Recycling Conveyor System: System Design and Buyer Checklist

Designing a recycling conveyor system begins with the material, not with the belt catalogue. Mixed municipal and commercial recovery streams arrive wet, abrasive, tangled and inconsistent, and every one of those properties pushes the layout toward hardware that a quarry or a cement plant would never pick. Across the recovery sites our engineers have walked, the stoppages that cost the most production were hardly ever tensile failures of the belt itself. Wrap points, transfer drops and blocked cleaning access did the damage.

This article gives the system-level view we work from at pre-engineering stage: where a recovery line begins and ends, how the physical properties of the waste stream constrain the machinery, how to test a tonnage target against belt width and belt speed, and what belongs on a purchase order before the first frame is welded. Belt compound choice gets a short mention only, because that subject already carries its own note.

Treat every tonnage figure below as a worked example of the arithmetic. Re-run it with your own bulk density, your own peak factor and your own shift pattern.

01System Boundary: Where a Recovery Line Begins and Ends

A recovery line runs from the tipping floor to the baler discharge, and everything between those two points has to be drawn as a single system. Enquiries often get split into a tipping package and a sorting package, and the two halves then show up sized for different peak flows. We keep one boundary, one flow sheet, and an explicit note of every place where material changes hands between machines.

One boundary, one flow sheet.

Where the boundary gets drawn too tight, the receiving hopper and the final transfer into the baler have usually been left off because a different purchase order covered them. A hopper that meters badly starves the whole line. Both pieces belong inside the system, even when another supplier delivers them.

Why the receiving hopper and pre-sort set the tone downstream

Receiving is where variability enters the plant in its rawest form, at whatever density the collection round happened to produce. The first conveyor therefore has to swallow surges that can reach twice the daily average inside a single hour. We size the hopper for the largest vehicle in the fleet plus a working margin, keep discharge plates at 45 degrees or steeper so wet paper cannot bridge across the opening, and line the impact zone with 10 mm plate, because a tipper releasing a mixed load from 1.2 m concentrates a surprising amount of energy into a small footprint. Too many projects pour their budget into optical sorting first and then rebuild the receiving end in year two.

What each stage must deliver before the next one can function

A recovery line is a chain of dependencies, and a stage that under-delivers simply hands the problem downstream as extra cost. Pre-sort that lets mattresses through destroys shredder hammers. Screening that produces an over-wide oversize fraction buries the manual cabin. Metal extraction mounted too high above the belt lets ferrous drop back into the flow, and then pickers spend the shift lifting what a magnet should have taken. This table is the checklist we use to test each handover before the layout is frozen.

Stage inside a complete recovery line	What this stage has to achieve in the flow	Design point we fix before releasing fabrication
Receiving, dumping and coarse feed-in from collection vehicles	Accepts unsorted loads, meters them onto the first incline and survives the impact of a full tipper drop.	Hopper volume set against the largest vehicle in the fleet, discharge plate angle above 45 degrees, and a 10 mm wear liner at the dump zone.
Pre-sorting of oversized, bulky and hazardous items	Removes mattresses, cable drums, gas bottles and demolition debris before anything reaches a shredder rotor.	Aisle width for a three-tonne wheel loader and a bypass chute sized for anything longer than 1,200 mm.
Size reduction and screening into controlled fractions	Breaks material down, then separates it into an oversize and an undersize stream suited to the machines that follow.	Shredder throughput checked at the design density and screen aperture matched to the downstream sorting equipment.
Magnetic separation and eddy current extraction of metals	Lifts ferrous with an overband magnet and throws non-ferrous clear of the flow with an eddy current rotor.	Magnet gap, suspension height and the belt speed that keeps metal from being buried under a layer of film.
Manual sorting cabins for the residuals machines miss	Places trained pickers beside a wide, slow belt to pull the last contaminants and grade the material by hand.	Belt width and belt speed that leave a picker a realistic 1.5 to 2.5 second window on each object.
Baling, wrapping and dispatch of the graded product	Compacts each recovered fraction into a stable bale, wraps it and feeds it out to the loading bay.	Transfer drop height into the baler hopper and the turning area a forklift needs in front of the discharge.

