Battery Line Conveyor Belts: Static, Solvent and Metal Control

The first question we ask a battery line engineer is not how long the belt lasts. It is what the belt leaves behind. On a cell assembly line a carrying surface that looks healthy under a torch can still be writing grey smears onto every case that passes, and nobody notices until incoming inspection pulls a tray and finds the wipe dirty. That is the moment the conversation changes from wear life to contamination, and it is the reason we treat a battery conveyor belt as a cleanliness component first and a wear part second.

This article is written for the people who specify, buy and qualify flat belts on lithium cell lines, module lines and pack lines. We build belts for those lines, and we also get called in after a rejection event, which is where most of what follows came from. If you are still comparing cover compounds at the catalogue level, the PVC conveyor belt product range is the right place to start on construction. What this page adds is the qualification logic that sits on top of it.

Send us your line layout and we will match a belt construction to it

01Debris and Metal Come Before Abrasion on a Cell Line

Almost every RFQ we receive from a battery plant opens with abrasion resistance. Buyers ask for a thicker cover, a harder durometer, a longer wear guarantee, because that is the language they inherited from mining and cement. It is the wrong first question for this industry. A belt on a cell line is not being punished by tonnage; it is being asked to stay chemically quiet while travelling at low tension over small pulleys, and the failure that hurts is the one where a fragment of the belt ends up inside a cell.

Put the three risks in order and the specification writes itself. Debris and metal contamination come first, because their consequence is a scrap event that reaches the customer. Static comes second, because its consequence is usually a fault at the tester or a shock report at the transfer, and because it can be verified with an agreed measurement rather than argued about. Abrasion resistance comes third, and on most cell lines it is a service-life issue rather than a quality issue. A belt that wears out in two years is an inconvenience. A belt that sheds carbon black into an open cell is a recall.

We saw this clearly on a 400 mm wide stacking line running at roughly 25 m/min. The belt had been selected on abrasion grade and had run for eleven months with almost no visible wear. The plant was rejecting cells at a rate they would not put in writing, and the reject reason was a dark transfer on the aluminium case. The belt was not worn. It was bleeding plasticiser and compound residue from the cover, and the wipe test that finally proved it took four minutes to run.

That is why we now insist on a debris conversation before a compound conversation. Ask what the carrying surface touches, ask what the wipe looks like after two hundred passes, and ask whether the splice sits under a transfer plate where a step will be scraped by every cell. Only then talk about cover thickness. Buyers who want to compare constructions across the whole catalogue can use our conveyor belt manufacturer overview as a map, but the ordering of priorities on a battery line is not the same as on a bulk line.

02Where the Black Shedding Actually Comes From

Shedding is rarely a single mechanism. It is a stack of small contributors that only becomes visible once the line is running clean enough for a grey wipe to stand out. The cover compound can release plasticiser that has not fully migrated into the polymer, and that film picks up dust, pouch fibre and glove residue until it looks like dirt from the belt itself. The slit edge can carry a burr or an exposed warp yarn, and every pass over a small pulley works that edge until a fine black fuzz appears on the frame. The splice can sit a fraction of a millimetre proud, and the cell case shears across that step thousands of times per shift.

Then there is the part nobody puts in the RFQ. What is the belt running against? A steel guide rail wearing through its coating, a roller with a dry bearing, a brush that has lost half its bristles and is now dragging its backing strip. We have opened guarding on a line that had been blamed for belt shedding and found the belt clean while an idler end cap was grinding itself into powder a few centimetres away.

The diagnostic we trust most is dull and cheap. Cut a white, lint-free wipe, mark a 100 mm square, and rub the carrying surface with a fixed number of strokes at a fixed pressure. Do it on a new roll, do it after a shift, do it after a wash-down. If the wipe goes grey on a belt with no visible wear, you have a compound or edge problem. If the wipe stays white but the frame under the return run is dusted, you have a mechanical interference problem and no compound change will fix it. This test is worth writing into the line's routine; our quality assurance team runs the same wipe on every roll before it is packed.

Edge quality deserves its own paragraph because it is the most common structural cause we see on thin belting. A blade that is dull, or a slitting speed that is too high for the compound, leaves a micro-rough edge with exposed fabric. On a 1.2 mm belt running over a 40 mm pulley that edge works hard. Buyers who compare only cover hardness never see this coming, which is why we send slit samples rather than photographs when a plant is chasing a debris complaint. The wider comparison of cover families sits in our PVC conveyor belts section, and it is worth reading alongside this page because edge behaviour changes with compound.

