Preheater Cyclone Design: Full Technical Guide

Subtitle: Advanced Preheater Cyclone Engineering for the Second Tower Stage — Second Edition

Within the tall concrete skeleton of every preheater tower, a set of seemingly identical steel cones rests one above the other, and to the casual eye they look like the same machine repeated. They are not. Each stage of the preheater operates at its own temperature, carries its own gas volumetric flow, deals with its own dust load, and imposes its own separation duty, and the stage that is the true referee of the entire tower is the second one. The cyclones design file (cyclones design.zip) in the cementequipment.org package collects the working engineering of preheater cyclones, and this article is the complete technical companion, written specifically around the design of the cyclone pair that sits in the second tower stage, the stage where the gas has already climbed from the kiln and where the meal is beginning its real heating. This second-edition article develops the subject to a level well beyond the introductory treatment: it derives the governing aerodynamics of the swirling flow, the vortex finder and the inner vortex, the significance of the inlet velocity and the Euler number for the pressure drop, the cut-size model and the grade efficiency curve for the collection, the laws of scaling that let the designer move a validated cyclone from one stage to another, the hazards of gas short-circuiting and dust re-entrainment that are most acute in exactly this part of the tower, and the operational and maintenance disciplines that keep a second-stage cyclone honest for the life of the plant.

The second tower stage deserves its own treatment because it sits at the heart of the thermal cascade. The gas entering the second stage cyclones has already been cooled by the meal in the first stage, but it is still hot, typically of the order of 700 to 900 degrees Celsius at the cyclone inlet in a precalciner kiln, it carries a substantial dust load because the meal feed to the first stage is dense, and it must be cleaned well enough that the gas running upward to the first stage does not carry so much meal that the top of the tower overloads, recirculates, and drags down thermal efficiency. The second stage is also where the pressure drops of the system begin to be earned, because the cyclones, and the pressure they cost, are part of the total draft balance that the fans must buy. An engineer who can design, or at least evaluate, a second-stage cyclone correctly holds one of the true passports to understanding the preheater as a whole, and this article aims to put the reader in that position, with numbers that can be reproduced by hand and with the physical reasoning that makes them stick.

The Preheater Cyclone and the Role of the Second Stage

Begin with the place of the machine. A suspension preheater works by cascading the meal downward against an upward gas stream, and at every cyclone the two streams separate: the gas turns upward into the next higher stage and the meal falls through the stage’s downpipe into the gas of the stage below. The cyclones are therefore both heat exchangers, in the sense that the gas-solid contact happens as the meal is dispersed in the gas en route to and through each cyclone, and dust collectors, in the sense that each must return the meal to the process rather than let it escape upward. The second stage from the top receives its meal feed by gravity from the first stage cyclone downpipe and its hot gas from the stage below, and it hands its separated meal down to the third stage while the gas it has cleaned climbs to the first. Its inlet temperature and its gas volume therefore fix a large part of its design, and those values are set by the overall thermal balance of the tower: the hotter the gas and the more the meal has already been heated at the stage above, the larger the actual cubic meters per second the cyclone must handle, because hot gas expands.

Why is the second stage special? Three reasons. First, it is a transition stage: the heat transfer at this elevation in the tower is still strongly convective, so the meal is relatively cold and the gas relatively hot, and the properties of the gas, density, viscosity, velocity, all derived from temperature, are in a regime that is neither the cold top stage nor the near-calcination bottom stage. Second, the dust load at the second stage is at its most punishing, because the first stage above it is fed with fresh cold meal, and the meal cascading down is dense, so the second-stage cyclone must separate a heavy burden of hot, sticky-meal-laden dust from gas, with any inefficiency immediately visible as top-stage or fan overloading. Third, the second stage is where pressure and erosion both peak relative to the upper tower: the gas velocity is still high, the dust is abrasive, and the pressure drop of the stage contributes to the total that the induction fan must create. Getting the second stage right is therefore not one problem among many; it is, in many cases, the single most consequential design decision at the upper end of the tower.