02Material Properties and the Constraints They Impose

No two recovery streams behave alike, yet a handful of physical properties repeat often enough to plan around. Film, moisture, glass, grit and fibre turn up in almost every commercial mix, and each one attacks a different part of the machine. Choosing a conveyor belt supplier on price alone, without naming the stream, is how a line ends up with the wrong cover grade on day one.

Start by sampling your own material rather than a competitor's specification sheet. We ask clients to fill a 200-litre drum at three points on three different shifts, then weigh it, spread it on a 1 m2 tray and photograph it. Density, average object size, moisture content and the proportion of long thin items all fall out of that exercise, and each one changes a dimension somewhere on the drawing.

Contamination, moisture and the carryback they create together

Film and moisture are far more damaging in combination than either one alone. Dry film slips off a pulley cleanly; wet film sticks to the cover, wraps the return idlers and climbs back on the carrying side. Add food waste or green garden material, and the fines turn into a paste that glues a layer of grit to the belt surface, which then acts as grinding paste on every pulley it touches. A two-stage scraper on the head pulley plus an independent return-side cleaner is the minimum we will design in, and on very wet lines we add a wash box and a squeeze roller.

Sharp contamination follows a slower but equally predictable path. Glass shards and torn metal sheet nick the top cover first, and a nick becomes a cut within a day because the same point passes the same pulley a few thousand times per shift. What matters is not only the cover thickness but where the object enters the system, because a single metal detector positioned before the first shredder removes more future damage than any repair programme afterwards.

Property of the incoming waste stream	How it attacks the conveyor in service	Design response we write into the specification
Thin film, rope and textile wrap	Loops around rotating parts and idler ends, then builds into a growing donut that drags on the drive.	Labyrinth-sealed idlers with no exposed shaft, lower transfer drops, and guillotine edge guards at every nip point.
Glass, ceramics and jagged metal sheet	Punctures and slashes the top cover, and the cut then propagates a little further on every pass over a pulley.	A 6 mm or heavier abrasion-grade cover, a cold repair compound applied inside one shift, and detection as close to the source as possible.
Moisture, leachate and putrescible food waste	Turns fines into a sticky paste that clings to the belt, loads the return strand and corrodes the frame from below.	Covers pitched to drain, a two-stage head scraper plus a return cleaner, and a floor that slopes to a collection sump.
Dust from paper, cardboard and wood fractions	Lofts into cabins and drive houses, settles on beams and structures, and turns a confined space into a real ignition risk.	Enclosed transfers, dry extraction at each drop point, and bonding and earthing of every isolated metal component.
Bulk density that swings with the weather	A load running at 0.30 t/m3 in a dry week drops to 0.20 t/m3 after rain, so a full belt suddenly carries less mass.	Transport belts sized on the lightest credible density, with the manual cabin decoupled from that variation entirely.

Why we ask for a bulk sample before quoting a belt

An industrial conveyor belt built for graded stone handles a uniform 50 mm feed all day. A recovery stream delivers the same tonnage as a mixture of flat film, rigid bottles, wet fibre and loose grit in a layer that is 40 mm thick in one place and 250 mm in another. The belt does not care about the average; it cares about the point load under the worst object in the load, which is why a bulk sample tells us more than a tonnage figure ever will.

03The Equipment Combination Matrix for a Recovery Line

Six machine families cover almost every recovery duty, and most lines use four or five of them. The mix is decided by the fraction being handled and by how dirty the environment is, not by preference. As a conveyor belt manufacturer that also supplies idlers, pulleys and splicing, we end up specifying the interfaces between these families, and that is where the arguments usually start.

Read the matrix as a starting point, then test it against your own peak flow.