Splices are the second structural cause and the one with the least tolerance. A hand-skived finger splice can be flat at the centre and proud at the shoulders, and where every cell crosses the same point you get a wear track that looks like a design fault. Ask for the step to be measured, not described. On thin belting we work to a step a fingertip cannot catch, and we put it on the drawing rather than in a sales conversation.

03Metal Foreign Matter Is a Separate Problem From Static

Buyers often fold metal contamination and static into one line item called "anti-static". They are not the same failure and they are not solved by the same belt property. Static is an electrical condition of the surface, and it can be measured, argued about and written into a tolerance. Metal foreign matter is a physical object that got somewhere it should not be, and no surface resistance value will keep it out of a cell. Our rubber conveyor belt lines are built for heavy duty, and the same discipline of thinking about where loose material travels applies here in a much smaller scale.

Start with the splice, because that is where metal enters a thin belt most often. A mechanical lacing with steel staples is fine on a grain conveyor and unacceptable on a cell line, not because it fails quickly but because it fails in a way that distributes small steel pieces along the run. We specify endless belts or a pressed splice for anything handling cells, and if the frame cannot be opened for an endless belt we say so at quotation stage rather than after installation. The second entry point is the edge reinforcement. A belt with a fabric or cord edge that has been stapled, riveted or clamped to stop curl is a metal reservoir sitting right where the product rides.

Then look outward from the belt. Guide rails, side guards, transfer plates, sensor brackets, tooling left inside the frame during a changeover, and the small springs inside a worn roller end cap all sit within a metre of the product path. Our conveyor components page lists the parts we supply for these frames, and when a plant asks us to quote a full conveyor we push for polymer or non-metallic inserts at every contact point near the cell path. It is a small design change that removes a category of risk permanently.

Detection is the last line and it should never be the first. Metal detection at the stacking station will catch a fragment that is already in the tray, but by then the fragment has travelled through several handling steps and the quarantine decision is expensive. The plants that handle this well work on exclusion first and detection second, and they treat every metal detector trip as a root-cause investigation rather than a reset button. If a trip is logged and cleared without anyone walking the frame, the same fragment source will produce another trip within a fortnight.

One more point that plants learn the hard way. A belt that is metal-free can still deliver metal if its pulleys are not. We have seen shafts with a worn keyway throw a sliver after a jam, and we have seen a transfer plate with a lifted corner shave aluminium off a cell case on every pass. When we quote conveyor pulleys for a cell line we ask for the jam history, because a pulley that has been through three stalls is a different risk from a new one.

04Static Is Second Priority, and It Needs an Agreed Basis

We are not going to tell you static does not matter. It matters, particularly on dry lines where cells and pouches slide across a cover at speed, and it matters most at the point where a charged part meets a grounded fixture. What we will tell you is that "anti-static" as a word on a datasheet is close to meaningless, because two suppliers can both claim it and mean numbers that differ by four orders of magnitude.

The only workable approach is a basis agreed in writing between the belt maker and the plant, covering the instrument, the probe, the test voltage, how long the voltage is applied before the reading, the conditioning of the sample, which face is measured, and whether the surface is clean or as-run. Change any one of those and the number moves. We have sat in a room where the plant's lab reported one figure on a sample and our lab reported another on a cut from the same roll, and both were correct. Different probe geometry, different contact pressure, different dwell.

So write the basis into the purchase order. State the resistance band you will accept, state the probe and the voltage, state the face, and state that the reading is taken after conditioning at a defined humidity and temperature. Then agree what happens on a fail, whether that is a retest on a second sample, a concession with a shorter replacement interval, or rejection. This sounds bureaucratic until the first dispute, at which point it is the only thing that keeps a line running.

Two practical notes from our own bench. First, resistance drifts with humidity, so a belt that passes in a monsoon month can fail in a dry winter if the basis does not fix the conditioning. Second, resistance changes with contamination. A belt that has picked up pouch dust reads differently from a clean one, and if your acceptance test is done on a wiped sample while the line runs soiled, you are qualifying a surface that does not exist in production. Where the plant runs a dry room, we ask for the room's own condition range and test inside it. The material comparison in our comparison of rubber, PVC and PU covers explains why the conductive pathway differs between compound families.

Resistance values are not a quality badge. A belt that reads far below the agreed band is not automatically better, because a compound loaded with conductive filler to chase a number usually gives up something in mechanical strength or in solvent resistance. The right target is a band, agreed by both parties, with a test method both parties can reproduce. That is what we mean when we say static is the second priority and a contractual one.

