A belt conveyor is a continuous bulk-handling machine that carries loose material on a flexible rubber-and-fabric belt running between two pulleys, supported by troughed idlers along its length. It is the backbone of every modern mining operation, moving more tonnes per dollar than any other ground-transport method. The drive pulley pulls the belt, friction transfers motion to the bulk load, and the troughed idlers shape the belt into a U so material does not spill. A single overland conveyor like the Curragh North in Australia moves over 8,800 tonnes per hour across more than 20 km without a single transfer point.
How the Belt Conveyer Works
A belt conveyor works on a simple principle — friction between the drive pulley and the belt drags the belt forward, and the belt drags the load. The drive pulley sits at the head end (the discharge end), wrapped with a high-friction rubber lagging, usually diamond-grooved or ceramic-tiled. The tail pulley sits at the loading end. Between them, troughed idlers — typically three rollers set at a 35° trough angle — cradle the belt into a U-shape so loose material like coal, iron ore, or crushed stone rides in the centre without spilling over the edge. The return run sits below, supported by flat return idlers spaced roughly every 3 m.
Belt tension is everything. Too little and the belt slips on the drive pulley — you'll see scorched lagging and a smell of burning rubber within minutes. Too much and you crush the idler bearings, fatigue the belt splice, and warp the pulley shafts. That's why every conveyor over a few metres long has a take-up unit — usually a gravity tower with a counterweight or a screw take-up on shorter runs — that holds tension constant as the belt stretches with load and temperature. A typical EP (polyester-nylon) belt stretches 1-2% under working load; steel-cord belts on overland conveyors stretch under 0.25%, which is why they dominate runs over 2 km.
Failure modes are predictable. Mistracking — the belt drifting to one side — is the number one issue, caused by off-square idler frames, uneven loading, or material build-up on the tail pulley. If you notice the belt edge wearing on the structure, check your idler squareness with a string line before you blame the belt. Splice failure shows up as a sudden bang and a belt on the floor; nearly always traceable to a vulcanised splice cured below 145°C or a mechanical fastener overloaded beyond its rated tension.
Key Components
- Drive Pulley: The powered pulley that transfers torque to the belt through friction. Diameter typically 400-1600 mm depending on belt class, lagged with 12-15 mm of rubber or ceramic tile to raise the friction coefficient to around 0.35-0.45. Drive pulley diameter must match belt class — undersizing causes carcass fatigue at the bend point.
- Tail Pulley: The non-driven pulley at the loading end that returns the belt. Often fitted with a self-cleaning wing design to shed material that drops onto the return run. Shaft must be parallel to the drive pulley within 1 mm per metre of face width, or the belt mistracks immediately.
- Troughed Idlers: Three-roller assemblies that shape the belt into a U-trough, usually at 35° but 45° on high-capacity overland systems. Roller spacing is 1.0-1.5 m on the carry side. Bearings are sealed double-row ball or tapered roller, rated for L10 life of 60,000 hours under nominal load.
- Take-Up Unit: Maintains constant belt tension as the belt stretches and contracts. Gravity take-ups use a counterweight tower sized to roughly 1.5-2× the slack-side tension; screw take-ups suit conveyors under 60 m. Travel must be at least 1.5% of belt length to absorb thermal and load stretch.
- Conveyor Belt: The carrying medium itself. Multi-ply EP belts handle most underground and aggregate work up to ST1000 strength class. Steel-cord belts (ST2000-ST10000) run overland conveyors carrying 10,000+ tonnes per hour. Belt cover thickness is sized to material abrasiveness — 6 mm top cover for coal, 12-18 mm for sharp iron ore.
- Skirt Boards and Loading Chute: Contain material at the loading point until it settles to belt speed. Skirt rubber must clear the belt by 3-5 mm — drag it on the belt and you'll groove the cover within a week.
- Belt Cleaners (Scrapers): Primary scraper sits against the drive pulley to remove carryback; secondary scraper cleans residual fines further along the return. Tungsten-carbide tipped blades last 6-12 months on coal, 2-4 months on hard rock.
