Cyclone Design for Cement Preheater: Guide

Few pieces of process equipment are as underestimated as the cyclone, yet none is more woven into the working of a cement plant. Every preheater tower stage is a cyclone, the clinker cooler vents its hot air through cyclones, the raw mill and finish mill circuits classify material in separators built on cyclone physics, and the baghouses that guard the environment are protected upstream by cyclone pre-cleaners. The cyclones design workbook (cyclones design (1).xls, approximately 0.46 MB) from the cementequipment.org package collects the complete engineering of these devices in one set of linked sheets, and this article is the full technical companion to it. It develops the subject from the fundamental equation of volumetric flow, Q, through the geometry of the cyclone body, the inlet velocity, the pressure drop expressed by the dimensionless Euler number, and the collection efficiency captured by the cut-size model, and it presents worked numbers throughout so the reader can reproduce the design by hand. It then connects cyclone design to the wider disciplines of the library: the heat balance and gas temperatures of the preheater that set the gas properties, the raw mix design and quality control that decide the dust load and the particle density, the ball charge and separator tuning that make the mill circuit efficient, and the cost accounting that prices every Pascal of pressure drop and every ton of escaped dust. Real formulas are used throughout, consistent with the library’s culture of transparent, auditable engineering.

Why the Cyclone Is the Right Machine for Cement Dust

A cyclone is a mechanical dust collector without moving parts inside the gas path: it uses the swirl of the gas itself to fling particles to the wall, where they slide down to the outlet, while the cleaner gas spirals up and out through the vortex finder. Its virtues explain its dominance. It handles heavy dust loads without blinding; it tolerates the high temperatures of the preheater; it runs continuously with near-zero maintenance beyond erosion checks; and its capital cost per cubic meter per hour of gas is low. Its limitation is efficiency in the fine range: below a few microns its collection falls off, which is why fine finish-mill dust, though it may pass a preheater cyclone without disaster, needs a baghouse or a high-efficiency separator at the end of the circuit, where the particle sizes and the value of the product change the economics.

In the preheater, the function is gas-solid separation between stages: the hot gas rises through the tower, and each stage’s cyclone drops the entrained raw meal into the next cascade while the gas passes upward to the previous stage. The efficiency of each cyclone decides how much dust recirculates, how much material takes the long way around the tower, and how clean the gas reaches the top stage and the fan. In the cooler, cyclones capture the fine clinker dust carried by the cooler exhaust air before the air is reused as secondary or tertiary combustion air or vented. In the mill circuits, the separator cyclones, or the cylindrical-conical high-efficiency separators, split product from return fines, and their sharpness sets the quality and the economy of the grind. All of these applications obey the same equations, and the workbook is built around that unity.

The Starting Point: Volumetric Flow Q and Gas Properties

Every cyclone design begins with the gas it must clean: its flow rate and its physical properties at the operating temperature. The volumetric flow Q, in cubic meters per second, is the design traffic. In the preheater the flow at each stage derives from the mass balance of the kiln system, the fuel and combustion air, and the temperature at that stage; hot gas expands, so the operational cubic meters per second at a stage at 700 degrees Celsius is far larger than the same mass flow at 200 degrees. The gas density follows from the ideal gas relationship: density is proportional to the molecular mass of the gas, dominated by nitrogen, carbon dioxide, water vapor, and oxygen, divided by the absolute temperature. Because the density appears in both the Euler number and the cut-size model, the design sheet takes the gas temperature and composition as inputs and computes density and viscosity, rather than accepting a single cold-air number, which is the classic beginner’s error in cyclone work.

With Q and the inlet area A, the inlet velocity is Q divided by A. The inlet velocity is the master dial of the design because it drives both performance and cost. Practical preheater practice holds the inlet velocity between 15 and 25 meters per second; below the range the particles have too little momentum to reach the wall, and above the range the pressure drop, erosion, and risk of re-entrainment climb out of proportion to the efficiency gained. The workbook sweeps the inlet velocity across a design range and prints a small table of the resulting pressure drops, fan powers, and collection efficiencies, so the user chooses the velocity as an economic decision, not an inherited habit.