Conveyor family and its role on the line	Where it earns its place on a recovery line	Anti-jam and anti-wrap measures we insist on
Troughed flat rubber belt running on three-roller idlers	The workhorse for transport runs, inclines and the feed into every separation stage on the line.	Sealed idlers with no shaft ends, skirting that follows the belt with a 10 mm gap, and a drip slope at every wet point.
Apron or plate feeder with overlapping steel pans	Survives uncontrolled impact under a hopper lip, where a rubber belt would be torn apart inside a week.	Pans that overlap to keep fines out of the chain, a sprung impact bed, and a chain take-up a fitter can set in ten minutes.
Modular plastic or steel mesh belt	Handles wet, sticky and moderately hot fractions and shrugs off the small sharp pieces that pass screening.	Open hinge geometry that drains and self-cleans at each turn, plus side guards that keep thin film out of the sprockets.
Powered roller conveyor for outfeed and bale handling	Sequences bales and rigid containers on the dispatch side where a continuous walking surface is not needed.	Sprocket-driven rollers instead of chain loops wherever wire, string or plastic strapping can reach the drive.
Sloped chutes with replaceable wear liners	Joins stages where a powered belt would add another drive and another maintenance point for no benefit.	A chute angle above 55 degrees for film-rich material, 8 mm liners, and no horizontal ledge where scrap can park.
Spreading and levelling unit ahead of the sorting cabin	Thins a thick, clumped stream into a single layer so that pickers can actually see what is passing.	A levelling drum on a shear-pin drive, with the gap set 160 mm above the belt so oversized items ride over it.

04Capacity Check: From Tonnes per Hour to Belt Width and Belt Speed

Sizing is arithmetic, and the arithmetic is short. The work sits in the assumptions, which is why so many lines are commissioned with an 800 mm belt doing a 1,200 mm job, or the reverse. We run the same four steps on every project before a frame drawing is issued.

Step one fixes the design tonnage.

Fixing the design tonnage and the peak factor you have to cover

Take the annual or contractual throughput, divide it by realistic operating hours, and then multiply by a peak factor. For a municipal contract, an average of 25 t/h across a two-shift week is normal, and we would apply a factor of 1.4, giving a peak design duty of 35 t/h. The peak factor matters more than the average, because the belts and the drives are bought once and then asked to survive the busiest hour of the year for a decade. Operators who size on the annual average usually add a second shift of maintenance later.

Step two converts mass into volume, and this is where recovery lines part company with mineral plants.

Mixed commercial waste that contains film and fibre typically occupies 0.20 to 0.30 t/m3, against roughly 1.6 t/m3 for crushed stone. At 0.25 t/m3, a peak duty of 35 t/h becomes 140 m3/h, or about 0.039 m3/s. That is the number the belt geometry has to swallow, and it explains why a recovery belt is often wider than the tonnage alone would suggest. Confusing the two densities is the single most common sizing error we are asked to correct.

Working the volume back to a belt width and speed

Divide the volumetric flow by the belt speed to get the load cross-section the belt must carry. At 1.2 m/s, 0.039 m3/s needs about 0.032 m2 of material on the belt. A 35-degree three-roller trough on an 800 mm belt develops roughly 0.043 m2 at normal surcharge, which covers the duty with room for a lumpy load. A 650 mm belt in the same trough offers only about 0.028 m2, so it would run overloaded at the peak and spill at the skirting. Between those two arithmetic results sits the whole decision, and it takes about five minutes to do properly.

Belt speed is also a quality decision, not only a capacity one.

Fast belts throw material, raise dust and hammer the skirt rubber, so where the tonnage allows we prefer width over speed. On a transport run, 1.2 to 1.5 m/s keeps a light load settled. On the manual sorting belt the picture reverses completely, and a different rule applies.

Matching the numbers to the sorting cabin rhythm

Pickers cannot read a belt moving at 1.2 m/s. That is why the cabin always gets its own slow belt, usually 0.30 m/s, and the transport sections are joined to it by a metering feed. Work the cabin duty separately. With material spread to a 50 mm layer across a 1,200 mm belt, 0.30 m/s delivers 1.2 by 0.05 by 0.30, which is 0.018 m3/s, or roughly 16 t/h at 0.25 t/m3. To clear a 25 t/h line at that spread and speed you need two sorting belts, or one belt plus a second cabin. The two figures have to be reconciled on the flow sheet, not left to the electrical contractor to discover at commissioning.