The Swirling Flow: Outer and Inner Vortex

The cyclone achieves separation by turning the gas into a vortex. Gas enters the body tangentially through the inlet duct, and the shape of the cyclone, the cylinder rotating into a cone, constrains the flow into a spiral, called the outer vortex, which descends toward the apex. At the apex, near the dust outlet, the spiral reverses and flows back up the center of the cyclone as a second, inner vortex, which exits through the vortex finder, the central tube that projects down from the top of the cyclone. The particles, being denser than the gas, are flung by the centrifugal acceleration of the rotating flow toward the wall of the outer vortex, where they spiral down with the boundary layer and exit through the dust outlet. Two consequences follow from this structure. The first is that the finest, most fragile separation happens not at the wall but near the center: particles that are swept into the inner vortex, or that are re-entrained off the wall by the up-flowing core, escape out the vortex finder in the overflow. The second is that the entire internal flow is intensely turbulent and asymmetric, which is why cyclone calculations are empirically anchored: the flow is far too complex for simple laminar analysis, and the design is built on measured geometric families and validated correlations.

Several classic geometry families embody the empirical knowledge. The Stairmand high-efficiency cyclone, with its long body, relatively narrow inlet, and long vortex finder, achieves fine collection at the price of higher pressure drop. The Lapple geometry, broader and shorter, trades efficiency for throughput and lower pressure. The tangential-entry plus scroll-inlet designs, in which the gas is introduced through a full or partial scroll, reduce the inlet losses and the local turbulence peak and are common in the biggest preheater applications, where pressure drop is a significant operating cost. For cement duty the practical design space is narrower than the academic one: the cyclones must handle very high temperatures, heavy loads of abrasive, sometimes sticky dust, and frequent thermal and erosive insult, so the geometry chosen is robust rather than fragile, and the fine-tuning is done with ratios of the key dimensions to the body diameter. The displacement of the work from the textbook Stairmand coordinates to a plant-resistant set of proportions is one of the things the experienced designer learns, and it is why the design file’s ratio tables are used more than its raw formulas.

Design Primaries: Flow, Inlet Velocity, and Geometry Ratios

Every cyclone design begins with the volumetric flow Q, in cubic meters per second, which at the second stage is the gas mass flow from the stage below converted at the local temperature and pressure. The gas density follows from the ideal gas law: density is proportional to the molecular weight of the gas divided by the absolute temperature, so at 850 degrees Celsius a flue gas of density about 1.3 kilograms per cubic meter at ambient has fallen to roughly 0.34 kilograms per cubic meter, while its volume has expanded about fourfold, and the dynamic viscosity has increased with temperature, changing both the pressure drop and the collection. The designer therefore never works with cold-air numbers at this point in the tower; the temperature, composition, and resulting properties of the gas at the stage are genuine inputs, and the design file tabulates them so that a single input of temperature and gas analysis yields the density, viscosity, and volume used everywhere downstream.

The master dial is the inlet velocity, the gas velocity through the cyclone inlet area. In preheater practice the inlet velocity is held in a band, typically between about 15 and 25 meters per second, with 18 to 22 being the most common working range: low enough that pressure drop and erosion stay under control, high enough that collection is effective. Below the band, the centrifugal acceleration is too weak, the cut size coarsens, and efficiency collapses; above the band, pressure drop grows with the square of velocity, erosion accelerates as approximately the cube or higher power of velocity, and the risk of dust re-entrainment from the wall rises, so the efficiency gained per unit of added velocity shrinks to nothing. The inlet area is then fixed by dividing Q by the chosen inlet velocity, and the geometry ratios convert that inlet area into the cyclone dimensions. The standard family used throughout the preheater places the inlet width in the range of 0.18 to 0.22 times the body diameter D and the inlet height about 0.4 to 0.6 times D, but these are starting points, not laws, and the work of design is the trade, not the copy.

