Šyc et al., 2020 (magnetic oxides); ReMA, July 2026 (fines limit).

The RFQ checklist and Sizing an eddy current separator cover the full list. Eddy current makers are listed in the directory.

8Questions

Why do eddy current separators struggle with small pieces?

The push on a small piece shrinks much faster than its weight. The force grows with about the fifth power of size and the mass with the cube, so below about 5 mm friction takes over on a conventional separator. Below about 2 mm, air drag also dominates: a 2 mm copper piece flies about as far as a 2 mm aluminum piece. Sources: Settimo, Bevilacqua and Rem, 2004; Bunge, 2019.

Is fine non-ferrous metal worth recovering?

It depends on how much metal your fines hold and how much capacity you give up. In bottom ash, Bunge puts about half of the non-ferrous value worth recovering in pieces under 8 mm. But in one maker's ratings, fines models carry less per metre of belt width: Bunting rates 0 to 20 mm aluminum dross at up to 5 t/h per metre on its fines models, against up to 20 for 20 to 200 mm dross on a coarser model. Bunting also advises testing your own material. Sources: Bunge, 2019; Bunting, Eddy current separators, undated; Bunting, Maximising non-ferrous metal separation, 2019.

What can recover metal that is too fine for an eddy current separator?

Bunting, an eddy current separator maker, points to electrostatic separation for conductive metal from about 4 mm down to 45 µm. In bottom ash, density separation recovers fine heavy metals such as copper and brass, but not aluminum, whose density is close to that of the ash. Sources: Bunting, Eddy current versus electrostatic separation, 2020; Šyc et al., 2020.

How narrow a size range does a fines separator need?

STEINERT asks that the largest piece fed to its fines separator be no more than three times the smallest. By that rule, a 2 to 6 mm cut fits and a 1 to 10 mm cut does not. Bunge reports that coarse pieces separate better with the splitter set low, and small pieces with it set high. Sources: STEINERT, EddyC FINES, undated; Bunge, 2019.

Does a fines separator still need a magnet ahead of it?

Yes. Separator makers' manuals call for magnets upstream, and a 2020 review of ash plants describes a drum magnet after the overhead magnet to remove iron oxides that lower separation. This guide follows the separator makers' manuals: magnets go ahead of the separator, and the separator is not relied on to remove steel. Sources: Industrial Magnetics, manual 900505, undated; Eriez, RevX-E manual RM-99, 2015; Šyc et al., 2020.

Read the eddy current reference Find manufacturers

9Sources

Every figure on this page comes from one of these sources. Manufacturers' pages and manuals state each maker's own claims; this site has not tested the equipment.

