Conveyor Belt Cleaner: Stop Carryback Before It Eats Rollers and Budget

At one plant the same cone of fines sat under the head pulley every morning. Someone on the early shift shovelled it away before the boss showed up, and nobody counted it as a problem. It was part of the morning, like making the coffee. Yet that cone was no act of nature. It was carryback: material the belt had just discharged and then picked straight back up on the return side.

Carryback: the material that keeps coming back

Carryback is the part of the load that does not jump clear at the discharge point but stays stuck to the belt and rides back with it. Along the way it settles wherever nobody wants it: on the return idlers, on the pulleys, in the steelwork and on the floor under the plant. So a conveyor belt has an awkward habit that no brochure likes to mention. It brings back a share of what it has just delivered.

Per revolution it is next to nothing. A gram here, a handful there, hardly worth mentioning. But a belt does not run one lap, it runs day and night, and the little bit per lap grows into a ridge by the week. Three percent sounds harmless. It is the dripping tap of conveyor technology: nothing on its own, but added up over a year it is a line item you notice. One maker of cleaning systems puts the loss from spillage, dust and carryback on a plant without effective cleaning at up to 3 percent of the total throughput. That is manufacturer data, not a verified standard, but it shows the order of magnitude.

That is exactly what makes carryback so treacherous. Early on it hurts nobody. The plant keeps running, the cone under the pulley gets cleared in the morning, and the bill arrives later. Cleaning a conveyor belt is therefore not a matter of going round now and then with a broom under the plant. It means stopping the carryback where it starts, which is at the discharge point.

What unchecked carryback really does

On a long belt at a transhipment operation the return idlers kept failing for months. Everyone blamed poor bearings and dutifully swapped them out. The real cause lay somewhere else. A badly set scraper was letting fines through; they stuck to the return side and loaded onto the idlers lap after lap until the bearings seized. The bearings were not the problem. The belt was coming back dirty. Wet fines on a cold return idler behave like dough on a rolling pin: they cling, they bake on, and they grow with every revolution.

Just how much material that adds up to becomes clear from a sober sum. Around 100 grams of carryback per square metre still counts as an acceptable level; the figure comes from a US mining authority and circulates in the trade by way of a manufacturer. A handful of sand per square metre sounds like nothing. On a belt 1,200 millimetres wide running at 2 metres per second around the clock, it still comes to around 7 tonnes a day that want sweeping up from under the plant. And that is the case that passes for clean.

The consequences do not stop at clean-up time. Carryback raises the friction at the rollers, which kills the bearings sooner, and it likes to build up on one side. Then the belt runs off track, and a cleaning problem turns into mistracking with all the knock-on damage that hangs off it. A German opencast lignite mining company once analysed its idler replacements: around three in ten of the return idlers it changed could be traced back to escaped material, not to normal ageing. Every one of those swaps costs parts, fitter hours and usually downtime. The link between a clean belt and a long-lived idler is no theory. It is written into every maintenance log where someone separates the causes honestly.

One manufacturer reckons carryback is behind around 85 percent of all maintenance problems on conveyor systems. That is an in-house ballpark figure, not an independent study, and I would not trust it to the decimal place. But the direction matches what you see out in the field. Where it is dusty under the belt and the rollers clog up, half the maintenance crew is busy cleaning up after a problem that starts at the discharge point.

Cleaning is a concept, not a single scraper

A plant that handled gypsum had everything the catalogue offers: two scrapers, a brush roller and a water spray bar. And still the carryback coated half the car park in dust. Wet gypsum clings, and in winter the spray bar was out of action anyway because it froze up. The reflex was always the same: we need another scraper. But a second scraper against wet gypsum is like a stronger hairdryer against wet cement. The direction is right, the physics is not.

In the end no extra component fixed it, but a realisation did: you have to dry the belt before cleaning and enclose the cleaning area, with an air knife and extraction. Only then do the scrapers really bite. The dust on the car park was the smaller problem. The bigger one was the air the people beside it were breathing all day.

That is the crux of it. No cleaning system reaches a hundred percent, the trade agrees, because neither the belt surface nor the blade edge is flawless. So you do not solve the problem with one oversized scraper but with a chain of matched steps, each doing its part. Cleaning a belt is system work, not parts-swapping. Once that has sunk in, you stop simply ordering the next blade every time.

