CEMA Idlers Explained: B, C, D and E Classes and What They Govern

When a mining or cement project sends an idler enquiry, the first line that decides the order is usually a two-word label. The buyer writes CEMA C beside the roller assembly, and that short code outranks every price discussion that follows it. It is not a brand and it is not a size. It is a design envelope that states how much load the roller is built to carry, how long it is expected to survive in service, and how its bearings, shaft, shell and seals have been assembled to reach that target. Read the class correctly and the conveyor runs quietly for years. Read it wrongly and the same conveyor eats bearings inside a single season.

CEMA idlers sit inside a grading system published by the Conveyor Equipment Manufacturers Association, and the letters B, C, D and E are the four rungs of that ladder. Four letters, four duty envelopes, four different bills of materials that look almost identical on a quotation sheet. The difficulty is that the label alone rarely tells you which rung you are actually buying, because every supplier describes its own product families with the same four letters and quietly defines the boundaries in its own catalogue.

This page deals with the standard and the grading logic only. It does not repeat general roller selection advice, it does not repeat incoming inspection routines, and it does not re-open the dust seal argument, because those subjects already have their own treatment elsewhere on this site. What follows is the class system itself, the way the American and European conventions diverge, how a real site condition is mapped onto a class, and how to write a specification that leaves a supplier no room to reinterpret.

A class is a promise about service life, not a catalogue grouping.

Send us your conveyor duty and we will map it to the correct idler class

01What the Class Label on an Idler Drawing Really Governs

The letters do not describe a single component. They describe a chain of decisions that a serious manufacturer makes together, because changing one link without the others produces a roller that is over-built in the wrong place and under-built where it matters. A heavier class moves the bearing to a larger bore, thickens the shaft, increases shell wall thickness, upgrades the seal arrangement from a simple labyrinth to a multi-stage labyrinth with a grease purge path, and shortens the permitted spacing between idlers under the belt. The class is therefore a roof over several sub-decisions, and the roof only stays up when they are raised together. Raise the bearing alone and you have simply moved the failure point.

That is why a quotation that names a class but never states bearing bore, shaft diameter, wall thickness or seal type is not a specification at all. It is a word. I have watched a Chilean copper port reject an entire shipment for exactly this reason: the purchase order said CEMA D, the delivered roller carried a bearing and shaft sized to a lighter family, and there was no line on the order that either party could point to as breached until the first failures appeared at eleven months. The class label had done no work at all.

Why the class travels with the duty and not with the part number

A part number is stable. A duty is not. The same troughing position on the same conveyor can start life as a moderate-duty installation and become a heavy-duty one simply because the operator raised the tonnage, added a second shift, or changed the feed so that the loading point now drops material with far more impact energy than it did during commissioning. Because the physical roller still looks the same, nobody re-opens the class decision, and the installation quietly operates above the envelope it was bought for.

Understanding what the label constrains is the whole point of the exercise, so the table below lays out the dimensions that move when a class moves, and the consequence of leaving any one of them behind.

The list matters more than any single line on it, because buyers routinely negotiate the cheapest visible item and leave the invisible ones to the supplier. That is how a project can honestly order a heavy class and still receive a light roller.

02The Four Classes, B through E, and the Envelope Behind Each Letter

Class B, C, D and E form an ascending sequence of load and life capability, and the sequence is deliberately not linear in cost. Each step up buys a disproportionate amount of bearing life in a dirty environment, which is why operators who start on a lighter class and "just replace them more often" usually spend more over a decade than the buyer who specified the heavier class on day one. The letters are graded against the service the station is expected to give under continuous operation, not against a single peak load.

Class B occupies the light end and suits well-behaved duties where the material is small, the belt is modest in width and tension, and the installation runs intermittently. Class C is the workhorse of general bulk handling and covers the majority of quarry, aggregate and general industrial conveyors where dust is present but manageable and the belt runs most of the working day. Class D moves into continuous heavy duty, the territory of primary crushing plants, high-tonnage overland lines and ports where the conveyor is the process and cannot be taken out of service on a whim. Class E sits at the top of the published ladder and is chosen where the combination of load, width, speed and uninterrupted operation leaves no margin, and where the cost of a stopped conveyor dwarfs the cost of the roller.