Pressure Drop and the Euler Number

The pressure drop across a cyclone is the toll the device charges for its service, and it is paid by the fan, hence by the power bill. The dimensionless grouping that makes pressure drop comparable across designs is the Euler number, Eu, defined as Eu equals two times the pressure drop across the cyclone times the gas density, divided by the square of the inlet velocity. In symbols, Eu equals 2 times delta-P times rho divided by v-squared. Rearranged for design, the pressure drop is delta-P equals one half times Eu times rho times v-squared. The Euler number is an empirical property of the geometry: a given cyclone shape has a characteristic Eu, often between about 3 and 8, and the workbook holds the characteristic values for the common geometric families, the tangential-inlet cyclone, the scroll-inlet cyclone, and the various high-efficiency ratios.

The quadratic dependence on velocity is the heart of the trade. Doubling the inlet velocity quadruples the pressure drop; a design that gains a few points of efficiency by racing the gas pays for it in fan power that grows with the square of the speed. The fan power is the pressure drop times the volumetric flow divided by the fan and drive efficiencies, and the workbook carries that conversion into kilowatts and, with the tariff, into dollars per year, so the design sheet ends in the same currency as the cost ledger. This single translation, from Pascals to dollars, is what turns cyclone design from an academic exercise into a commercial decision, and it is present in the file by design.

Collection Efficiency and the Cut-Size Model

Efficiency is summarized by the grade efficiency curve, which states, for every particle diameter, the fraction collected. The conventional single number that anchors the curve is the cut size, the diameter at which 50 percent of the particles are collected. The classic and still most instructive approach is the Lapple model, which estimates the cut size from the cyclone geometry and flow. In the simplified form used by the workbook, the cut diameter dc is proportional to the square root of the quantity nine times the gas dynamic viscosity mu times the cyclone body diameter D times the inlet width b, all divided by the product of the effective number of vortex turns N, the gas density rho-g, the particle density rho-p, and the inlet velocity v. Written compactly, dc is proportional to the square root of (9 mu D b) divided by (N rho-g rho-p v).

Every term in that expression carries a design lesson, and the workbook annotates each one. A smaller body diameter D reduces the cut size: the particles travel a shorter radial distance, so the same centrifugal force collects finer material; this is why several smaller cyclones in parallel collect finer than one giant cyclone of the same total throughput. A higher inlet velocity v raises the centrifugal force and reduces the cut size, subject to the pressure penalty. The particle density rho-p in the denominator shows why cement dust, at roughly 2.6 to 3.1 tons per cubic meter, is comparatively easy to collect, while soot from alternative fuels is the notorious hard case. The gas viscosity and density come from the operating temperature and composition, and the effective number of turns N is the design’s geometric signature, counting how many times the gas circles the body before it escapes up the vortex finder; longer bodies, meaning more effective turns, collect finer.

Beyond the cut size, the full grade efficiency curve in the workbook follows the standard approximation that collection rises steeply with particle diameter, steeply between roughly half the cut size and three times the cut size, capturing the vast bulk of the mass of cement dust, since most kiln feed mass sits far above even the largest realistic cut. The total mass efficiency then results from convolving the grade curve with the particle size distribution of the dust, which is why the raw mix section of this article, with its particle sizes and densities, is not decoration: the dust load and its distribution are genuine inputs to the efficiency sheet.

Cyclone Geometry: The Rules of the Species

A cyclone is defined by a handful of ratios that fix its entire geometry, and the workbook presents the standard family. The cyclone body is a cylinder on top of a cone. The tangential inlet has a width a and a height b, dimensioned from the body diameter D by the family ratios, typically inlet width around one-fifth to one-quarter of D and inlet height about half of D. The outlet gas tube, the vortex finder, projects down from the top with a diameter of about half of D and a length that determines how far the escaping vortex must travel. The cone tapers to the dust outlet, whose diameter is a small fraction of D; the cone’s angle and length set the effective number of turns and hence the efficiency and the pressure drop. The dust outlet must be isolated from the inlet region: a gas leak up the dust duct destroys efficiency, which is why every real cyclone installation is sealed and why a failed rotary valve or flapper on the dust outlet shows up immediately as a plume of dust in the stack.

