Cement Grinding in Ball Mill: Circuit & Control

This is the second of two companion articles on cement grinding in ball mills, and it assumes the fundamentals covered in the first: the purpose of finish grinding, the feed chemistry, the circuit equipment, the Bond theory and the three control loops of load, quality and temperature. This part goes deeper into the practice that separates good plants from the best: the advanced audit of the mill and separator, the interaction of the particle size distribution with cement performance, the role of grinding aids and their optimization, the control strategies that hold the mill on its optimum, the management of feed variability, the economics of energy, and the troubleshooting of the difficult cases that defeat the standard procedures. It is written for the process engineers, production managers and senior operators who run the finish grinding department, and it is designed to be read alongside its companion article and used as the reference for the plant’s own grinding studies.

1. The Two-Article Structure and What Part 2 Adds

The first article established the complete framework: the purpose of the process, the materials, the circuit, the theory, the internals, the separator, the temperature control, the daily practice, the economics and the quality loop. This second article adds the depth that the daily operation cannot carry: the quantitative methods that find the last 10 percent of performance, the understanding of why the particle size distribution behaves the way it does, and the organizational practices that keep a grinding department excellent over years, not months. The two articles together are the complete package: Part 1 is the operator’s handbook, Part 2 is the engineer’s workbook.

The structure of this part follows the logic of the improvement cycle: understand the distribution and its quality implications, measure the circuit and its efficiency, then act — on the internals, the separator, the aids, the control and the feed. Each section is written to be actionable: it names the measurement, the target and the expected effect, and it closes with the case that illustrates the principle in a real plant context.

2. The Particle Size Distribution as the Product Definition

The cement’s behavior in concrete is the product of its particle size distribution, and the ball mill circuit is the instrument that shapes it. The distribution is described by the Rosin-Rammler equation, R = exp(-(x/x0)^n), where R is the fraction retained on size x, x0 is the characteristic size at 36.8 percent retained, and n is the distribution slope. The slope n is the shape parameter: a steep distribution with n of 1.0 to 1.2 is characteristic of a well-classified ball mill product, with a narrow size range and a small coarse tail; a flat distribution with n below 0.8 has a long coarse tail and is typical of overgrinding in an open circuit. The n value is the most underused quality parameter in the industry: it is measured in minutes by laser diffraction, and it explains the strength and the water demand behavior that the Blaine alone cannot.

The practical relationships are documented across decades of cement research: for a given Blaine, a steeper distribution gives higher early strength because the fine fraction is concentrated where it hydrates fastest, while a flatter distribution gives more late strength from the surviving coarse particles but higher water demand. The cement types therefore specify not just a Blaine but a distribution shape: the plants that control n as well as Blaine deliver cement whose concrete behavior is stable, and the plants that chase Blaine alone deliver cement whose strength varies despite identical fineness. The separator is the instrument of n: the cage speed sets x0, the air flows and the bypass set n, and the daily control of both is the real quality loop.

3. The Advanced Mill Audit

The advanced audit goes beyond the standard stop inspection to quantify the circuit’s efficiency. Its components are: the charge audit with the screened grading and the filling measurement; the liner and diaphragm wear measurement; the separator audit with the feed, product and reject sampling that yields the Tromp curve; the ventilation measurement with the mill pressure and the air velocity; and the instrumentation verification of the power, the feed and the temperature signals. The output is the complete picture: where the energy goes, where the coarse material circulates, and where the bottlenecks sit.

The Tromp curve is the audit’s centerpiece. The separator feed, the fines and the reject are sampled simultaneously, sieved and weighed, and the recovery of each size fraction to the fines is plotted against the particle size. The ideal curve is a step: 100 percent recovery below the cut size, zero above. The real curve has a characteristic shape whose features quantify the separator’s health: the cut size x50, the sharpness index, and the bypass, the fraction of coarse material that passes to the fines at large particle sizes. A bypass of 10 to 20 percent is typical of a tired third-generation separator; a bypass above 25 percent with a flat curve is the signature of worn vanes, leaking flaps or a poor dispersion, and its cost is the regrinding energy of the bypassed coarse fraction — commonly 2 to 5 kWh per tonne of the circuit’s consumption. The Tromp curve, measured twice a year, is the separator’s health certificate.