  1. Berkhout, S.P.M., Oudenhoven, B.P.M., and Rem, P.C. Optimizing non-ferrous metal value from MSWI bottom ashes. Journal of Environmental Protection 2(5):564-570, July 2011 (peer-reviewed). https://content.scirp.org/pdf/jep20110500017_42862674.pdf. Read 2026-10-06.
  2. Biganzoli, L. PhD thesis, Politecnico di Milano, March 2013 (findings also published in Waste Management and Research, 2013). https://www.politesi.polimi.it/retrieve/a81cb05a-5011-616b-e053-1605fe0a889a/2013_03_PhD_Biganzoli.pdf. Read 2026-10-06.
  3. Bunge, R., UMTEC, OST Eastern Switzerland University of Applied Sciences. Recovery of metals from waste incinerator bottom ash (technical monograph). June 2019. https://www.ost.ch/fileadmin/dateiliste/3_forschung_dienstleistung/institute/umtec/fachartikel/2019/19-03_recovery_of_metals_from_waste_incinerator_bottom_ash.pdf. Read 2026-10-06.
  4. Bunting. 5 key eddy current separator operating parameters (blog). November 10, 2020. https://buntingmagnetics.com/blog/5-key-eddy-current-separator-operating-parameters. Read 2026-10-06.
  5. Bunting. Eddy current separators (product page, model table, capacity table and options). Undated. https://buntingmagnetics.com/product/magnetic-separation/eddy-current-separators. Read 2026-10-06.
  6. Bunting. Eddy current versus electrostatic separation (blog). July 20, 2020. https://buntingmagnetics.com/blog/eddy-current-versus-electrostatic-separation. Read 2026-10-06.
  7. Bunting. Maximising non-ferrous metal separation (blog). December 3, 2019. https://buntingmagnetics.com/blog/maximising-non-ferrous-metal-separation. Read 2026-10-06.
  8. Cao, B., et al. Effects of particle size on the separation efficiency in a rotary-drum eddy current separator. Powder Technology 410:117870, 2022 (accepted manuscript). https://eprints.bournemouth.ac.uk/37428/. Read 2026-10-06.
  9. Chen, B., et al. Characterization and mechanical removal of metallic Al embedded in weathered MSWI bottom ash. Waste Management 176:128-139, March 2024. https://pubmed.ncbi.nlm.nih.gov/38281344. Read 2026-10-06.
  10. Dholu, N., Nagel, J.R., Cohrs, D., and Rajamani, R.K. Eddy current separation of nonferrous metals using a variable-frequency electromagnet. KONA Powder and Particle Journal 34:241-247, 2017. https://www.jstage.jst.go.jp/article/kona/34/0/34_2017012/_html/-char/en. Read 2026-10-06.
  11. Dings Magnetic Group. Eddy current separator (product page). Undated. https://dingsmagnets.com/products/eddy-current-separator/. Read 2026-10-06. Dings has supplied magnets to Remcon Equipment, the publisher's affiliate.
  12. Duranceau, C., and Spangenberger, J., Argonne National Laboratory. All Auto Shredding: evaluation of automotive shredder residue generated by shredding only vehicles, ANL/ES-C0201801. July 11, 2011. https://publications.anl.gov/anlpubs/2011/09/70931.pdf. Read 2026-10-06.
  13. Eriez. Eddy current separators (product page). Undated. https://www.eriez.com/Products/Metals-Recycling/Nonferrous-Recovery/Eddy-Current-Separators. Read 2026-10-06.
  14. Eriez. Eddy current separators for nonferrous recovery, brochure RB-780. September 2026. https://www.eriez.com/Documents/Literature/Brochures/Products/Magnetic-Separation/RB-780-Eriez-Eddy-Current-Separator.pdf. Read 2026-10-06.
  15. Eriez. RevX-E eddy current separator, installation, operation and maintenance manual RM-99. April 2015. https://www.eriez.com/Documents/Literature/IOMs/Metals-Recycling/RM-99-Eriez-RevXE-Eddy-Current-Separator-IOM.pdf. Read 2026-10-06.
  16. Goudsmit Magnetics. What determines the separation efficiency of eddy current systems (knowledge base). Undated. https://www.goudsmitmagnetics.com/en-us/knowledge-base-efficiency-of-eddy-current-separators. Read 2026-10-06.
  17. Goudsmit Magnetics. EddyFines eddy current separator 1500 mm (product page). Undated. https://www.goudsmitmagnetics.com/en-us/products/eddyfines-eddy-current-separator-1500-mm/e0136908. Read 2026-10-06.
  18. Goudsmit Magnetics. EddyFines eddy current separators (product page). Undated. https://www.goudsmitmagnetics.com/en-us/products/belt-conveyor-magnetic-separators/eddyfines-eddy-current-separators. Read 2026-10-06.
