When 90μm sieve residue starts running above the upper end of the operating window, raw meal is no longer being classified the way the burning side assumes. Coarse particles take longer to calcine, push up free lime, and force the kiln to compensate with higher burning zone temperature or lower throughput. The cause is almost always inside the separator-and-classifier system rather than in the grinding zone itself, but quarry hardness changes and worn grinding surfaces can mimic the same signal. A structured walk through separator speed, blade condition, and feed quality usually identifies the driver in one shift.
Common Causes
1. Separator speed too low
Reduced rotor speed lets coarse particles bypass into finished product. This is the most common cause when residue creeps up gradually with no other parameter changes — typical drift after a control-room set-point change or a restart sequence.
2. Worn classifier blades or guide vanes
Once rotor blades and guide vanes lose their profile, classification sharpness drops and the cut-point spreads. Even at the original speed setpoint, more coarse material passes through. The signal is residue rising while specific power stays flat.
3. Quarry zone change increasing feed hardness
Harder feed needs more grinding energy. If grinding pressure and table liner condition are not matched to the new Bond Work Index, the mill produces a coarser cut even at design throughput.
4. Worn table liner or roller tyres
Lost profile on grinding surfaces reduces the nip and the bed compaction. The mill can still meet tonnage, but the particle size distribution shifts coarser without an obvious vibration or power signal.
5. Excessive throughput overloading the classifier
Pushing throughput beyond the classifier's design loading collapses separation efficiency. Residue rises sharply with feed rate and recovers when feed is backed off.
How to Diagnose
-
01Pull a fresh 90μm sieve residue from freshly-ground product and confirm the lab result against the trend — exclude sampling and instrument drift before going further.
-
02Increase separator speed in 5 rpm increments and re-sample after each step. If residue responds normally, the issue is set-point drift; if it does not move, suspect mechanical wear in the classifier.
-
03Inspect classifier rotor blades and guide vanes through the inspection door. Look for rounded edges, missing tip segments, or asymmetric wear that distorts the cut-point.
-
04Measure roller tyre and table liner profiles against the wear template. If either is below minimum thickness, schedule replacement and adjust hydraulic pressure as an interim measure.
-
05Sample fresh feed and run a quick Bond Work Index check or hardness comparison against the baseline — a quarry zone change is a common silent cause.
-
06Reduce feed rate by 10% and confirm whether residue recovers — this isolates classifier overload from a true grinding problem.
-
07Cross-check 90μm residue against kiln inlet LSF standard deviation; if both have moved together, the mill, not the lab, is the driver.
Process Impact
Coarse raw meal does not burn the way the kiln control loop assumes. Particles above the design cut take longer to calcine, the burning zone has to compensate with higher temperature or longer residence time, and free lime in clinker tends to creep up. Heat consumption rises in step with that compensation. Quality follows: clinker C₃S falls, 28-day strength softens, and the cement mill has to be pushed harder to recover finished-product Blaine. Raw meal homogeneity also worsens, because coarse fractions segregate in storage and extraction. The longer residue sits above target, the more these effects compound — and once the kiln has absorbed the change as the new normal, recovery becomes a multi-shift exercise rather than a single setpoint move.
Operating Targets
| Parameter | Target | Action threshold |
|---|---|---|
| Raw meal residue (90μm) | 10–14% R90μm | Action above 18% sustained |
| Kiln inlet LSF stdev | < 1.0 | Investigate if > 1.5 |
| Product moisture | < 0.5% | Drying review above 0.8% |