4. The Energy Reconciliation and Its Findings

The energy reconciliation is the accounting that locates the losses. The measured inputs are the circuit’s electrical energy — the mill motor, the separator, the fans, the conveyors — and the measured outputs are the fresh feed and the product fineness. The reconciliation compares the measured specific power with the Bond calculation for the feed and the product, and the difference is allocated by the audit: the separator bypass, the charge inefficiency, the overgrinding of the circulating load, the ventilation losses and the mechanical losses. The typical findings of a mature reconciliation on an average plant are instructive: 5 to 10 percent of the energy in the separator bypass, 5 to 15 percent in an imperfect charge, 3 to 8 percent in the ventilation and conveying, and the balance in the theoretical grinding energy and the motor losses.

The findings translate directly into projects. A 5 percent bypass reduction at 30 kWh per tonne is 1.5 kWh per tonne, worth 0.15 dollars per tonne, or 225,000 dollars per year on a 1.5-million-tonne plant. The audit’s other classic finding is the overgrinding of the circulating load: a mill grinding its reject at the final fineness is spending the fine-grinding energy on material that will be ground again, which is why the first-compartment efficiency and the separator cut are reviewed together. The reconciliation is the annual report of the grinding department, and the projects of the following year are its action list.

5. Grinding Aids: Chemistry, Selection and Optimization

The grinding aids are the chemical layer of the mill’s performance, and their selection and optimization is an engineering study in itself. The active molecules — alkanolamines such as triethanolamine and diethanolisopropanolamine, glycols, and the proprietary polymer blends — adsorb on the fresh particle surfaces and reduce the surface energy that drives agglomeration and ball coating. The measurable effects are a lower mill power at the same fineness, a higher capacity at the same power, a finer product at the same energy, a better separator performance through better dispersion, and, for the quality-active formulations, an influence on the hydration: the alkanolamine fraction can accelerate the early strength development and modify the distribution.

The selection of the aid is a laboratory and plant trial: the candidate products are tested on the plant’s clinker, first in the laboratory mill and then in the plant mill, with the Blaine, the residue, the distribution slope, the strength and the mill data all measured. The optimization is a dosing matrix: the dose is stepped from 200 to 800 grams per tonne, and the capacity and the power are recorded at each level until the dose-response curve identifies the economic optimum — the dose where the value of the capacity and energy gains exceeds the aid cost. The documentation of the trial, the batch number and the results is the plant’s aid record, and the re-optimization is repeated when the clinker, the additions or the cement types change materially.

6. Control Strategies Beyond the Basic Loops

The basic three-loop control holds the mill; the advanced strategies optimize it. The mill load controller is the foundation: it manipulates the feed rate against the mill power or the differential pressure, and its tuning — the gain, the dead band and the rate limits — decides whether the mill rides its maximum stable load or oscillates around it. The advanced layer adds the feed-forward: the clinker grindability, measured or predicted from the kiln data, adjusts the load set-point before the mill feels the change, instead of after. The quality layer adds the fineness feedback: the online Blaine or residue measurement, where installed, closes the loop on the separator speed with a cycle time of minutes instead of hours.

The supervisory and model predictive control layers coordinate the loops: the optimizer computes the operating point that maximizes the production at the target fineness and temperature, subject to the separator, the elevator and the silo constraints, and it moves the loops as a system. The documented gains of the advanced layer over well-tuned conventional control are 3 to 6 percent capacity and a measurable reduction in fineness variability. The precondition of the advanced layer is the basic layer: a plant whose load loop oscillates cannot be fixed by a supervisory layer, and the sequence of tuning — basic loops first, then the supervisory, then the optimizer — is the only sequence that works.

7. Managing Feed Variability

The feed to the finish mill is the least controllable variable of the circuit: the clinker arrives from the kiln with its grindability, its temperature, its moisture and its lump size all varying with the kiln operation and the season. The management of the variability has four layers. The first is the silo buffer: the clinker storage between the kiln and the mill, sized for 5 to 10 days, averages the short-term variation and is managed by a reclaim strategy — reclaim from the oldest, blend the new with the old. The second is the measurement: the clinker grindability is tested on the standard schedule, and the hard batches are flagged to the mill shift. The third is the feed-forward: the grindability result adjusts the mill’s load set-point and the separator’s fineness loop before the batch reaches the mill. The fourth is the mix: the feed bins blend the clinker, the slag and the limestone so that the mill sees a target average, and the blend proportions are adjusted by the quality department against the grindability results.

