Classification efficiency evaluates how particle sizes in the feed report to the fine product and coarse return streams. Circulating load is the internal material flow returned from the classifier to the mill; it is not a universal good-or-bad ratio by itself.
What does classification efficiency measure?
Classification efficiency examines how particle-size classes divide between fine and coarse streams, not only the total quantity of product. A sound assessment measures representative feed, fine-product and coarse-return samples with the same method.
Why is one overall percentage insufficient?
A single total percentage can hide misplacement around the boundary size. Size-by-size partition behaviour is more informative for product control.
Misplaced fines and coarse particles
Fine material reporting to the coarse return can be reground unnecessarily, while coarse material entering the product can shift indicators such as d90 and d97. Both directions of misplacement require evaluation.
Bypass
Bypass describes flow or particles reaching a stream without experiencing the expected classification mechanism. It influences interpretation of the partition curve and circuit efficiency.
Partition / Tromp curve
A partition curve shows the probability of each particle-size class reporting to a selected product or return stream. Its characteristic cut point and transition width help describe both the centre and sharpness of separation.
Advanced but useful control
A Tromp curve does not select equipment by itself. Without representative sampling, a closed mass balance and consistent measurement it can be misleading. It is therefore retained here as an advanced section rather than expanded into a separate thin article.
What is circulating load?
In a closed circuit, circulating load is the material rejected by the classifier and sent back for further grinding. Dry/wet basis, circuit boundary and sampling points must be stated in any calculation or report.
Why is there no universal optimum?
Suitable internal load depends on material, product target, mill capacity, classification sharpness and circuit configuration. A percentage copied from another plant is not a reliable design target.
How can circulating load affect the mill?
| Observation | Possible circuit meaning | Checks |
|---|---|---|
| Recycle is increasing | Feed or separation conditions may have changed | Fresh feed, PSD, rotor/air and sample balance |
| Coarse product tail is increasing | Coarse misplacement or measurement shift | d90/d97, partition curve and preparation |
| Return contains excessive fines | Bypass or poor separation sharpness | Fine fraction, flow distribution and fan/filter conditions |
| Mill load is unstable | Classifier and feed interaction may be involved | Stream trends, throughput and control loop |
| Pressure loss is changing | Air path, filter or build-up may be involved | Fan, duct, sealing and cleaning |
The table lists diagnostic topics; it is not a stand-alone fault diagnosis or operating set point.
How is a classification system monitored?
- Plan sampling of feed, fines and return streams
- Record flow and mass balance on a stated basis
- Keep the PSD method consistent across streams
- Trend d50/d90/d97 in the product context
- Assess partition behaviour with sufficient size classes
- Match fan, filter and pressure-loss trends to process data
- Change one setting at a time under a recorded test plan
Limits in circuit optimisation
The classification system efficiency approach published by AusIMM combines circulating-load ratio, separation sharpness and classifier bypass. Each plant still requires its own sampling plan, circuit boundary and product target.
Do not confuse an air separator with a hydrocyclone
Both are classifiers, but their process medium, physics and measurement conditions differ. This page discusses general closed-circuit concepts and does not claim that Estilo supplies hydrocyclones.
This guide publishes no universal optimum circulating load, efficiency percentage or operating setting.