05Thin PVC, PU and Fabric Constructions Compared for Cell Lines

Once contamination and static are settled as the governing risks, construction becomes a shortlist rather than an open field. On cell and module lines we mostly see three families. A thin PVC cover over a fabric carcass is the workhorse, because it is dimensionally stable, easy to splice and available in a range of surface finishes. A PU cover is chosen where solvent contact or hydrolysis is the dominant worry and where a higher cut and tear resistance is wanted at the same thickness. A fabric-based belt with an impregnated surface turns up on lighter transfer sections where the load is small and the pulley diameters are tiny.

Thickness is where buyers over-specify. A 2 mm cover feels more robust and costs more in bending resistance, and on a 40 mm pulley it may crack at the back long before a 1.2 mm belt would. On a cell line the belt is not carrying rock; it is carrying a few hundred grams to a few kilograms per metre. Going thinner usually improves tracking, reduces the power needed at the drive and lowers the splice step, and all three of those help the contamination story. Our lightweight conveyor belts range exists for exactly this class of duty.

Surface finish deserves a line of its own. A heavily textured or fabric-impressed cover holds dust in the valleys, and no amount of brushing gets it out once compound residue has cured into the pattern. A fine matte finish wipes clean and shows contamination early, which is what you want on a line where a grey wipe is your alarm. Glossy finishes look cleaner in a photograph and are less forgiving in a dry room, because a glossy surface slides and generates charge more readily than a matte one.

The carcass matters for a reason that is not obvious. A fabric carcass with a loose weave can wick liquid along the belt through capillary action, so a small solvent spill at one station appears as a swollen edge three metres away. A tightly woven carcass with a well-bonded skim coat resists that. If the line has any wet station at all, ask how the carcass is sealed at the slit edge, because that is where wicking starts. Buyers comparing grades across the whole catalogue can start from our conveyor belt supplier index page, and the guide to PVC belt types and grades goes deeper into cover families than this page can.

Modular plastic belts are the alternative some plants raise at this point. They are easy to clean and they can be built with conductive grades, but the hinge gaps trap debris and the rod ends are a wear item, so on a cell line they tend to move the contamination problem rather than remove it. Our modular conveyor belt pages set out where that construction earns its place, and the short answer for cell handling is that it usually does not unless the line is washed down heavily.

One purchasing note. Battery projects rarely buy one belt; they buy a spare set and a replacement for every station, and they buy again when a line is duplicated. That pattern makes a wholesale conveyor belts arrangement sensible, provided the compound and the splice are frozen as a controlled specification rather than re-quoted each time. We have seen a plant accept a substitute compound on a repeat order because the price looked better, and spend a quarter chasing a debris problem that had been solved two years earlier.

If your plant runs both cell contact and module handling, do not use one construction across both. Split the specification at the boundary where the product stops being sensitive, and let the module side have the heavier belt it needs. We quote these as two line items and label them by station, because a single part number covering both ends of a plant is how contamination ends up on the wrong conveyor. For plants that also run drive belts on the same site, the transmission belt manufacturer range is a separate conversation with its own rules, and the V-belt manufacturer page sets out the drive belt options that sit on that side of the plant.

Availability is part of the specification. A six-week lead time is fine for a greenfield line and unacceptable for a plant running three shifts with one spare on the shelf. Ask for the compound code to be recorded on the belt edge, so a replacement two years later matches the original rather than the nearest catalogue grade.

06Solvent, Electrolyte and Wash-Down Contact on the Carrying Cover

A cell line is not a wet process, and that is exactly why solvent contact catches people out. The spill is small, it happens at one station, and it is wiped up within the shift. What remains is a patch of cover that has been softened, and a cut edge that has drawn liquid into the carcass. Weeks later the belt is curling at that edge and the plant is looking for a mechanical reason that is not there.

The chemistry matters less than the exposure pattern. A brief splash of a polar solvent on a PVC cover can extract plasticiser from the surface layer, and the cover that was flexible comes back stiffer and slightly shrunken. Repeated small exposures compound that, and the shrinkage shows up first as a wave along the edge, then as tracking drift, then as a crack at the splice. Electrolyte is a different story because it leaves a conductive salt residue behind, and that residue changes both the surface resistance reading and the corrosion risk on any steel nearby.