05Durability Design Against Tear, Impact and Wrap

Recovery belts fail from the outside in. Tension is rarely the limiting factor; the surface is. A belt that would run for eight years on screened aggregate can be scrapped in eighteen months on a mixed municipal stream, and the difference is entirely in how wear and impact were handled at the drawing stage. Any conveyor belt factory will quote a cover grade, yet far fewer will ask what actually lands on the belt.

Choosing cover grade and thickness for a mixed, sharp feed

Cover thickness follows the sharpest object in the stream rather than the annual tonnage. Where glass and torn sheet metal are present in any quantity, we start at 6 mm and often specify 8 mm on the receiving and shredder-outfeed belts. The rubber itself needs to be abrasion resistant rather than merely oil resistant, because the load carries grit that behaves like grinding paste once it is pressed into the surface. Our note on cover grade selection explains how those wear figures are read. A rubber conveyor belt carrying the heavier grade on its dirtiest leg, rather than its longest one, is usually the better value on the order.

Managing impact at the points where material actually lands

Impact energy scales with drop height, and a modest 1 m fall behind a shredder produces the same punch as a much heavier object dropped from 300 mm. Under the hopper and the shredder outfeed we mount an impact bed with rubber-covered rollers, extended 1.5 times the chute width on both sides of the load point, because material scatters across that span.Chute liners of 8 to 10 mm replaceable plate take the direct hit and are far cheaper to renew than a belt section and the labour to splice it.Where the drop cannot be shortened, a rock box or a short dead bed of material is preferable to any steel the stream can reach.

Every joint is a weak point in a recovery belt, and a mechanically fastened joint that survives on stone will often fail early on a stream carrying string and wire. We favour vulcanised splices on the main runs and reserve mechanical fasteners for short, low-tension segments that a crew can change in an hour. Edge damage deserves equal attention, because a belt that has been trained badly will wear through both edges before the centre cover shows any wear at all. The table groups the protective measures we routinely cost into a recovery line, together with the condition that makes each one worth its money.

Protective measure on the belt or the frame	What failure it prevents in service	The condition that makes it worth the spend
Abrasion-resisting top cover at 6 to 8 mm	Delays the first puncture and slows the sliding wear under a fast, gritty load.	Justified whenever glass, ceramics or sheet metal appear in the feed more than occasionally.
Breaker fabric or steel cord carcass	Carries tension and resists the longitudinal rip that a trapped bar or rod will start.	Worth adding on any belt working directly below a hopper or a shredder discharge.
Sprung impact bed under the loading zone	Absorbs the energy of a 250 mm object landing from a metre, which rigid idlers cannot.	Essential wherever the chute cannot be shortened or a rock box cannot be fitted.
Replaceable liners in chutes and skirting	Wear away in place of the belt and can be renewed inside a single shift.	Cost-effective at any transfer with a drop above 600 mm or a high grit content.
Cleaner blades, ploughs and edge seals	Keep carried material off the return strand, which is where most wrap begins.	Sensible on every line, and non-negotiable on wet or food-waste-bearing streams.
Cold-repair kit and spare belt length held on site	Converts a 2 m slit into a thirty-minute repair rather than a two-day belt change.	Store belt for the widest splice on the line plus three metres, and audit the kit quarterly.

06Cleaning, Carryback and Transfer-Point Discipline

Carryback is the quiet destroyer of recovery lines. Material that clings to the belt past the head pulley falls onto the return strand and starts the sequence of mis-tracking, edge wear and wrap that eventually stops the line. On a wet stream that spill can reach 3 percent of throughput, which on a 25 t/h line means three quarters of a tonne an hour on the floor.

