Belt Weigh Feeders in Cement Plants

 Belt Weigh Feeders in Cement Plants: Working Principle, Calibration, and Field-Proven Fault Tracing

If you've spent any time in a cement plant control room, you already know the belt weigh feeder is one of the most argued-about instruments on the site. Production wants accurate raw mix proportioning. Process wants stable kiln feed. Maintenance wants it to just stop drifting. And somewhere in between sits the I&C technician, staring at a load cell reading that doesn't match what's actually on the belt.

This article is written from twenty-five years of hands-on troubleshooting on Schenck and Hasler feeders across raw mill, coal mill, and additive dosing lines. The goal isn't to repeat the theory you'll find in a vendor manual — it's to walk through how the system actually behaves in the field, where it fails, and how to fix it without guesswork.

How a Belt Weigh Feeder Actually Works

A belt weigh feeder combines three measurements into one continuous mass flow calculation:

Belt load — measured by a load cell (or load cell pair) mounted under an idler frame, sensing the weight of material on a fixed section of belt.

Belt speed — measured by a tachometer or encoder coupled to the drive pulley or a dedicated speed roller.

Belt length calibration factor — the mechanical span between weighing idlers, which converts load-per-unit-length into a mass flow rate.

The integrator (Schenck Disomat, Siemens Milltronics, or a PLC-based totalizer) multiplies load × speed continuously, producing an instantaneous flow rate in tons/hour and a running total in tons. This looks simple on paper. In practice, every one of those three inputs is vulnerable to mechanical drift, and the feeder doesn't tell you which one is lying — it just gives you a wrong number.

Why Feeders Drift: The Real Root Causes

Most technicians jump straight to "let's do a material test" when a feeder reads inaccurately. Before you burn a shift doing that, check these mechanical culprits first — in my experience they account for the majority of drift complaints:

Belt tension changes. As the belt ages and stretches, or as the take-up tension is adjusted, the load distribution across the weighing idler changes even though the actual material weight hasn't. A belt that's too loose will sag between idlers, causing the weighing idler to read artificially high because it's carrying more of the belt's own dead weight in addition to material.

Idler misalignment or bearing wear. A worn or seized idler bearing anywhere near the weigh span adds friction that doesn't show up as "load" in the traditional sense but does show up as speed variation, which corrupts the flow calculation just as badly as a bad weight reading.

Material buildup on the belt or skirting. Sticky raw meal or coal fines building up on return idlers or skirt rubber changes belt tracking and adds unaccounted mass. This is especially common on coal feeders during humid weather — check skirting rubber wear monthly, not just when you get a complaint.

Load cell zero drift from temperature. Strain gauge load cells are temperature-compensated, but not infinitely so. A feeder installed near a kiln inlet or in an uninsulated shed will show seasonal zero drift. If your zero calibration keeps wandering in the same direction every few months, check ambient temperature swings before suspecting the load cell itself.

Speed sensor slippage. If the tachometer is friction-coupled to the belt or pulley rather than directly geared, wear on the friction wheel causes speed under-reading — and because flow = load × speed, this quietly under-reports tonnage even when the load cell is perfectly accurate. This is one of the most missed faults because the load cell "checks out fine" on a static test, and nobody thinks to check the speed reference independently.

Static vs Dynamic Calibration — Do Both, Not Either

A lot of confusion in the field comes from technicians treating static (zero and span) calibration as sufficient on its own.

Static calibration confirms the load cell and electronics are reading correctly at rest, using test weights or a calibration chain. This tells you the load cell path is healthy but says nothing about speed accuracy or dynamic behavior under running conditions.

Dynamic (material) calibration is the only way to confirm actual delivered tonnage, done by running a known batch across the feeder and comparing the integrator total against a verified weighbridge or belt-cut sample. This is the test that actually matters to production, and it's the one most often skipped or rushed because it takes the feeder offline.

My practical rule: static calibration monthly, material test calibration quarterly, and immediately after any mechanical work on the belt, idlers, or drive — even something as minor as replacing a bearing can shift the calibration factor enough to matter for raw mix accuracy.

Fault Tracing Sequence: A Practical Checklist

When a feeder reading looks wrong, don't start with electronics. Work in this order:

Visual belt check — tracking, skirting condition, material buildup on idlers, belt tension by eye.

Zero check with belt running empty — if zero is off, it's almost always mechanical (buildup, tension) rather than electrical.

Load cell signal check — verify mV/V output against the calibration certificate at a known static load. A drifted mV/V reading with clean wiring points to the load cell itself or moisture ingress at the junction box.

Speed reference check — compare tachometer/encoder pulse output against a handheld tachometer on the pulley. Don't assume the coupling is intact just because the display is moving.

Cable and junction box inspection — cement plant environments are brutal on cabling. Moisture ingress at load cell summing boxes is one of the most common causes of erratic (not just offset) readings, and it's frequently misdiagnosed as a "faulty load cell" when the cell itself tests fine on the bench.

Integrator configuration check — confirm the belt length, calibration factor, and filter/damping settings haven't been altered, especially after a firmware update or card replacement.

A Word on Documentation

Every calibration — static or dynamic — should be logged with date, technician, ambient conditions, and the specific reason (routine, post-maintenance, complaint-driven). When a feeder's accuracy gets questioned by production months later, this log is what separates "we know exactly when and why it drifted" from an argument with no data behind it. If you don't already have a simple calibration/inspection form for your weigh feeders, it's worth setting one up — it pays for itself the first time there's a dispute over raw mix consistency.

Takeaway

A belt weigh feeder isn't just a load cell with a display. It's a mechanical-electrical system where the belt, idlers, drive, and sensors all have to agree with each other before the number on the screen means anything. Most "instrument faults" reported against weigh feeders are actually mechanical issues wearing an electrical disguise — and the technicians who get to the root cause fastest are the ones who check the belt before they touch the electronics.