Real-World Applications of the Belt Conveyer
Belt conveyors dominate bulk handling because no other system matches their tonne-per-kilowatt-hour efficiency over distance. They handle everything from sticky tar sands to abrasive iron ore to fragile grain. The reason mining engineers reach for a belt over a truck fleet, a pipeline, or a rail loop comes down to three numbers — capital cost per tonne-km, energy per tonne-km, and labour per tonne-km. A belt conveyor wins on all three above roughly 2 km of haul distance and 500 tonnes per hour throughput. Below that, trucks usually still win on flexibility.
- Coal Mining: The Curragh North overland conveyor in Queensland, Australia, runs 20+ km from pit to wash plant at 8,800 tonnes per hour using a steel-cord ST3150 belt.
- Iron Ore: Rio Tinto's Brockman 4 mine in the Pilbara uses a network of belt conveyors feeding the primary crusher to the train loadout at over 6,000 tonnes per hour.
- Hard Rock Mining: Underground belt conveyors at LKAB's Kiruna iron mine in Sweden lift ore 1,365 m to surface using a series of inclined belts with intermediate transfer points.
- Quarry and Aggregate: Stationary belt conveyors at Lafarge limestone quarries feed jaw crushers, secondary cone crushers, and stockpile radial stackers — the Telestack TC-421 radial telescopic stacker is a common pick for variable stockpile management.
- Cement Production: Long-distance belt conveyors at Holcim's Untervaz plant in Switzerland move limestone from quarry to plant across mountainous terrain, replacing a truck haul that ran 24 hours a day.
- Port Handling: Ship-loading belt conveyors at the Port of Hedland in Australia load Capesize bulk carriers with iron ore at rates over 10,000 tonnes per hour using shuttle and travelling boom designs.
- Potash and Salt: Nutrien's Rocanville potash mine in Saskatchewan uses underground belts to convey ore from the mining face to the production shaft at 3,500 tonnes per hour.
The Formula Behind the Belt Conveyer
The single most important calculation in belt conveyor design is the volumetric carrying capacity — how many cubic metres per hour the belt can move at a given speed and trough geometry. Get this right and the conveyor matches the upstream feed rate without spillage or starvation. At the low end of the typical belt-speed range (around 1.5 m/s) the belt runs quietly with minimal dust, but capacity is constrained — fine for short underground sections. At the nominal range (2.5-3.5 m/s) you hit the sweet spot for most mining applications. Push past 5 m/s on overland conveyors and capacity climbs, but idler bearing life drops sharply and dust generation becomes a serious problem.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Q | Mass flow rate (carrying capacity) | tonnes/hour | tons/hour |
| A | Cross-sectional area of material on the belt at the trough angle | m<sup>2</sup> | ft<sup>2</sup> |
| v | Belt speed | m/s | ft/min |
| ρ | Bulk density of conveyed material | tonnes/m<sup>3</sup> | lb/ft<sup>3</sup> |
| k | Inclination factor (1.0 on flat, drops with belt angle above 0°) | dimensionless | dimensionless |
Worked Example: Belt Conveyer in an iron ore mine overland conveyor
An iron ore operation in northern Brazil is sizing an overland belt conveyor to move 6,000 tonnes per hour of crushed hematite from the secondary crusher discharge to the rail loadout 3.4 km away. Material has a bulk density of 2.4 tonnes/m³ and a surcharge angle of 20°. The route is essentially flat (k = 1.0). The engineering team has chosen a 1600 mm wide belt with a 35° three-roll trough. They need to verify what belt speed delivers the required throughput.