How much belt speed changes what a cleaner can remove

A scraper blade works by forcing the belt to change direction while the blade holds the material back, so its effectiveness falls as speed rises. That is the practical reason we match belt width to the duty before we reach for belt speed, and it is why a cleaner tuned on a 1.0 m/s belt will not perform the same way at 1.6 m/s. On recovery lines with a wet, paste-like carryback, a single primary scraper is rarely enough. We specify a primary blade at the head pulley, a secondary blade behind it, and an independent return-side cleaner positioned where the belt has already started to flatten. A conveyor belt cleaner chosen on price alone, without checking the blade compound against the surface, will either cut the cover or leave the paste behind.

Transfer points deserve the same discipline. Every chute should be enclosed, drop heights held as low as the layout permits, and skirting rubber set with a 10 mm gap that follows the belt instead of pressing on it. Where a chute is the only practical link, angle it above 55 degrees for film-rich material and avoid any horizontal shelf where scrap can park and then release in a slug.

Some fractions resist scraping entirely. Sticky wet paper and food residue will smear rather than release, and forcing a blade harder only wears the cover. In those cases we switch the strategy to washing: a spray bar followed by a squeeze roller removes the layer before it reaches the first return idler. Whatever the method, the sequence we follow when diagnosing carryback is documented in our note on belt cleaning methods and when scrapers will not work, and it starts with checking the belt surface and the load profile before touching the blade pressure. A conveyor belt distributor who can supply the cleaner, the blade compound and the splice material from one source saves a great deal of argument when a line is down.

07Dust, Leachate, Noise and the Working Environment

Environmental design on a recovery line is not decoration. It decides whether the plant keeps its permit and whether the people in the cabin can work a full shift without respirators.

Dust control and the explosion risk at a transfer point

Paper, cardboard and wood dust is combustible, and a conveyor transfer is exactly the kind of confined, aerated, ignition-rich location where a dust cloud becomes dangerous. The standard response is to keep dust from becoming airborne in the first place, then to remove it safely where it does. We enclose each transfer, fit dry extraction at the chute lip, keep the enclosure under slight negative pressure, and bond and earth every isolated metal part, because a single ungrounded sensor bracket can hold a static charge all shift. A dust-resistant conveyor belt selection is part of the answer, but the belt alone cannot compensate for an open transfer.

Every tonne of wet municipal waste carries water that will leave the load somewhere on the line. If the floor does not direct it, that water collects under the frames, corrodes the structure and produces the smell that neighbours notice before the plant does. We slope the floor beneath each conveyor to a collection sump, keep conveyor legs out of the water path on 150 mm plinths, and specify a covered drain rather than an open channel wherever the odour load is high. For the belts themselves, the low point of every run gets a drip slope so the liquid leaves the belt instead of riding back. Where spray cleaning is used, the water volume and the drainage capacity have to be balanced, which is a common oversight we cover under spray cleaning on conveyor belts.

Noise, walkway clearance and the space a crew actually needs

Steel chutes, unsupported spans and loose covers all radiate noise, and a recovery line that measures above 85 dB(A) at the walkway triggers hearing protection rules and slows inspections. Lining chutes, closing open spans and fixing covers properly brings most transfers down several decibels. Clearance matters just as much, and it is the first thing sacrificed when floor area is tight. We hold a 700 mm walkway on both sides of every drive station and provide a maintenance opening of at least 900 mm where a pulley or a bearing has to be changed. A crew that cannot reach a bearing safely will postpone the job, and the postponed job becomes the breakdown. When a plant needs a large volume of matched spares across many segments, ordering wholesale conveyor belts against one common specification, rather than buying piecemeal, keeps the store simple and the changeover predictable.

08Drives, Take-Up and Stopping Logic

The drive end of a recovery line carries more stress than the tonnage suggests, because starting torque under a full, heavy belt is far higher than running torque. Getting the drive and the take-up right costs little at design stage and prevents a class of problems that no belt can solve.