So the question to ask is not "is this belt solvent resistant" but "which chemicals, at what concentration, for how long, and how often". Ask the plant to list every wipe-down agent and every process chemical that can plausibly reach the belt, including the ones used by the cleaning contractor at weekend. We have seen a line qualified on process chemicals alone, then degraded by a general-purpose degreaser that the cleaning crew brought in on their own initiative. On the material side, our chemical resistant conveyor belt range covers the compounds we would shortlist, and the oil resistant conveyor belt page explains the swell behaviour that also governs many solvent exposures.

Wash-down also interacts with tension. A belt that has been wetted and dried repeatedly will creep slightly, and a line that was tensioned correctly at installation can be over-tensioned a month later without anyone touching the take-up. Over-tension shows up as edge wave, then as splice failure, then as a debris complaint. Mark the take-up position at commissioning and check it at every service. Every industrial conveyor belt we ship leaves with a commissioning sheet for this reason, and our conveyor belt factory keeps the tension record against the serial number so a replacement can be pre-set.

If the plant is buying through an integrator, make sure the chemical list travels with the order. A conveyor belt distributor who never sees the cleaning schedule cannot protect you from a degreaser choice made by a contractor. We would rather have an awkward conversation at quotation than a curl complaint at month six.

Two bench checks tell you most of what you need before a line trial. Cut a strip, measure its length and thickness, immerse it in the actual chemical at the actual concentration for the actual contact time, then measure again after a recovery period. A swell of a fraction of a millimetre in thickness and a length change that does not recover is a warning. The second check is a wipe after immersion, because a cover that has been softened releases compound far more readily, and that is the mechanism that turns a spill into a debris event.

Edge sealing is the cheapest insurance on this list and the one most often skipped. A sealed edge stops wicking, keeps the carcass dry and removes the fray source at the same time, so one process step pays for itself twice. On belts under about 2 mm we seal as standard for battery customers, and we say on the drawing which faces are sealed.

07How This Differs From Anti-Static Drive Belts

There is a page on this site about anti-static belts for power transmission, and readers sometimes arrive here expecting the same content with a different product photograph. The two are different problems and it is worth being explicit about the boundary. A drive belt sits between two pulleys and transmits torque; the static question there is about charge building on a belt that is running at high linear speed in a confined guard, often near a drive motor and its electronics. A conveyor belt carries product; the static question here is about charge on a surface that touches the cell, the pouch and the operator's gloves.

The failure consequences differ too. On a drive, a static event is an electrical nuisance that can damage nearby electronics or give a shock during maintenance, and the belt is not in contact with the product. On a conveyor, a static event happens on the same surface the product sits on, so the charge can transfer to the cell, upset a protection board, or attract dust to a surface that was clean a minute ago. That attraction effect is the one people forget. A charged cover pulls airborne particles onto itself, and then the belt becomes its own contamination source without shedding a single fragment of compound.

Read the drive belt material for what it is, which is a treatment of charge on a transmission element running under tension at speed, with its own measurement basis and its own installation rules. The anti-static power transmission belt factory page covers that ground, including how a conductive pathway is built into a V-belt or a timing belt and why the drive end of a machine needs a different conversation from the carrying end. Nothing in this article repeats it.

What we cover here is narrower and, for a battery plant, harder. The carrying surface has to be clean, low-shedding, free of metal, compatible with the chemicals on the line, and only then conductive enough to sit inside an agreed band. Those five requirements pull against each other. A compound loaded for conductivity can be more brittle, a cover softened for flexibility can migrate more plasticiser, and a finish that wipes clean may slide and charge more readily. Choosing a belt is the act of deciding which of those five you can trade, and on a cell line the answer is almost always that shedding is the one you protect first.

One practical consequence of the split. If your plant has both conveyors and drives, keep two specifications and two acceptance sheets. We have seen a maintenance team substitute a drive belt's conductive grade onto a conveyor because the compound code looked similar, and lose a week of output to a debris problem that the substitution created. The timing belts and rubber V-belt range are there for the drive side; this page stays on the carrying side.

08Acceptance, Incoming Inspection and a Line Trial That Proves Something

Most battery plants have a thorough supplier audit and a thin belt acceptance sheet. The audit asks about the quality system, the acceptance sheet asks for a datasheet, and neither proves that the roll in the crate will not mark a cell. Closing that gap is mostly a matter of writing down four things and holding both sides to them.