The two broad families the workbook supports are the standard (high-throughput, moderate-efficiency) cyclone and the high-efficiency cyclone, the latter with a longer body, a smaller body diameter relative to flow, and consequently a smaller cut size at the price of a higher Euler number and hence more pressure drop. The user chooses the family by the duty: raw-mill and preheater cyclones tend toward the high-efficiency geometry because the entrained dust is abrasive, valuable, and hot; the simplest pre-cleaner ahead of a baghouse may accept the standard geometry because the baghouse catches the remainder. Whatever the family, the ratios are held constant internally so that scaling the cyclone to a new flow is a matter of one size factor, which is exactly how the workbook lets the user scale a validated design up or down.

Scaling and the Choice of Number of Cyclones

The design sheet’s real craft is the decision of how many cyclones and what diameter. Because efficiency favors small body diameter while the flow demands a large total inlet area, the designer partitions the gas stream among several cyclones in parallel. The workbook asks for the total volumetric flow Q and offers the design sweep: a single large cyclone versus two, four, or six smaller ones, with each layout recomputing the inlet velocity, the Euler-number pressure drop per unit, the total fan power, the cut size, and the total mass efficiency. The result is the familiar economic optimum: a moderate number of cyclones, each running at the chosen inlet velocity, balances efficiency against the capital cost of multiplicity and the pressure drop.

The gas distribution to an array matters as much as the number. Real installations suffer unequal division when the inlet duct geometry short-circuits one cyclone or when a plug builds in a manifold leg; the fan then drives most of the gas through the open leg, over-speeding it, eroding it, and starving the others. The workbook cannot fix a bad manifold, but it computes the pressure-drop equality the manifold must aim for, and the commissioning sheet includes the measurement points for the velocity or static pressure at each cyclone inlet so the plant can confirm balanced distribution before the fire is lit. This practical discipline, scale with the ratios, isolate the dust outlets, balance the distribution, measure and confirm, is the difference between a cyclone that silently works for twenty years and a tower that erodes a campaign in eighteen months.

Erosion, Materials, and the Abrasive Reality

Cement dust is abrasive, and the cyclone’s duty, flinging it to the wall at the highest velocity in the device, guarantees a fight against erosion. The wall speed peaks at the inlet transition and at the vortex-finder tip, and the design sheet flags the erosion-critical zones: the lower barrel opposite the inlet, the cone taper, and the region under the vortex finder. The material selection follows the temperature and the abrasiveness: mild steel with wear liners for moderate services, white cast iron or ceramic-lined segments at the hot and erosive points, and the same refractory and wear protection families the plant already uses inside the preheater. The workbook carries the erosion allowance guidance and the liner thickness input, and the maintenance sheet tracks the thickness gauging points so that a thin shell is detected before it becomes a hole.

The dust outlet sealing is the quiet half of erosion control. A leaking rotary valve at the dust outlet bypasses gas up the duct, at first merely wasting efficiency, and in an erosive service creates a localized sandblasting jet that can wear a hole in a campaign. The design sheet reminds the user that the pressure at the dust outlet is below the inlet region’s, so any gas short-circuit flows that way with depressing force, and it lists the standard isolations, the double flap, the rotary valve with proper pocket geometry, and the continuous level-discharge hoppers. Many of the efficiency problems that consultants are called to diagnose in a “failing” preheater trace, at the end of every probe, to a dust outlet whose seal had quietly failed, and the workbook treats that diagnostic as part of the design rather than a footnote.

Preheater Cyclones and the Heat Balance Connection

In the preheater the cyclone is inseparable from the heat balance of the tower. The gas temperature and composition at each stage set the gas density and viscosity that drive the cut-size model; the thermal efficiency of the stage decides how much heat the gas carries upward and how much the meal absorbs; and the temperature profiles that the quality and control engineers chart against design are the signature of how the cyclones, the ducts, and the dispersion of the meal are actually performing. This is where the cyclone design sheet meets the heat balance family of the library, which is built on the higher and lower heating value framework of the fuel and the complete enumeration of input and output terms around the kiln system, including the sensible heat of the hot gas at each stage, the heat absorbed by the meal, the heat of the carbonate reaction, the losses through the shell, and the heat leaving in the clinker, the cooler exhaust, and the preheater exhaust.

