Idler Roller Case Study: Eight Kilometres Through the Wet Season

An iron ore operation in West Africa lost more return rollers in the four wet months than in the other eight combined. The cause was not the rollers. It was carryback: sticky fines that pass the scraper in the rain, pack onto the return side, and grow the roll diameter until the shell stops turning. We changed the roller build, changed the wash-down, and the monthly seizure count fell from forty-one to three.

Project Snapshot

The Problem: The Rain Changed What the Scraper Could Do

Dry ore behaves. It leaves the belt at the head pulley, the scraper takes what is left, and the return side stays reasonably clean.

Wet ore does not. Fines turn to paste, the paste passes the scraper, and it rides back on the belt to the return rollers. There it packs into the gap between the shell and the frame, and the effective diameter of the roll grows.

Once the diameter grows, two things follow. The belt stops tracking true, and the shell starts dragging against material that is no longer moving. A seized return roller then gets a flat spot, and a flat spot cuts the belt cover. The operator was changing rollers almost daily in August, and each one was a stoppage on a line that feeds a ship loading schedule.

Why We Chose a Sealed Return Build Over More Wash-Down, Scrapers Alone and a Full Non-Metallic Fleet

Wash the return side harder and more often. Excluded. High-pressure water directed at the roller ends drives water past the seal, and the bearing then fails from the cleaning rather than from the ore.

Fit another stage of belt scrapers. Excluded. Scrapers help, and they do not remove paste in a four-month rainy season; the carryback load simply exceeds what any scraper arrangement holds back.

Replace the whole fleet with non-metallic shells. Excluded. Non-metallic shells resist build-up well, but the loading zone and the impact areas still need their own specification, and a blanket change is not a specification.

Rebuild the return side with better sealing, keep non-metallic shells where build-up is worst, and change the wash-down method. Selected. It treats the mechanism rather than the symptom.

What We Manufactured

Every return roller in the wet-season zone was built with a multi-stage labyrinth seal and a greased-for-life bearing, and each one passes a 72-hour water ingress test before dispatch. Radial runout is held below 0.35 mm, because a roll that does not run true both vibrates and wears the belt.

Where build-up was worst, close to the head end, the specification moves to a non-metallic shell. Paste bonds less readily to a non-metallic surface, and where it does bond it releases under the scraper rather than under hand tools.

Carry-side tracking was addressed at the same time. Self-aligning idler sets were fitted at fixed intervals so a belt that wanders under a wet load corrects itself instead of running into the structure.

Installation: How the Programme Ran

Change-out was staged across six weeks of planned stops rather than done as one shutdown, because the ship loading schedule had no window for a full-day stop.

Rollers were delivered numbered by chainage. That sounds administrative, and on a 7.4 km line it is the difference between a six-week programme and a six-month one, because crews otherwise fit whatever is closest to hand.

Wash-down procedure changed in the same month. Pressure at the roller ends was reduced and the nozzle angle moved off the seal face, which was the single change that stopped the bearings being washed out.

Testing and Quality Control

Bearings are checked on arrival and shells are checked for runout and wall thickness before assembly. Every roller is then run and measured for drag, held below 0.010, and sampled for the 72-hour water ingress test.

Sealing is verified on the assembled roller rather than on the components, which is the point most suppliers skip. A seal that measures correctly and seats badly is a seal that fails in month two.

The failure mode we were designing against is set out in more detail in our note on common idler roller failures. Design life for the return build is 30,000 hours, and the operator now tracks against that figure per chainage.

Results: What the Numbers Show

Seized rollers fell from 41 per month in the previous wet season to 3 per month across the four wet months after the change, on the same ore and the same rainfall.

Belt cover wear over the following twelve months was visibly lower at the head end, and the operator has deferred a belt section replacement that was in the previous year's budget.

Tracking complaints have essentially stopped. The aligning sets plus the reduced wash-down pressure removed the two causes together, which is why the result held through the whole season rather than for a few weeks.

What This Case Means for Your Project

Carryback is a wet-season problem with a dry-season budget. If your seizure count is seasonal, the fix is rarely a stronger roller on its own; it is a roller that keeps water out and a wash-down that stops putting it in.

Send us your belt width, trough angle, chainage length and where the seizures cluster, and we will specify the return side by zone rather than as one part number. Curved sections need their own specification, and you can send your conveyor data and curve radius to our engineering team.