Selecting and protecting the drive package

A shaft-mounted gearbox is adequate on short runs where the belt is never fully loaded at start. On long inclines and on any belt fed from a hopper, we prefer a foot-mounted unit with a fluid coupling or a soft-start arrangement, because it lets the belt accelerate in three to five seconds instead of in one. That single change reduces the peak tension on the splice and on the take-up. Where the plant runs several drives from one supply, a transmission belt manufacturer can supply the drive belts from the same source as the conveying belts, which keeps the spare list short. Selecting a V-belt manufacturer on profile availability rather than on price alone avoids the familiar situation where a belt is out of stock and the line is idle for a week.

Take-up travel has to cover the elastic stretch of a new belt plus the extra stretch that a recovery belt develops in its first month. For a fabric-carcass belt we normally allow 1.5 to 2 percent of the belt length in the shortest centre distance, and more on a long incline. Stopping logic matters just as much on a mixed line: the belt that carries the material away must stop last, and the feed must stop first, so the line empties from the discharge end backward. A pull-cord running the full length of every accessible segment, with the trip arranged to stop the feed before the discharge, is the cheapest safety decision on the drawing. Idlers and rollers on these runs should be selected for slow speed and dirty conditions rather than for high load, and our guide to conveyor roller selection sets out the sealed, corrosion-resistant types that survive a wet recovery hall.

Bearings are where that choice pays off. A sealed, greased-for-life bearing in a wet, gritty environment will outlast a regreasing type simply because nobody has to find access to it every month. Choosing corrosion-resistant conveyor rollers up front costs a little more per unit and removes a recurring labour cost that rarely appears in a capital estimate.

09Matching the Sorting Cabin Rhythm to the Line

A sorting cabin is a production station with a human being in the loop, and its output is set by biology rather than by horsepower. The belt speed, the width of the picking lane and the number of pickers per lane decide how much material a cabin can actually clean, and none of those numbers appears on a drive nameplate.

Picking windows, spread and cabin geometry

A picker needs roughly half a second to see an object, decide on it and reach for it, and considerably longer if the object is small or partly buried. Everything the layout does should protect that window. A 1,200 mm sorting belt lets one picker work a lane without leaning across the stream, and a 0.30 m/s belt speed gives each object about four seconds in front of a fixed worker on a 1,200 mm reach. Narrow the belt to 800 mm and the reach shortens, which lets you place two pickers per side; widen it to 1,600 mm and the far edge becomes useless because nobody can reach it safely. Set the picking height between 900 and 1,100 mm above the platform, keep the belt close enough that an object can be dropped into a chute without a full turn, and provide a footrest so the worker is not standing on the frame.

Spread matters more than hours.

Pickup rate falls through a shift in a predictable way, which is why we size cabins on a realistic 70 to 80 percent of the first-hour rate rather than on a peak figure. Rotation between sorting and other duties helps. So does positioning the cabin after the strongest separation stages, since every tonne that a magnet or a screen removes is a tonne the pickers never see. When a client reports that a cabin cannot keep up, the cause is usually upstream: the spread is too thick, the shredder has produced a long, tangled fraction, or the screen aperture is letting mixed material through. Our notes on conveyor belt splicing checks and on sealed rollers for fine dust both come into play here, because a cabin that is fed by a mistracking belt or a seized idler loses capacity faster than any staffing decision can restore it.

10Purchase List and Acceptance Criteria

A recovery line is bought in pieces by several trades, and the biggest single cause of a poor start-up is that nobody wrote down what each piece had to deliver. The purchase list below is the version we attach to a tender so that every bidder prices the same scope.

What each line item should carry

Write the duty, not just the hardware. A line item that says only "conveyor belt, 800 mm" will be supplied with whatever cover grade happens to be in stock, and the argument starts at the first puncture. Each belt should appear with its measured duty, its cover grade and thickness, its carcass, its splice method and the segment it serves. Idlers should be listed by diameter, bearing and sealing class, pulleys by face width and lagging, and the cleaning system as a complete set with a spare blade. Our guide to the conveyor belt splicing methods is a useful cross-check on the joint specification, and the full range of conveyor belt products can be matched against the duty list before the enquiry leaves the office.