Between the letters there is no universal numeric line, and that is the first trap for a buyer who wants to compare quotations. The standard defines the classes and the relationships inside each one, but the exact load and life figures attached to a class are carried by the manufacturer's published tables and by the standard publication itself. A responsible spec therefore never says "class C, one hundred kilos" and expects that number to mean the same thing in two different catalogues. It says class C and then pins the design down with the mechanical detail that follows from it.

Reading the letters as a family rather than four separate products

Manufacturers who build the full range usually design the family so that bearings, seals, shafts and end caps are shared upward as far as practical. That is good for spares, but it also means the catalogue photographs look nearly identical across the range, and a buyer comparing two websites can easily mistake a light roller for a heavy one. The differences live in the numbers printed in the tables and in the cutaway drawing, never in the photograph.

03Bearings and Shaft Diameters as the Mechanical Backbone of a Class

If you want to know which class a roller really belongs to, ignore the label and measure the bearing seat and the shaft. Those two dimensions are the physical signature of the class, and nearly every other feature of the assembly is arranged around them. The bearing sets the load the station can carry and the life it will return under that load. The shaft determines whether the bearing actually receives that load in a clean, square, unstressed way, or whether it receives it through a bending moment that steadily destroys the seal and then the raceway. A family that pairs a large bearing with an unchanged shaft is not a heavier class at all. It is a heavier bearing on a light roller.

On a 1,400 t/h coal overland line we once replaced a batch of rollers that had been ordered to a heavy class and failed inside a year. The bearings measured correctly for the class. The shafts did not. Every failed station showed the same footprint, a polished ring where the seal lip had been deflected by shaft flex. The buyer had specified the class and the bearing, and had said nothing about the shaft, so the manufacturer filled the gap with the light shaft that was already in the machine shop. That is a specification failure, not a manufacturing failure.

Class also governs the fit between the shaft and the bearing seat, the concentricity of the end cap, and the runout of the tube. Heavier classes hold these tolerances more tightly because they are expected to run at higher belt tension and for longer uninterrupted periods, and because the cost of an unplanned stop at a port is measured in demurrage rather than in parts. A light class running at low tension can absorb a degree of runout that would be unacceptable in a heavy class under continuous load.

Why the drive-side and the loading zone always punish a marginal class first

Load on an idler is never uniform along the conveyor, and the class system assumes the designer knows where the peaks sit. The loading zone receives impact that a return station never sees, and the drive-side trough carries the highest tension and therefore the highest radial load. When a class is marginal, it is always these positions that fail first, and operators often read the pattern as a bad batch rather than as a class that was never adequate for the duty. If you are investigating repeated failures, the three articles on impact rollers at the loading zone, self-aligning roller behaviour and bearing and seal pairing take the diagnosis further than this page needs to go.

A wise buyer treats these four lines as non-negotiable, and treats everything else on the quotation as commercial. It is also worth telling your conveyor belt manufacturer which positions on the line carry the peaks, because a supplier that understands the duty will often advise a mixed-class belt line rather than one uniform class everywhere. Many conveyor belt supplier catalogues hide exactly this advice behind a single price per station, and the buyer who asks for the station-by-station view usually discovers that the loading zone wanted a heavier family all along.

04The Roller Shell, the Wall Section and the Seal Envelope

The shell of a roller looks like the simplest part of the assembly and is often the least specified. It is simply a tube with two machined ends, yet its wall thickness, its straightness and its balance determine whether the bearing ever sees the clean, steady load the catalogue promised. A thin wall deflects under belt pressure, and once it deflects, the outer surface is no longer a rotating cylinder but a slightly flattened shape that rubs rather than rolls. Belt cover wear follows directly from that, long before anyone suspects the roller.