A poorly performing cyclone shows up in the heat balance as a colder-than-design meal at a given stage, a hotter-than-design gas leaving the top stage, and a heavier dust load circulating in the loop, all of which cost fuel because the kiln must work harder to compensate for heat the tower failed to transfer. The two workbooks therefore share their input conventions: a change in the cyclone efficiency prediction, or a retrofit to a more efficient geometry, flows through to the predicted stage temperatures, the exit gas enthalpy, the fuel consumption on the LHV basis, and finally the cost line. The integrated view, cyclone to heat balance to cost, is the library’s great teaching, and the cyclone design file carries the bridges to the heat balance sheets so the effect of any design change is visible all the way to the dollar sign.

Mill-Circuit Separation: The Same Physics, a Different Duty

The classification inside the raw mill and finish mill circuits uses the same swirl physics as a cyclone but aims at a different target: an arbitrary cut on particle size rather than all the dust possible. The high-efficiency separator, whether static, dynamic, or the modern cage-type variety, draws gas and material through a rotating distribution, and the returning coarse fraction flows back to the mill while the fine product exits to collection. The governing quantities are the same volumetric flow Q, the inlet and chamber velocities, the Euler number that prices the fan power, and a cut (or separation) sharpness that plays the role of the cut-size model in reverse: here the line between product and return is set to a specification, and the imperfection of the cut, the fraction of product-size particles that take the return path, and the fraction of oversize that leaks to product, is the headline performance number.

The connection to the ball charge is close and practical, as the library’s ball charge workbooks explain in parallel. A sharper separator cut returns less finished material, lowering the circulating load, cooling the mill’s wasted work, and shifting the balance of power from re-grinding to new surface, with the physics that a few decisions upstream, the filling degree, the grading, the ventilation, all interact with the separator. The cyclone design sheet serves the mill circuit by providing the separator velocity, Euler number, and cut behavior, and the mill audit answers whether the circuit is in its economic optimum. This is why a single plant can trace a single thread, separator vane angle, through the fan power, the circulating load, the media wear, the quality charts, and the cost of a ton of cement, and why the library’s files, cyclone, ball charge, heat balance, quality, and cost, are published as one coherent package rather than as islands.

Raw Mix, Dust Load, and Quality Control as Inputs to Design

The dust the cyclone must handle is decided upstream, in the raw mix design and quality control family. The clinker chemistry is governed by the three classical moduli, the limestone saturation factor LSF equal to CaO divided by the sum of 2.8 times SiO2, 1.18 times Al2O3, and 0.65 times Fe2O3, the silica modulus SM equal to SiO2 divided by Al2O3 plus Fe2O3, and the alumina modulus AM equal to Al2O3 divided by Fe2O3, and those moduli decide the composition, the burnability, the clinker phases, and indirectly the dust’s particle density and its behavior in the gas stream. The quality control charts of the laboratory guarantee the stability of the kiln feed, and by extension the stability of the dust load reaching the cyclones and the stability of the load the separator sees in the mill. A plant whose raw mix wanders sees the consequences in the preheater as drifting dust loads and in the mill as drift in circulating load and product fineness, and the cyclone design, though nominally fixed by geometry, is actually being re-qualified every shift by the stability of everything upstream of it.

The designer’s habit, taught by this workbook, is to design for the worst realistic combination of dust load, temperature, and distribution, then to verify at the normal case, and then to flag the instruments that will confirm in operation, so that the assumptions made at the desk are checked against the plant in the first campaign. This is the culture of the package: every input documented, every calculation transparent, every design decision traceable to a measurement that can be made. The clerk in the lab, the operator at the panel, and the accountant at the desk all stand on the same evidence, which is the quiet contract this library signs with every plant that uses it.

A Full Design Worked in Numbers

To commit the method to the reader’s hands, follow one design through the sheet. A preheater stage must handle a volumetric flow Q of 90 cubic meters per second of gas at 650 degrees Celsius, where the gas density is about 0.38 kilograms per cubic meter and the dynamic viscosity about 3.6 times 10 to the minus 5 Pascal-seconds. The chosen family ratio gives an inlet width of 0.22 times D and an inlet height of 0.55 times D, so the inlet area is 0.121 times D squared. With three cyclones in parallel each taking 30 cubic meters per second and a target inlet velocity of 20 meters per second, each cyclone needs an inlet area of 1.5 square meters, so 0.121 times D squared equals 1.5 and the body diameter D comes to about 3.5 meters. Entering the geometry, the effective number of turns N is about 4.5 for the family’s body length, and the particle density of kiln feed dust is about 2.7 tons per cubic meter.