The moisture deserves its own paragraph: the clinker moisture, normally below 1 percent, rises with the season and the cooler performance, and the feed moisture above about 1.5 percent degrades the mill’s performance measurably — the moisture sticks the fines to the media, raises the temperature and floods the diaphragm at the limit. The plant’s defense is the moisture measurement at the mill feed, the drying capacity of the mill ventilation, and the cooperation with the kiln and cooler departments, because the clinker moisture is made in the cooler, not in the mill.

8. The Circuit Bottleneck Analysis

The optimized mill eventually hits a bottleneck, and the bottleneck analysis finds where the circuit’s true capacity limit sits. The candidates are the mill itself — its power, its ventilation and its charge; the separator — its rotor capacity, its air flow and its dispersion; the elevator and the conveyors — their capacity at the circulating load; the dust collector — its cloth area and its pressure drop; and the feed and the product systems. The analysis measures the actual flow through each element against its rating, under the condition of maximum mill load, and identifies the element that reaches its limit first: the bottleneck. The practical method is the incremental test: raise the feed until a limiting signal appears, and the first signal that saturates names the bottleneck.

The bottleneck analysis drives the investment plan: a separator with a saturated rotor justifies a new rotor or a new separator; a saturated dust collector justifies a cloth upgrade; a saturated elevator justifies a speed or bucket upgrade. The order of the investments follows the bottlenecks: each investment moves the limit to the next element, and the plan sequences the moves so that every dollar is spent on the element that is actually limiting. The classic error — installing a new separator on a mill whose charge is degraded, or a new charge on a mill whose separator is the limit — is avoided by measuring the bottleneck before the capital project, not after.

9. The Troubleshooting of the Difficult Cases

The standard symptom matrix covers the common cases; the difficult cases need the deeper methods. Case one: the capacity falls slowly over months at constant settings and constant power — the signature of a charge that has degraded while being topped up with the wrong additions; the diagnosis is the screened grading, and the fix is the re-grade. Case two: the fineness is on target but the strength is falling — the signature of a distribution change: the n value has flattened, the coarse tail has grown, or the clinker phase composition has shifted; the diagnosis is the laser diffraction and the clinker microscopy, and the fix is the separator sharpness and the raw mix review. Case three: the mill power oscillates at a period of tens of minutes — the signature of an unstable load loop or a periodic feed disturbance; the diagnosis is the control loop data and the weigh feeder data, and the fix is the loop tuning or the feeder maintenance. Case four: the reject rate climbs while the fineness holds — the signature of a separator wearing or a dispersion fault; the diagnosis is the Tromp curve, and the fix is the vane, the cage or the feed distribution. Each case is solved by the data, and the case log of the plant is the training material of its engineers.

10. The Economics of the Grinding Department

The grinding department’s economics are decided by four numbers: the specific power, the capacity, the media consumption and the aid cost. The specific power at 28 to 34 kWh per tonne is the electricity bill; the capacity decides the dispatch and the margin; the media at 200 to 600 grams per tonne is the consumable bill; and the aid at 200 to 800 grams per tonne is the chemical bill. The four interact: the aid that raises the capacity lowers the specific power and the media consumption per tonne, and the charge policy that extends the media life at the cost of capacity is a false economy on most plants. The annual economic review of the department — the actual versus the target of the four numbers, with the reconciliation of the differences — is the meeting where the grinding strategy is set, and the investments are ranked by their payback on the department’s own data.

11. Organizational Practices of the Best Grinding Departments

The organizational layer decides whether the technical practices survive. The practices of the best departments are consistent across the industry: a monthly grinding performance review with a fixed agenda; a documented change control that logs every setting change with its author and its measured effect; a stop-based inspection calendar that protects the measurement discipline; a KPI dashboard that shows the specific power, the capacity, the fineness variability and the media consumption against the targets; and a training program that keeps the shift teams current with the circuit’s state. The department’s knowledge is institutionalized: the mill file holds the design charge, the Tromp curves, the audit reports and the case log, and the engineer who leaves does not take the knowledge with them.