The tube material matters as much as the section. Standard carbon steel is the default for dry, neutral duties. Where the material is wet, acidic or salt-laden, the shell corrodes from the outside and the resulting pitting acts like a lap that polishes the belt. A heavier class in a corrosive duty with a plain steel shell is still the wrong roller, which is why class and material must be specified together and never treated as alternatives to one another.

Sealing is graded with the class because sealing is what allows the bearing rating to be realised in a real plant. A single-stage labyrinth suits a dry, light duty. A multi-stage labyrinth with a long grease path suits the dusty, continuous duties that the heavier classes exist for. Nobody should read this as a recommendation to buy the heaviest seal available, because an oversized seal arrangement on a light roller wastes money and can even increase drag. The point is alignment: the seal family should match the class, and both should match the dust load and moisture the station will actually meet.

How shell and seal decisions escalate from the light end to the heavy end

For a deeper treatment of the sealing detail itself, including the labyrinth geometry and the ways dust defeats a well-made seal, the dedicated article on labyrinth seal design is the right place to continue. What matters here is only that the seal family is a function of class, and that a quotation which names a heavy class while hiding a light seal is internally inconsistent.

05Service Life Targets and What a Class Actually Promises

A class is ultimately a statement about bearing life under a stated load, and bearing life is where the whole grading system earns its keep. Any bearing, given a load, a speed and a clean environment, will return a calculable life. The class system exists because the environment is never clean in a real plant, and because the load is never uniform. It bundles the bearing selection with the protection that lets the bearing approach its theoretical life in the field, so that a heavy class is not merely a bigger bearing but a bigger bearing plus the conditions it needs to last.

That is why two rollers carrying the same bearing can have wildly different service lives on the same conveyor, and why the operators of long overland lines keep records of failure position rather than failure count. A return station in dry, clean air and a troughing station under a wet loading point are two different worlds even though they sit forty metres apart. I have seen a single heavy-class trough position fail four times in three years while identical neighbours on the same span, ordered in the same batch with the same label, ran untouched, and the whole difference traced to a dribbling chute that kept one station permanently damp. The class was right. The location was wrong for that class.

Life expectation should therefore never be quoted as a single number for the whole conveyor. It should be quoted per duty zone, and the class should be chosen so that the zone with the worst combination of load, dust and moisture still clears the maintenance interval the operator can actually staff. A plant that can only reach a station twice a year needs a class that survives two years of neglect, not one that performs well under monthly greasing.

Telling a life problem from a class problem

When rollers fail early, buyers usually blame the supplier and suppliers usually blame the site, and neither side examines the specification. The honest test is whether the failures correlate with position or with time. Failures that cluster at specific positions point to a duty that outruns the chosen class. Failures that arrive in a wave across the whole line, all within a narrow band of operating hours, point to a batch, a bearing source or a grease that was never suited to the duty. Separating the two patterns costs nothing and usually settles the argument before it reaches the accounts department.

It also helps to know what the standard does and does not fix. CEMA classes describe a family of idler designs with graded load and life capability, and the publication sets out the relationships that define each class and the way a manufacturer should present them. The absolute load and life figures belong to the standard publication and to each manufacturer's table, and they are the numbers you must ask for rather than assume. Any quotation that offers a class without offering those tables, or without a bearing designation that can be traced, is asking you to take the promise on trust.

One more practical point, and it is the one most often skipped at the enquiry stage. The conveyor's belt itself sets part of the load the roller must carry, so an idler decision is never fully separable from the belt decision. A heavier, thicker rubber conveyor belt running at high tension puts more load on every station than a lighter belt on the same route, and a change of belt specification mid-life can push a marginal class over the edge. Buyers who treat the belt and the rollers as two unrelated purchases are the ones who discover this the expensive way.

06CEMA Versus DIN and ISO: Why the Two Families Do Not Convert

European and international practice approaches the same engineering problem from a different direction. Where the American convention groups idlers into lettered load classes that a buyer can order by name, the DIN and ISO traditions tend to describe dimensions, fits and test methods directly, and to leave the grouping to national or company practice. The result is a genuine mismatch of vocabulary, not a difference of quality, and treating one as a translation of the other is one of the most common and most expensive mistakes in idler procurement.