  19. IFE Aufbereitungstechnik. Eddy current separators (brochure). March 2026. https://www.ife-bulk.com/fileadmin/user_upload/ife-bulk/Produkte/Wirbelstromscheider/IFE-eddy-current-separators_web.pdf. Read 2026-10-06.
  20. IFE Aufbereitungstechnik. The new NANOS eddy current separator (news). 2024. https://www.ife-bulk.com/en/company/news/the-new-nanos-eddy-current-separator. Read 2026-10-06.
  21. Industrial Magnetics, Inc. (IMI). Eddy current separator service manual, endless belt, 900505. Undated. https://assets.magnetics.com/assets/files/manuals/900505_ECS_Service_Manual_Endless_Belt.pdf. Read 2026-10-06. Industrial Magnetics supplies eddy current separators and magnets to Remcon Equipment, the publisher's affiliate.
  22. Industrial Magnetics, Inc. (IMI). Eddy current separators, data sheet TG-13A. 2022 to 2023. https://assets.magnetics.com/assets/files/technical-data-sheets/TG13A_EddyCurrentSep.pdf. Read 2026-10-06. Industrial Magnetics supplies eddy current separators and magnets to Remcon Equipment, the publisher's affiliate.
  23. Mika, S., Mühl, J., and Lederer, J. Extraction of metallic aluminum from fluidized bed and grate incineration bottom ash. Waste Management and Research 44(9):1460-1471, September 2026 (abstract). https://pubmed.ncbi.nlm.nih.gov/42068213/. Read 2026-10-06.
  24. Rahman, M.A., and Bakker, M.C. Sensor-based control in eddy current separation of incinerator bottom ash. Waste Management 33(6):1418-1424, June 2013 (abstract). https://pubmed.ncbi.nlm.nih.gov/23490354/. Read 2026-10-06.
  25. ReMA (Recycled Materials Association). ReMA's ISRI Specifications. July 2026. https://www.isrispecs.org/wp-content/uploads/2026/09/2026-July-ReMA_ISRI_Specifications.pdf. Read 2026-10-06.
  26. ReMA (Recycled Materials Association). Material recovery facility-derived 3-color mixed container glass (MRF glass) (grade page). Effective July 14, 2022. https://www.isrispecs.org/orpheus_resource/material-recovery-facility-derived-3-color-mixed-container-glass-mrf-glass/. Read 2026-10-06.
  27. Settimo, F., Bevilacqua, P., and Rem, P. Eddy current separation of fine non-ferrous particles from bulk streams. Physical Separation in Science and Engineering 13(1):15-23, 2004. https://repository.tudelft.nl/record/uuid:a219d415-14cf-47ee-84de-719786333cdb. Read 2026-10-06.
  28. SGM Magnetics. Eddy current separators (product page). Undated. https://www.sgmmagnetics.com/en/products/eddy-current-separators/. Read 2026-10-06.
  29. STEINERT. EddyC FINES (product page). Undated. https://steinertglobal.com/us/sorting-systems/magnetic-separation/eddy-current-separators/steinert-eddyc-fines/. Read 2026-10-06.
  30. STEINERT. Incineration bottom ash (application page). Undated. https://steinertglobal.com/us/applications/metal-recycling/incineration-bottom-ash/. Read 2026-10-06.
  31. Šyc, M., Simon, F.G., Hykš, J., Braga, R., Biganzoli, L., Costa, G., et al. Metal recovery from incineration bottom ash: state-of-the-art and recent developments. Journal of Hazardous Materials 393:122433, 2020 (postprint). https://cris.unibo.it/bitstream/11585/762085/5/S%cc%8cyc2020_postprint.pdf; DOI https://doi.org/10.1016/j.jhazmat.2020.122433. Read 2026-10-06.
  32. Terex Recycling Systems. First plant install (case page). Undated. https://www.terex.com/recycling/first-plant-install. Read 2026-10-06.
  33. US Environmental Protection Agency. Energy recovery from the combustion of municipal solid waste (MSW). Page last updated March 19, 2026. https://www.epa.gov/smm/energy-recovery-combustion-municipal-solid-waste-msw. Read 2026-10-06.
  34. Remcon Equipment, Inc., the publisher's affiliate. Eddy current aluminum separators (product page). Undated. https://www.remconequipment.com/eddy-current-aluminum-separators. Read 2026-10-06.
  35. SGM Magnetics. BVIS model (product page). Undated. https://www.sgmmagnetics.com/en/products/bvis-model/. Read 2026-10-06.

Updated . Next review due [date].

Found an error on this page? Write to [corrections address].