Quick answer
When the cement mill separator reject rate climbs while the Blaine fineness remains unchanged, the most common root causes are separator wear, airflow imbalance, or feed‑distribution problems. Verify separator plate condition, check the separator’s air‑flow set‑point against the OEM recommendation, and confirm even feed to the classifier. Corrective actions typically involve cleaning or replacing separator plates, adjusting the separator fan speed, and re‑balancing the feed distribution system. After each adjustment, monitor reject rate and product fineness until both return to the baseline range.
Problem / job
The mill’s separator is delivering a higher percentage of reject material to the bypass loop, yet the measured Blaine fineness of the product remains within the target range. The plant aims to restore the reject rate to its historical level without compromising product quality or throughput.
Evidence
- Reject rate increased from the typical X % to Y % over the past 48 hours (exact values to be confirmed from plant data).
- Online Blaine measurements have shown no statistically significant drift; the average remains within the specified tolerance.
- Separator fan vibration levels are within normal limits, but a slight rise in motor current has been logged.
- Visual inspection of the separator plates during the last scheduled shutdown revealed moderate wear on the downstream plate set.
- Air‑flow set‑point on the separator control panel reads at the nominal value, but the OEM manual notes that the optimum set‑point can shift with wear.
Decision
Proceed with a systematic diagnostic workflow that isolates the separator, validates airflow and plate condition, and checks feed distribution. Implement corrective actions in the order of least‑intrusive to most‑intrusive, documenting each change and its effect on reject rate and Blaine. Accept the solution when the reject rate stabilizes at the historical baseline and product fineness remains within the target range.
Practical 7‑step workflow
- Data capture and baseline confirmation
Gather the last 72 hours of reject‑rate data, Blaine readings, separator fan current, and vibration logs. Plot the trends to confirm that the increase is sustained and not a transient spike.
- Secure the mill for inspection
Apply lock‑out/tag‑out (LOTO) to the mill drive, separator motor, and any pneumatic actuators. Verify isolation with a voltage tester and post a “Do Not Start” tag. Only qualified personnel may remove LOTO after the inspection is complete.
- Visual and dimensional inspection of separator plates
Remove the separator housing according to the OEM disassembly procedure. Inspect the upstream and downstream plate sets for wear, fouling, or damage. Measure plate spacing at several points; compare with the OEM design drawing. If wear exceeds the OEM‑specified limit, plan for plate refurbishment or replacement.
- Air‑flow verification
Measure the actual separator fan airflow using a calibrated anemometer or flow‑meter. Compare the measured value with the set‑point displayed on the control panel. If the measured airflow deviates, adjust the fan speed or control valve to bring the flow back within the OEM‑recommended range. Document the OEM‑specified airflow range for future reference.
- Feed‑distribution assessment
Check the distribution of mill feed to the classifier. Verify that the feed inlet is centered and that the distribution nozzles are not clogged. Use a temporary flow‑meter or visual tracer (e.g., dyed water) to confirm even distribution across the classifier surface.
- Re‑assemble and re‑commission
Re‑install the separator plates, ensuring proper alignment and torque of all fasteners per the OEM manual. Remove LOTO only after a complete visual check and a functional test of the interlocks. Restart the mill at a reduced load and observe separator performance for at least one full production cycle.
- Performance verification
Record reject rate, Blaine, fan current, and vibration for the next 24 hours. Compare the values against the baseline established in step 1. If the reject rate has returned to the historical range and Blaine remains stable, close the investigation. If not, repeat steps 3–5 focusing on the most divergent parameter.
Diagnostic matrix
| Evidence observed | Likely direction of root cause | Next check / corrective action |
|---|---|---|
| Reject rate rising, Blaine stable | Separator is over‑classifying fine particles | Verify separator airflow and plate condition |
| Separator fan current increase, vibration normal | Airflow restriction due to plate wear or fouling | Inspect and clean/replace downstream plates |
| Plate spacing measured larger than design | Excessive wear on downstream plate set | Replace or refurbish plates; re‑set spacing |
| Measured airflow lower than set‑point | Fan speed set too low or valve partially closed | Adjust fan speed or valve; confirm with flow‑meter |
| Uneven feed distribution across classifier | Clogged distribution nozzles or mis‑aligned inlet | Clean nozzles; realign feed inlet; verify with tracer |
| No change after plate cleaning | Possible internal wear beyond cleaning benefit | Replace plates; re‑evaluate reject rate |
| Reject rate unchanged after airflow adjustment | Root cause likely feed‑distribution or mechanical wear | Proceed to feed‑distribution check; consider mechanical inspection of classifier rotor |
Acceptance criteria
All corrective actions are considered successful when the following measurable conditions are met for at least two consecutive production cycles:
- Reject rate – Returns to the historical baseline range documented in the plant’s performance log (e.g., 2 % ± 0.5 %). Verify with the plant’s reject‑rate monitoring system.
- Blaine fineness – Remains within the target tolerance (e.g., 3500 ± 150 m²/kg) as measured by the online Blaine analyzer.
- Separator fan current – Stabilizes at the OEM‑specified operating range for the current production load. Confirm with the motor’s current monitor.
- Airflow – Measured flow matches the control‑panel set‑point within the OEM‑approved deviation (typically ±5 %).
- Plate spacing – Measured spacing is within the OEM‑defined tolerance after any plate replacement or adjustment.
- Feed distribution – Visual or tracer test shows uniform distribution across the classifier surface, with no more than a 10 % variation between quadrants.
If any of the above criteria fall outside the verified limits, repeat the diagnostic loop, focusing on the parameter that failed.
Common mistakes and repeat‑failure traps
- Adjusting fan speed without confirming plate condition – A higher fan speed can mask airflow restriction caused by worn plates, leading to recurring reject spikes after the fan controller reverts to its default set‑point.
- Cleaning plates but not measuring spacing – Residual wear may still be present; without a dimensional check, the root cause remains unresolved.
- Skipping LOTO during plate removal – Unexpected motor start‑up can cause injury and may also damage the separator housing, creating new sources of imbalance.
- Relying solely on reject‑rate trends – Ignoring concurrent changes in fan current, vibration, or temperature can lead to mis‑diagnosis of mechanical versus aerodynamic issues.
- Changing the set‑point based on a single data point – Transient fluctuations are common; always base adjustments on averaged data over multiple cycles.
- Neglecting feed‑distribution checks after plate work – Plate removal can disturb the feed inlet geometry; failing to re‑verify distribution may re‑introduce the original symptom.
- Re‑using damaged plates – Attempting to re‑install plates with visible cracks or severe wear often results in immediate reject‑rate escalation.
By systematically avoiding these pitfalls, the plant can achieve a durable resolution and maintain consistent product quality.
Related Infinity guides
Build a stronger cement maintenance library
The Infinity for Cement Organization technical package contains 900+ technical files. It is a one-time $249.99 purchase, with instant download immediately after payment and lifetime updates forever at no additional cost.