This is not an argument that one system is better. It is an argument that the two are not interchangeable, and that a specification must name the system it is written in. The mistake appears most often in international projects where an engineering office on one continent issues a specification, a fabricator on another quotes against it, and a third party inspects against a different document again. Everyone believes they are following the same standard because the words look familiar. Then the delivered roller has the wrong shaft end, or the wrong bearing seat tolerance, and the first physical proof of the mismatch arrives when the brackets will not line up.

Because a heavier industrial conveyor belt line can be assembled from components sourced on both sides of that divide, the mixing risk is real and routine. The practical defence is simple. State the system, state the edition or year of the document you are referencing, and require the supplier to declare any dimension it has taken from a different convention. That one sentence in the purchase order prevents most of the confusion.

Independent writers on belt and roller standards also warn against quoting a class as if it were a global index of quality. A conveyor belt factory that builds to both conventions will usually keep two sets of drawings for what the catalogue calls the same roller, and a buyer who requests one set while inspecting against the other will reject a perfectly good product.

Power transmission components belong to a third conversation entirely, and it is worth separating them clearly at the enquiry stage. The engineers who specify a transmission belt manufacturer for the drive, or consult a V-belt manufacturer about the motor end, are asking a different question from the one this page answers, and a quotation that bundles everything into one price is usually hiding a shortcut somewhere inside it. If you need several duty zones priced together, a wholesale conveyor belts schedule with the roller classes broken out line by line will tell you far more than a lump sum, and a conveyor belt distributor who handles both belts and components can usually say which class each zone should carry. That is the practical value of writing the standard down properly. It lets a buyer split the order across two or three suppliers without losing control of which class each duty zone was designed around.

07Mapping a Real Site Condition to a Class Before the Spec Is Written

Class selection should be a deliberate step that happens before the specification is drafted, not a line copied from an old purchase order. The inputs are the tonnage the line must carry, the belt width and speed, the nature of the material, the presence of dust and moisture, and whether the conveyor runs continuously or intermittently. Those five inputs, read together, will point to a class with reasonable confidence, and it is normal for one conveyor to carry two or three classes across its length.

Tonnes per hour alone is a poor guide. A modest tonnage of fine dry material can be gentler on a roller than half the tonnage of damp, heavy ore dropped from height, because the second duty loads the station through impact and keeps the seal wet. Belt speed matters too, since a higher speed raises the cycles the bearing sees per hour and therefore consumes life faster for the same radial load. Dust and moisture decide how much of the bearing's theoretical life is actually reachable. The continuous or intermittent nature of the duty decides how much accumulated damage the station can absorb between maintenance windows.

My habit, after too many remedial projects, is to write the duty words down on one page before any class letter appears. A line described as continuous, twenty-four hours, damp material, with a dribbling transfer point, tells you more about the class than a tonnage figure ever will. Only after that page exists does the letter get chosen, and only then does the specification get written.

Notice what the table does not contain, which is a single hard number. That is not an oversight. The correct load and life figures come from the standard publication and from the supplier's own tables, and any writer or salesperson who hands you a fixed figure for a class without those references is guessing on your behalf. Ask for the table. Then check that the roller you receive was built from the same table.

08Writing the Specification and Accepting What Arrives

A specification that names only a class is an invitation to interpret. The document that works is short, boring and precise, and it settles the argument before the rollers are even made. Four things have to be inside it: the standard system the class is drawn from, the mechanical detail that the class implies, the environment the roller must survive, and the evidence the supplier must supply to prove the roller matches. Everything else is commercial negotiation and does not belong on the technical sheet.

The environment line is the one buyers most often leave out, and it is the one that ends up in the failure report. Dust load, moisture, temperature range, washdown chemicals and whether the station sits indoors or on an open stockyard all change the sealing and coating that should accompany a given class. A supplier who is not told about the damp transfer point will supply a roller that meets the class and fails at the load-out.