The cut size follows from the cut-size proportionality: dc is proportional to the square root of (9 times 3.6E-5 times 3.5 times 0.77) divided by (4.5 times 0.38 times 2700 times 20); with the family constant calibrated, the math resolves to a cut size of the order of a few tens of microns, and the grade curve built through it captures nearly all the mass of the raw meal. The pressure drop comes from the Euler number: with the family’s Eu near 5, delta-P equals half times 5 times 0.38 times 400, about 380 Pascals per cyclone, and the fan must also pay the duct losses. The fan power for the whole stage is the total pressure rise times the 90 cubic meters per second divided by the fan and motor efficiencies, and at the local tariff the workbook converts that into dollars per year so the design’s appetite is stated in the management’s own language. Every one of these numbers is recomputable in the sheet, which is the entire spirit of the exercise.

Frequently Asked Questions

Why is there an optimum inlet velocity and not a maximum?

Because efficiency rises with inlet velocity but pressure drop rises with its square. Beyond the practical band, roughly 15 to 25 meters per second in preheater cyclones, the added efficiency is tiny while the fan power, erosion, and re-entrainment risks climb rapidly. The economic optimum therefore sits inside the band where the marginal efficiency is not worth the marginal power.

What is the difference between the Euler number and the cut size?

The Euler number captures the pressure drop: it is twice the pressure drop times the gas density, divided by the square of the inlet velocity, and it prices the fan power. The cut size captures efficiency: it is the particle diameter captured at 50 percent, set by the geometry, flow, and particle properties. Both are used together, because the whole design is a trade between the cut you need and the pressure you can afford.

How many cyclones should I use in parallel?

Sweep the number in the design sheet: fewer, larger cyclones give lower capital and lower pressure drop but coarser cut sizes, while more, smaller cyclones give finer collection at higher pressure and multiplicity cost. Choose the number that hits the required cut size at the minimum total of fan power and maintenance cost, and verify the gas division between legs at commissioning.

Why does a leaking dust outlet destroy efficiency?

Because the dust outlet region sits at a lower pressure than the cyclone inlet, a leaking seal or valve creates a gas short-circuit flowing up the dust duct, dragging collected dust back into the stream and disturbing the vortex. The result is a sharp drop in collection efficiency and, in abrasive service, a localized erosion jet. Sealing the dust outlet is as much a design requirement as the body itself.

Can I use the same sheet for preheater cyclones and mill separators?

Yes, because both obey the same swirl physics, the volumetric flow, the Euler number, and the cut-size model. The preheater sheet optimizes total mass collection at temperature; the separator sheet controls an arbitrary cut for product quality. The workbook carries both modes so the user does not re-learn the physics between the kiln tower and the mill circuit.

Summary

The cyclones design workbook and this companion article together carry the complete engineering of the cement plant cyclone: the volumetric flow Q and the gas properties at temperature, the geometry of the family, the inlet velocity as the master dial, the pressure drop through the dimensionless Euler number, the collection efficiency through the cut-size model, and the practical economics of fan power translated all the way to the cost ledger. The article has connected the cyclone to the heat balance of the preheater, where stage temperatures and exit gas enthalpy respond to cyclone performance, to the mill circuits and ball charge, where separator sharpness and circulating load respond to the same physics, and to the raw mix design and quality control that decide the dust load and its stability. It has worked through a full design in numbers and through the operational disciplines of balanced gas division, isolated dust outlets, and erosion protection that keep a design honest for decades.

The enduring lesson of the cyclone is that nothing is free and everything connects. Every Pascal of pressure drop and every micron of cut size is an economic decision, and every design choice echoes through the heat balance, the mill, and the cost sheet. The engineer who masters the volumetric flow, honors the Euler number, respects the cut-size model, and above all closes the loop from the drawing to the measured plant, is the engineer who keeps the tower efficient, the fans cheap, and the dust where it belongs. That is the standard this library teaches, and the cyclones design workbook is its precise, auditable instrument.

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.