The first is a controlled specification, revision-numbered, naming the compound, the carcass, the thickness with a tolerance, the finish, the edge treatment, the splice type and the resistance band with its test basis. The second is a sample roll or a cut sample from the production run, not from a laboratory press, tested against that specification and retained. The third is a first-article inspection at the belt maker with photographs and measured values, which is what our quality assurance record contains for every battery order. The fourth is a line trial with a written pass criterion agreed before the belt is installed.

The trial is where most projects go wrong, because the criterion is usually "run it and see". Run it and see tells you nothing when the belt is new and everything looks fine. Set the criterion in advance. A defined number of production hours, a wipe test at the end of each shift on a marked square, a daily resistance reading on the carry face, a metal detector log with zero unexplained trips, and a visual check of the splice step. If any of those fails, the trial fails and the belt comes off. That is a harder conversation at the start and a much easier one at month six.

Incoming inspection should take less than an hour and does not need a laboratory. Measure the width and thickness at three points. Check the edge with a magnifier for exposed fabric. Run a gloved hand along the splice to feel for a step. Wipe a marked square with a white cloth. Read the resistance if you have a meter and the basis says how. Photograph the edge marking so the compound code is on record. If the belt arrived coiled, check for a flat spot or a crease from storage, because a crease that has set will produce a tracking fault that no amount of take-up adjustment will cure.

Commissioning deserves the same discipline. Set tension to the figure on the sheet, mark the take-up, run the belt empty for an hour and check tracking at low speed before any product goes on. Then run product and watch the transfer gaps, where a splice step or a lifted plate will show itself. Our service team attends commissioning on cell lines when the plant asks.

Two cross-references are worth making, because buyers often read several application pages at once. On a feed mill the return side matters because pellets stick and the carryback hardens into an abrasive layer, which is a very different mechanism from anything on a cell line; that story sits in our feed mill belt article. On a tobacco line the belt is stained by aroma oil rather than worn, and the tobacco line guide deals with that. Here the return side matters only because it is where a dropped fragment can lie unnoticed until the next cell passes.

Keep a failure sample. When a belt comes off for any reason other than scheduled replacement, cut a metre from the damaged area, label it with the station and the hours run, and store it. Those samples settle a contamination argument in five minutes, and plants that keep them stop repeating the same specification mistake.

Ask for a belt specification review before your next order

09Frequently Asked Questions

Why does a belt that looks clean still leave marks on the case?

Because the marking usually comes from the compound rather than from visible wear. Plasticiser that has not fully migrated sits in the cover as a film, and every pass transfers a trace of it to the case. A rough slit edge adds fibre to the same film. Neither shows up as a worn patch, which is why a wipe test on a marked square is the check that finds it and a visual inspection is the check that misses it.

Can we use a mechanical lacing instead of an endless belt?

On a cell line we advise against it. The staples are steel, they wear at the hinged points, and when they fail they distribute small metal pieces along the carry run rather than dropping them in one place. If the frame cannot be opened for an endless belt, a pressed splice with no metal parts is the alternative, and the frame modification is usually cheaper than the quarantine that follows a staple failure.

What resistance band should we put in the purchase order?

We cannot give you a number that fits every line, and any supplier who does is guessing. What we can say is that the band has to be agreed with the probe, the voltage, the dwell, the conditioning humidity and temperature, the measured face and the clean or as-run condition all stated. Fix those six items and the number becomes meaningful; leave any of them out and two laboratories will disagree on the same roll.

How do we check shedding before the belt is installed?

Cut a white lint-free cloth, mark a hundred millimetre square on the carry face and rub it with a fixed number of strokes at a fixed pressure. A grey transfer on a belt with no visible wear is a fail. Do the same test after a shift and after a wash-down once the belt is running, because a belt that passes dry can still release compound once it has been softened by a cleaning agent.

Does a thicker cover reduce contamination?

No, and it often makes the mechanical side worse. A thicker cover resists abrasion for longer, but on the small pulleys used at transfer gaps it bends harder, which raises the stress at the splice and at the edge. On a cell line the belt is carrying very little mass, so thickness buys nothing that matters and costs tracking quality. Thin, well-slit and correctly spliced beats thick and stiff.

Can one belt type serve both cell contact and module handling?

It can be made to work, but we do not recommend it. Cell contact demands a low-shedding, chemically quiet cover and a splice with no metal, while module handling rewards a heavier carcass that tolerates mass and impact. A single part number covering both ends of a plant tends to be over-specified where it does not matter and under-specified where it does, so split the specification at the boundary and label each belt by station.