The second organizational practice is the relationship with the other departments: the kiln, because the clinker grindability is made there; the quality, because the cement is sold on its distribution; and the logistics, because the silo capacity and the dispatch decide the production plan. The grinding department that optimizes in isolation optimizes the wrong thing: the optimum of the plant — the clinker quality, the grinding capacity and the market demand together — is the real objective, and the coordination meetings are where it is set.

12. The Benchmark and the Road Ahead

The benchmark of the best-in-class finish grinding department is a set of numbers that the industry recognizes: 24 to 28 kWh per tonne of specific power on a closed-circuit ball mill with a modern separator and aids; a fineness variability of 2 to 3 percent around the target; a distribution slope n held within its band; a media consumption at the low end of its range; and a quality rejection rate near zero. The road to the benchmark is the sequence repeated every year: measure, audit, correct, optimize, verify — and the annual cycle compounds: the plant that runs the cycle for five years has a grinding department that the new plants cannot match, because the knowledge is in the mill file and the discipline is in the team.

The road ahead is the technology that will change the benchmark: the vertical roller mill at lower energy for the bulk products, the high-pressure grinding roll as the pregrinder that halves the ball mill’s work, the continuous online quality measurement that closes the loop in minutes, and the machine learning layer that predicts the charge condition and the separator state from the daily data. The ball mill will not disappear: its distribution quality and its flexibility keep it at the center of finish grinding for the premium products, and the engineer who masters the practice of this article will master whichever machine the future brings, because the principles — measure, understand, control — are the same.

Frequently Asked Questions

What is the difference between Part 1 and Part 2 of this topic?

Part 1 is the complete foundation: the purpose, the materials, the circuit, the theory, the operation and the quality control, written as the operator’s handbook. Part 2 is the engineer’s workbook: the particle size distribution science, the Tromp curve, the energy reconciliation, the aid optimization, the advanced control and the organizational practices.

What does the Tromp curve tell the engineer?

The Tromp curve plots the separator’s recovery of each particle size to the fines. Its cut size, sharpness and bypass quantify the separator’s health, and the bypass — the coarse material escaping into the fines — is the direct measure of wasted regrinding energy, typically 2 to 5 kWh per tonne of the circuit’s consumption.

Why is the distribution slope n as important as the Blaine?

The Blaine measures the total surface area but not its distribution. The slope n describes the shape: a steep distribution gives high early strength and low water demand, a flat one gives more late strength from the coarse tail at the cost of workability. The separator controls n, and the plants that control both deliver consistent concrete behavior.

How is the circuit bottleneck found?

By the incremental test: raise the feed until the first signal saturates — the separator rotor, the elevator, the dust collector or the mill itself. The first element that limits names the bottleneck, and the investment plan follows the order of the bottlenecks.

How much can the advanced practices save on an existing mill?

The full sequence — audit, charge correction, separator sharpening, aid optimization and control improvement — typically delivers 10 to 20 percent more capacity and 5 to 10 percent lower specific power on the same mill, worth 1 to 3 dollars per tonne of cement in energy and fixed cost terms on a modern plant.

Summary

This second part of the cement grinding in ball mill package has taken the practice beyond the fundamentals: the particle size distribution as the product definition, the Tromp curve and the energy reconciliation as the quantitative audit, the aid chemistry and its dosing optimization, the advanced control layer, the management of feed variability, the bottleneck analysis, the difficult troubleshooting cases, the department economics and the organizational practices that sustain excellence. The theme of the whole package is measurement: every improvement in this part is measured before and after, every loss is located by its instrument, and every practice is documented in the mill file. The ball mill circuit is the most measurable machine in the plant, and the plant that measures it comprehensively is the plant that controls it completely — and that control, in the end, is the whole difference between the average and the best-in-class grinding department.

13. The Closed Circuit Configuration and the Classifier Integration

The closed-circuit cement grinding couples the ball mill with the classifier loop: the mill discharges to the bucket elevator, the elevator feeds the separator, the separator splits the stream into the finished product and the rejects, and the rejects return to the mill inlet through the feed chute: the circulating load of the typical closed circuit is the 100-300% of the fresh feed (the rejected mass versus the fresh feed mass), and the circulating load is the key operating variable of the circuit. The classifier integration determines the circuit behavior: the 3rd generation dynamic separators with the rotor speed control, the guide vane settings and the air flows deliver the adjustable cut size and the steeper classification, while the static and the 2nd generation units fix the classification at the lower efficiency. The circuit tuning balances the mill feed moisture, the ventilation and the separator settings for the target product: the integrated control of the classifier is the heart of the modern closed-circuit cement grinding.