The lines that belong on every idler specification

If your team would rather work from a ready-made list of questions, a good idler factory RFQ checklist covers the commercial side of the same exercise, and the detailed specification breakdown in the mining roller specification guide is worth reading alongside this page. Buyers who want the broader selection picture rather than the standard itself should start with the complete roller guide, and those at the ordering stage will find the practical buyer notes before ordering useful. All four stay at the general level. This article stays at the standard level, and the difference is deliberate.

The misuses that cost the most money

Almost every idler complaint I have investigated came down to one of a small number of mistakes, and none of them were exotic. They were ordinary commercial shortcuts taken by capable people under time pressure. The table below sets out the pattern, how it reveals itself and what it costs.

Nothing in that table is about the standard being complicated. It is about a one-line order carrying more weight than it can bear. Write the class, then write the detail behind it, and the standard does exactly what it was designed to do.

Two related points before the questions. First, the station type is a separate decision from the class, and a line will normally use several: a troughing station on the carrying run, a return station underneath, and a heavier impact station at the loading point, with training stations at intervals to hold the belt. Second, the class applies to the roller assembly and not to the conveyor pulleys that drive and tension the belt, which are selected from a different set of calculations even though they share the same frame. A specification that confuses the two will be wrong at both ends.

One closing boundary is worth setting, because it catches people out at the ordering stage. Whether the rollers are inspected at the component level or against a full quality assurance programme, the test is the same, and it is whether the delivered roller matches the table it was ordered from. The class system reaches only as far as the roller assembly itself, so the pulleys that drive and tension the belt sit outside it, and they are chosen from their own calculations even when they mount on the same frame as the idler stations.

Ask for our class table before you order

09Frequently Asked Questions

Do the letters B, C, D and E describe the bearing size only?

No. The letter describes a whole design envelope that moves together. A higher class does bring a larger bearing bore, but it also brings a stiffer shaft, a thicker shell section, tighter tolerances and a more elaborate seal arrangement, and it usually tightens the spacing permitted between stations. Reading the letter as a bearing size alone is how buyers end up with a large bearing on a light shaft, which simply relocates the failure from the raceway to the seal lip.

How do I know which class the conveyor actually needs?

Work from the duty, not from the last purchase order. The inputs are the tonnage the line must carry, the belt width and speed, the character of the material, the dust and moisture level, and whether the conveyor runs continuously or intermittently. Read them together, because a heavy, damp, impact-loaded duty can be harder on a roller than a larger tonnage of dry fines. Most conveyors end up using more than one class across their length, with the loading zone and the drive side carrying the heaviest.

Can I replace a class roller with a DIN or ISO equivalent?

Not by translation, and that is the trap. The conventions approach the problem differently. The American practice groups designs into lettered classes that a buyer can order by name, while the European and international practice tends to set out dimensions, fits and test methods directly and leaves the grouping to national or company practice. There is no reliable one-to-one mapping, and shaft ends, spigot diameters and bracket interfaces often differ. If you must change system, name it, name the edition, and have any deviation approved in writing.

Does a heavier class always last longer on my line?

It lasts longer where the load and the environment justify it, and it can be wasted money everywhere else. A heavier class in a clean, lightly loaded, single-shift duty buys life that the station will never consume. Worse, an oversized seal arrangement on a light roller adds drag for no benefit. The right class is the one whose life envelope clears the maintenance interval the plant can realistically staff, and no more.

What must be on the purchase order besides the class letter?

The system the class comes from and the document edition, the bearing designation and bore, the shaft diameter at the seat and the spigot, the tube material and wall thickness, the seal arrangement and whether the roller is sealed for life or relubricated, the duty and environment the roller must survive, and the documentation package that must accompany the first article. A purchase order that carries only a letter and a quantity has left every decision that matters to the supplier.

Is it acceptable to mix classes on one conveyor?

It is normal, and it is often the correct answer. The loading zone and the drive-side trough carry the highest loads and should be ordered heaviest, the general carrying run usually sits one step lower, and the return side can be lighter still where the duty is clean. What is not acceptable is mixing classes without telling the supplier where each one goes, or switching class mid-project without recording it, because that is how a correct-looking set of rollers ends up fitted in the wrong positions.