Given
- Qrequired = 6000 tonnes/hour
- Belt width = 1600 mm
- Trough angle = 35 degrees
- Surcharge angle = 20 degrees
- ρ = 2.4 tonnes/m<sup>3</sup>
- k = 1.0 dimensionless
Solution
Step 1 — calculate the cross-sectional area of material on a 1600 mm belt at 35° trough with 20° surcharge. From CEMA tables, the standard edge distance leaves about 1450 mm of effective contact width, and the resulting cross-section is approximately 0.27 m²:
Step 2 — solve for the required belt speed at the nominal design point of 6,000 tonnes/hour:
Step 3 — check the low end of the typical operating range. If the upstream crusher de-rates to 3,500 tonnes/hour during liner changeouts, what belt speed do you need?
At 1.5 m/s the belt runs quietly, generates almost no dust, and the load profile sits stable in the trough. This is the kind of speed you'd see on a short underground gathering belt. Step 4 — check the high end. The owner is asking whether the same conveyor can be future-rated to 9,000 tonnes/hour:
3.86 m/s is feasible but you're entering territory where idler bearing L10 life drops to roughly 35,000-40,000 hours, transfer-chute wear accelerates, and dust suppression becomes a serious operating cost. Most overland iron ore conveyors live happily between 4.0 and 5.5 m/s, but only with upgraded idler series and ceramic-tile lagging on the drive pulley.
Result
The nominal design speed is 2. 57 m/s to deliver 6,000 tonnes per hour on a 1600 mm belt at 35° trough — well inside the comfort zone for steel-cord overland conveyors. At the low-end de-rate of 1.5 m/s the belt is loafing along; at the future-rated 3.86 m/s you're working the bearings and lagging hard, but it's still within accepted CEMA practice. If commissioning measurements show throughput 15-20% below the predicted 6,000 t/h at 2.57 m/s, the usual suspects are: (1) loading chute geometry forcing material to skid on the belt for several metres before reaching belt speed, which reduces effective fill cross-section; (2) belt tracking off-centre, so the load profile rides asymmetrically and spills before reaching the rated A; or (3) actual bulk density running 10-12% below the lab figure because the secondary crusher is producing finer material than spec, lowering ρ and Q proportionally. Always weigh a sample at the discharge with a belt scale before blaming the design.
Belt Conveyer vs Alternatives
Choosing between a belt conveyor, a haul truck fleet, and a rail loop comes down to distance, throughput, and terrain. Each option wins in a specific zone of the operating space. Here's how they compare on the dimensions that actually drive the capital and operating decision.
| Property | Belt Conveyor | Haul Truck Fleet | Rail / Train Haulage |
|---|---|---|---|
| Throughput capacity | 500-15,000+ tonnes/hour continuous | Limited by truck count; typically 200-2,000 t/h per route | 5,000-30,000 t/h at peak loadout |
| Energy per tonne-km | 0.5-1.5 kWh/tonne-km — best in class | 8-15 kWh/tonne-km equivalent (diesel) | 0.3-0.6 kWh/tonne-km on rail |
| Capital cost (per km installed) | $2-8 million for overland; high upfront | Low fixed cost; cost lives in fleet purchase | $5-20 million per km plus rolling stock |
| Maintenance interval | Idler replacement every 30,000-60,000 hours; belt 8-15 years | Engine, tyres, and brakes on rolling 500-2,000 hour cycles | Rail grinding annual; locomotive overhaul every 8-12 years |
| Flexibility / route changes | Fixed installation; relocation is a project | Highly flexible — change route in a shift | Very rigid; new spurs are major capital |
| Best application zone | Distances above 2 km, throughput above 500 t/h, fixed source-to-destination | Short hauls, variable destinations, ramp-up phases | Long-haul (50+ km) port-bound bulk export |
| Labour intensity | Very low — 1-2 operators monitor several km of belt | High — one operator per truck, plus support | Moderate — small crew operates large train |
| Slope capability | Up to 18° for smooth belt; 30° with cleated/sidewall | Limited to about 10% grade for loaded haul | Limited to about 2.5% grade for heavy haul rail |
Frequently Asked Questions About Belt Conveyer
If you've checked idler squareness with a string line and the belt still drifts, the next two suspects are off-centre loading and pulley taper. Loading off-centre by even 50 mm at the chute will pull the belt to that side because the load itself shifts the centre of mass — fix it at the chute, not with self-aligning idlers downstream.