14. The Moisture and the Ventilation Control of the Circuit

The moisture and the ventilation control the material behavior inside the grinding circuit: the clinker moisture of the 0.5-2% and the gypsum water of crystallization must be evaporated by the mill ventilation, and the ventilation air of the 0.8-1.5 m/s through the mill carries the fine particles, removes the water vapour and cools the mill interior. The insufficient ventilation causes the coating on the liners, the diaphragm clogging, the temperature rise (the mill outlet above the 110-120 degrees dehydrates the gypsum into the hemihydrate and risks the false set) and the capacity loss; the excessive ventilation wastes the energy and lifts the fines prematurely. The modern circuits monitor the mill outlet temperature, the pressure drop, the filter differential and the gas flows, and the automatic damper control maintains the ventilation window: the moisture control of the circuit is the quiet variable behind the mill stability.

15. The Product Quality Control in the Grinding Circuit

The product quality control of the closed circuit operates on the particle size distribution: the laboratory samples the finished cement every 2-4 hours for the Blaine, the 45 micron residue and the SO3, and the quality control adjusts the separator speed, the feed rate and the gypsum content to hold the product inside the specification: the finer Blaine targets (the 380-420 m2/kg for the high early strength) come at the higher specific energy (the 40-50 kWh/t), and the coarser targets (the 300-330 m2/kg for the masonry products) at the 25-30 kWh/t. The quality control also monitors the strength development of the mortar cubes (the 2, 7 and 28-day series), the setting times and the soundness: the grinding circuit is the last quality gate of the cement plant, and its control loop closes the production chain from the quarry to the silo.

13. The Closed Circuit Configuration and the Classifier Integration

The closed-circuit cement grinding couples the ball mill with the classifier loop: the mill discharges to the bucket elevator, the elevator feeds the separator, the separator splits the stream into the finished product and the rejects, and the rejects return to the mill inlet through the feed chute: the circulating load of the typical closed circuit is the 100-300% of the fresh feed (the rejected mass versus the fresh feed mass), and the circulating load is the key operating variable of the circuit. The classifier integration determines the circuit behavior: the 3rd generation dynamic separators with the rotor speed control, the guide vane settings and the air flows deliver the adjustable cut size and the steeper classification, while the static and the 2nd generation units fix the classification at the lower efficiency. The circuit tuning balances the mill feed moisture, the ventilation and the separator settings for the target product: the integrated control of the classifier is the heart of the modern closed-circuit cement grinding.

14. The Moisture and the Ventilation Control of the Circuit

The moisture and the ventilation control the material behavior inside the grinding circuit: the clinker moisture of the 0.5-2% and the gypsum water of crystallization must be evaporated by the mill ventilation, and the ventilation air of the 0.8-1.5 m/s through the mill carries the fine particles, removes the water vapour and cools the mill interior. The insufficient ventilation causes the coating on the liners, the diaphragm clogging, the temperature rise (the mill outlet above the 110-120 degrees dehydrates the gypsum into the hemihydrate and risks the false set) and the capacity loss; the excessive ventilation wastes the energy and lifts the fines prematurely. The modern circuits monitor the mill outlet temperature, the pressure drop, the filter differential and the gas flows, and the automatic damper control maintains the ventilation window: the moisture control of the circuit is the quiet variable behind the mill stability.

15. The Product Quality Control in the Grinding Circuit

The product quality control of the closed circuit operates on the particle size distribution: the laboratory samples the finished cement every 2-4 hours for the Blaine, the 45 micron residue and the SO3, and the quality control adjusts the separator speed, the feed rate and the gypsum content to hold the product inside the specification: the finer Blaine targets (the 380-420 m2/kg for the high early strength) come at the higher specific energy (the 40-50 kWh/t), and the coarser targets (the 300-330 m2/kg for the masonry products) at the 25-30 kWh/t. The quality control also monitors the strength development of the mortar cubes (the 2, 7 and 28-day series), the setting times and the soundness: the grinding circuit is the last quality gate of the cement plant, and its control loop closes the production chain from the quarry to the silo.

This file is part of the Complete Cement Technical Package (931 files) available from cementequipment.org. Respective rights holders; library copy for the licensed single user.