Purchase line item on the enquiry sheet	What it has to state so the bid is comparable	How we verify it before accepting delivery
Belt specification for each individual segment	The duty in tonnes per hour, the cover grade and thickness, and the carcass, written per belt rather than as one general line.	Measure cover thickness with a gauge on arrival and compare the certificate against the ordered grade.
Idler and roller schedule with sealing class	One common diameter and bearing across the plant, with the seal type named so nothing exposed can wrap.	Inspect the seal design by eye and confirm that no idler shaft end protrudes beyond the frame.
Pulley drawings with face width and lagging	A face at least 100 mm wider than the belt on each side, and lagging matched to the wet or dry duty.	Check face width against the belt and confirm the lagging pattern before any crowning is machined.
Splice method and joint position schedule	Which joints are vulcanised, which are mechanical, and where each one sits on the layout drawing.	Walk the line after start-up, inspect every joint and mark its position on the as-built drawing.
Cleaner, skirting and liner set with spares	The complete cleaning system, including blade compound, liner thickness and one spare set held on site.	Confirm the spare set is physically on the shelf, not promised in a delivery note.
Documentation: flow sheet, drawings and certificates	A flow sheet with the design tonnage on every segment, the drawing set, and the material and test certificates.	Refuse the handover payment stage until the as-built flow sheet and certificates are signed off.

Acceptance tests we run before signing a handover

Acceptance should be measurable, not a matter of opinion. We run a loaded trial at the design peak for at least four hours, record the belt tracking at the head, tail and mid-span, and log any material left on the floor after the run. A belt that cannot hold position under peak load will mis-track within a month, whatever an empty tracking test showed. Noise is measured at the walkway, dust is checked at each transfer with the extraction running, and every emergency pull cord is tested by the person who will use it.

11Common Design Mistakes on Recycling Lines

Most of the design faults we are called in to correct follow one of four patterns, and all four are visible on a drawing long before steel is cut.

Wrap dead spots and where they come from

A wrap dead spot is a place where film and rope inevitably collect because a rotating part has an exposed end that the material can reach. Every idler with a protruding shaft, every gap between a guard and a pulley hub, and every unsupported span where the belt flutters creates one. The cure is design rather than maintenance, because a dead spot that has to be cleaned by hand every shift will be cleaned for a few months and then abandoned. Fit sealed idlers with no exposed shaft, keep guards within 5 mm of the rotating face, and remove the pockets where material can settle. A related failure is mis-tracking, and the sequence of causes is set out in our EP conveyor belt tracking guide, which starts with the belt and the frame rather than with a training idler.

The second mistake is a cleaner that physically interferes with the structure it is meant to help. A return cleaner fitted where the belt has not yet left the head pulley will rub, and a plough set too close to a chute mouth blocks at the first wet day. Leave 300 mm of belt travel between the primary and secondary blades, and check the cleaner against the chute drawing before fabrication rather than on site. Excessive drop height is the third pattern, and it is almost always created by the decision to raise one stage rather than lengthen a chute. A 1.5 m drop onto a belt that already carries glass will double the puncture rate, so shortening the fall with a rock box is usually the cheapest durability improvement available. The fourth mistake is a plant built around four different roller diameters, three belt widths and two bearing types, which turns a ten-minute fault into a store-room search and a two-day wait.On aggregate lines we see the same lesson, and our note on belt selection for crusher and aggregate plants makes the argument for standardisation from a different direction.

12Commissioning Follow-Up and the First-Year Watch List

Commissioning is when a design meets its material for the first time, and the first three months decide whether the line will settle or keep generating work orders. We keep a short watch list and revisit it at the end of the first month, the first quarter and the first full year.

Record belt tracking at every station weekly for the first month, because a new fabric belt stretches unevenly and the take-up moves more in thirty days than in the following six months. Photograph each splice, note any cover damage and log the position so the next inspection has a baseline to compare against. Where a wet stream is in use, check the drip points and the sump carefully in the first rain-heavy week, since that is when leachate behaviour is worst. Our observations on heavy-duty belts in mining-style duty transfer directly to a mixed recovery stream, where the same impact and wrap mechanisms dominate. A line that is quietly wearing faster than expected almost always shows it first in the carryback volume on the floor, so we weigh that floor sweep and record it. Where a plant also handles port or bulk transfer duty alongside the municipal stream, our note on port bulk material handling covers the extra interface points.