Beyond that, drum and tail pulleys aren't always perfectly cylindrical out of the box; even 0.5 mm of crown taper or shaft non-parallelism over a 1600 mm face will steer a belt visibly. Put a precision level across both pulley shafts and check parallelism with a piano wire pulled along the conveyor centreline. Carryback build-up on the tail pulley behaves like a soft crown and produces the same symptom — clean it before you blame the structure.
The cleanest decision rule is conveyor length and required tension class. Below about 1,000 m centre-to-centre and tension class ST1000 or lower, multi-ply EP (polyester-nylon) belts are cheaper, easier to splice in the field with mechanical fasteners, and forgiving of impact. Above that, the elastic stretch of EP belts forces a take-up travel that becomes physically impractical — a 2 km EP belt stretching 1.5% needs 30 m of take-up travel.
Steel-cord belts stretch under 0.25%, so the take-up tower stays manageable on multi-kilometre runs. The trade is upfront cost (roughly 2-3× per metre) and the need for a heated vulcanising press on site for splicing — you cannot mechanically fasten steel-cord. If your site is remote and you have no splice crew, that alone can swing the decision back to EP.
The inclination factor k in CEMA tables assumes the material rides stably in the trough at the rated surcharge angle. On an incline above roughly 12°, two things happen that the textbook k doesn't capture cleanly: the surcharge angle effectively reduces because the material wants to slide back, and at the loading point material rolls backward on the belt before reaching belt speed, lowering the effective cross-section.
The fix is usually at the loading geometry, not the belt. Reposition the chute so material lands moving in the conveying direction at close to belt speed. If you can't, you need cleated belt or sidewall belt — at which point the formula changes entirely and you size from the manufacturer's load-pocket capacity, not the troughed cross-section.
CEMA and DIN both publish minimum drive pulley diameter as a function of belt tension class — for example, an ST2000 steel-cord belt requires a minimum 1,000 mm drive pulley, and going below that fatigues the steel cords at the bend point. The symptom of undersizing isn't immediate — it shows up 2-4 years in as cord fractures starting near the splice, then propagating until you get a sudden longitudinal rip.
If you inherit a system and want to check, measure the pulley diameter and compare to the belt's stamped class. A common mistake is matching pulley diameter to belt width — the two are unrelated. A 1400 mm wide ST500 belt only needs a 630 mm pulley; a 1000 mm wide ST3150 belt needs a 1,250 mm pulley.
The most common cause is idler rolling resistance running higher than the calculation assumed. CEMA and DIN 22101 use a base friction coefficient of around 0.020-0.025 for clean, well-aligned idlers in good condition. In real installations with dirty idlers, frozen bearings on the return run, or build-up on rollers, the effective coefficient can climb to 0.035-0.045 — and that's a linear factor in the total power draw.
Walk the return run with an infrared thermometer. Any idler reading more than 15°C above ambient is dragging — replace it. On a 3 km conveyor with 1,000 idlers, even 30 stalled return rollers will measurably move the power draw. Other contributors: skirt rubber dragging on the belt, belt cleaner blade pressure set too high, and a take-up tension running above the design slack-side value.
Ceramic-tile lagging makes sense when you're operating in wet or sticky conditions where rubber lagging glazes over and loses friction coefficient. On a coal handling system that runs through monsoon weather or a potash conveyor where dust turns to brine, rubber lagging can drop from 0.35 to under 0.20 friction coefficient when wetted — at which point the belt slips, the lagging glazes hotter, and you cascade into a belt fire risk.
Ceramic tiles maintain 0.40-0.50 friction wet or dry. The trade is cost (roughly 3-4× rubber) and the fact that ceramic is aggressive on the belt back cover, so you spec a thicker pulley-side cover (typically 4-6 mm). For dry, clean aggregate operations, plain diamond-grooved rubber is still the right call.
References & Further Reading
- Wikipedia contributors. Conveyor belt. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