The building blocks of a cleaning system

A cleaning system is no black art. It is a handful of building blocks, each with a clear job, and only together do they get the belt clean. In catalogues the bladed parts go by belt cleaner or belt scraper. On a conveyor they always mean the same thing: an element that relieves the belt of what it will not give up on its own. Which design and which blade material fit in detail is a topic of its own.

What decides it is the interplay, not the single part. Two trade portals arrive independently at the same finding: a pre-cleaner on its own does not get the belt clean; only the combination of primary and secondary scraper lifts the cleaning performance above 90 percent. The pre-cleaner takes the bulk, the second the rest, and the return-belt plough makes sure that what the blades up top let through does not stay lying on the return side. When one link in that chain fails, you do not notice it straight away. You notice it at the rollers, a few weeks later.

Moist and sticky material: the hard cases

Dry, coarse gravel jumps clear at the discharge point almost by itself. A decent pre-cleaner often does the job there, and the belt comes back clean. The hard cases are the others: wet fines, sticky gypsum, material that freezes in winter. Here a single scraper fails not because it is a bad scraper but because the blade is fighting moisture. Wet fines smear their way under the edge instead of being scraped off.

That is why the real cleaning performance in daily service often sits well below the brochure value. On plants with average to poor maintenance a cleaning system frequently reaches only 40 to 60 percent, according to one manufacturer. On paper the design is listed at over 90 percent. In service, with no re-tensioning and no enclosure, half of that arrives. The gap between the two numbers is not a question of material but of upkeep and surroundings.

For the hard cases, then, what the table above lists comes together: a brush against the sticky fines, water or air against the rest, and an enclosure so that neither dust nor moisture flies off uncontrolled. Run a spray bar in a freezing shed and you have to heat it or clean quite differently in winter; otherwise it will stand idle on exactly the days you need it. And one special case in passing: profiled belts such as chevron or sidewall belts cannot be scraped clean with a straight blade; they need a cleaning concept of their own. That would take us too far here, but it belongs on the list if your material runs over cleats.

What better cleaning really saves

Now to the sums, because this is where the tiresome cone under the pulley turns into a line item worth caring about. One manufacturer worked out, for a plant with 40 conveyor belts in coal handling, what a small difference in cleaning performance amounts to. By this calculation, just 2 percentage points less cleaning means around 10,876 tonnes of extra material loss a year. At a 10 percentage point difference the loss grows to over 54,000 tonnes. That is a manufacturer’s model calculation, not a measured field figure, and it holds for a large plant. But it makes tangible that the second decimal place of cleaning performance is money, not a detail for the brochure.

It gets more concrete in the before and after at a single plant. The same manufacturer gives these figures for one works after it improved its scraper cleaning. It is a manufacturer’s example from one plant, not an industry norm, but a picture every maintenance hand recognises.

Skimp on the cleaning system here and you pay for it at the rollers and the belt, usually many times over. Yet that is the exception, not the rule. One manufacturer reckons only about a quarter of all conveyor belts have a scraper at all, and of those only about a quarter again are properly maintained. On the maths, that leaves roughly six belts in a hundred running with a well-maintained cleaning system. All the rest supply the material for the figures in this section.

One more warning about the euro figures you find online. For a works in India, one manufacturer puts a 2 percent difference in efficiency at around 7 million rupees of extra cost a year, and itself points out that labour costs in Europe run to roughly forty times that. You cannot simply convert that to a plant over here, and none of these sources names a sound euro figure for the German-speaking market. So you will find none here. What your carryback costs you hangs on your rollers, your belt and your wages. Ask your own operation, not a brochure from another country.

From diagnosis to fix: where to start

Before you rebuild a cleaning system, watch it run. What follows is field experience, not a standard. Stand at the discharge point and watch what the belt still carries behind the pulley. Does it come back almost clean, or is there a film of fines on the return side? Then walk the length of the belt and note where the material drops off. Where the floor is buried thickest, there sits your problem, not where you happen to look first.

From that observation the order almost picks itself. Cause first, symptom second. If the belt brings a lot back, check the pre-cleaner and whether it still sits against the belt at all, before you think about a third building block. If the material clings, the question is not how many blades you stack but whether you have moisture, brush and enclosure under control. And if the rollers are dying, look for the link to the discharge point instead of endlessly swapping bearings.

None of this needs special tools, only an honest look under the belt and the willingness to chase the cause instead of the cone. A clean belt at the head is the cheapest life insurance for the expensive parts behind it. The plant tells you fairly precisely what is going on, once you stop and listen to it, instead of just shovelling away the ridge every morning.