By the end of the first year, the useful questions are simple. Has the carryback mass changed? Has any splice moved from its recorded position? Has any segment been re-specified because the actual density came in below the design figure? A yes to any of those three means the original assumptions need revisiting before the next belt order goes out, and that conversation is far cheaper than a second redesign.

13Questions Buyers Ask Us Most Often

How do I size a recycling conveyor belt when the tonnage is only an estimate?

Size on the lightest density you expect, never on the annual average. Convert the peak hour into cubic metres and confirm that the load cross-section at your chosen speed still has about 30 percent headroom. When the tonnage figure is soft, buy one width larger and keep the speed down, because width is paid for once and excess speed is paid for every shift.

Why does film wrap around idlers so much faster on a recycling line?

Because an exposed shaft end is a hook, and a light, springy material finds it within a single shift.

Can one belt serve both the transport runs and the sorting cabin?

No, and forcing it is a false economy. Transport wants 1.2 to 1.5 m/s to move the tonnage, while a cabin wants roughly 0.30 m/s so a picker can act on what passes. Link the two with a metering feeder and give the cabin a dedicated slow belt.

What cover thickness should a recycling belt have?

Start at 6 mm and step up to 8 mm wherever glass, ceramics or torn metal sheet appear in the feed, then measure the cover with a gauge when it arrives instead of trusting the label.

Does the metal detector belong before the shredder or after it?

Before it, wherever the layout allows, because a shredder that has already swallowed a steel bar has already damaged its hammers and scattered fragments through the rest of the stream. A second detector downstream still earns its place on high-value lines by catching wear metal the first unit cannot see. Where only one aperture is available, put it on a short straight run after the last contamination risk and keep forklift routes and compressors away from the head. Always pair the detector with a rejection device that clears the object before the next stage, not after it.

How do we deal with dust and odour inside the building?

Dust is controlled by enclosure first and extraction second, since no extraction system keeps pace with an open transfer. Odour is controlled by keeping material moving and the floor dry, because a wet, stagnant sump is what the neighbours smell first. Hold the enclosure at slight negative pressure and bond and earth every isolated metal part. Measure the dust loading at each chute lip after start-up rather than assuming the design figure still holds.

Is a modular belt always better than a rubber belt for mixed waste?

Not automatically, and that assumption has cost our clients real money, because a rubber belt on a troughed frame handles heavy, lumpy, abrasive mixed waste far better than a plastic modular belt, which earns its place only on wet, sticky or hot fractions and on the short segments where drainage matters most.

How many pickers does a sorting cabin need?

Work backward from the cabin duty rather than from the labour budget. A 1,200 mm belt at 0.30 m/s with a 50 mm spread clears about 16 t/h, so a 25 t/h line needs two sorting belts or two cabins. Then add a margin, because pickup rate falls through a shift and a cabin sized on the first-hour rate will lag by mid-afternoon.

What is the most expensive design mistake on a recycling line?

Underestimating wrap and carryback, which only shows up later as seized idlers, torn covers and a floor that has to be swept every shift.

Related Products You May Need

• Rubber conveyor belts for receiving and transport runs
• EP fabric carcass belts for long incline sections
• Steel cord belts for the heaviest recovery duties
• Chevron belts for steep infeeds and ramps
• Heat resistant belts for hot fraction handling
• Sealed conveyor rollers and idler sets
• Full conveyor belt and component catalog

Related Blog Posts

• Impact rollers and loading zone protection
• Dust resistant conveyor belts for enclosed transfers
• Conveyor belt cleaning methods and schedules
• Return side belt cleaning
• Abrasion resistant belts for quarry and aggregate duty
• Conveyor roller guide 2026: types, materials and load ratings
• EP conveyor belt tracking guide