Pressure Drop and the Euler Number

Nothing about a cyclone is free, and its fee is the pressure drop, the resistance the device imposes on the gas, which the induction fan must overcome and which therefore appears on the power bill and, translated through the tariff, on the cost statement. The dimensionless number that lets different designs be compared is the Euler number, Eu, defined as Eu equals two times the pressure drop times the gas density divided by the square of the inlet velocity; rearranged, the pressure drop delta-P equals one half times Eu times rho times v-squared. The Euler number is an empirical property of the geometry: a well-proportioned tangential-inlet cyclone has an Eu characteristically of about 4 to 6, a high-efficiency long-bodied cyclone more like 6 to 9, and a thoughtfully designed scroll-inlet cyclone can bring it toward the lower end of the range. Two consequences follow that every operator knows. Doubling the inlet velocity multiplies the pressure drop by four, so a small change in velocity is a large change in fan power; and the pressure drop is what balances the tower, because the gas will take the path of least resistance, and cyclone designs that differ in Eu, or installation legs that differ in plugging or throttle, divide the gas unevenly, a failure mode the section on distribution addresses.

The fan power that pays the toll is the volume flow times the pressure rise, divided by the fan and drive efficiencies: for a stage handling 100 cubic meters per second at a pressure drop of, say, 1,000 Pascals, the ideal gas power is 100 kilowatts before fan and motor inefficiencies, and at a plant tariff of, say, $0.10 per kilowatt-hour that is a six-figure annual bill for one stage. The design file computes this transition explicitly, from Pascals and cubic meters to kilowatts and then to dollars per year, because it is the discipline that binds the engineering and the commercial views of the same number. It is a defining habit of this library: every Pascal is translated into money, so that the design conversation, vane angle, velocity, number of cyclones, is held in the language the plant’s management actually speaks. A cyclone that “looks efficient” in isolation but costs $150,000 a year in fan power against a cheaper slightly coarser competitor is not efficient at all, and the file teaches the reader to make that judgment with the same arithmetic the accountant uses.

Collection Efficiency and the Cut-Size Model

The purpose of the cyclone is collection, and collection is summarized by the grade efficiency curve, the fraction collected as a function of particle diameter, plus a single anchoring number, the cut size, the diameter at which 50 percent of the particles are collected. The classical model that anchors the thinking is the Lapple cut-size expression, which, in the form used by the design file, states that 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, the particle density, and the inlet velocity. Every term in that expression is a design lever, and the designer reads it like a sentence. Smaller body diameter D collects finer, because the particle has a shorter radial distance to travel. Higher inlet velocity v raises the centrifugal force, and finer collection, but at the price shown by the Euler number. The particle density in the denominator shows why kiln feed dust, at a density of roughly 2.6 to 2.8 tons per cubic meter, collects readily, while light soot or volatile sublimates are the hard cases. Longer bodies, more effective turns, collect finer; and viscosity, which rises with temperature, is the quiet enemy that coarsens the cut as the gas heats up.

Beyond the cut size of the model lies the real collection of the stage: the grade efficiency curve, which in well-designed cyclones rises steeply with particle diameter through and beyond the cut size, so that particles somewhat smaller than the cut are largely collected and particles substantially larger are essentially all collected. The total mass efficiency of the second stage is then the convolution of that grade curve with the particle size distribution of the entering dust, and here the second stage reveals its true burden: the dust it receives includes the fresh meal cascade plus any material that the first stage failed to catch, and if the first stage is inefficient, the second stage sees even more of the fine end. A useful engineering truth is that the cyclone’s mass efficiency is dominated by the coarsest-class fractions because they carry the mass, so a second-stage cyclone with a modest collection efficiency on the mass basis, say in the high 90s of percent, can still let a noticeable amount of fine material through, which then recirculates in the tower. The stability of the whole tower, and its resistance to the classic self-destructive dust cycle, depends precisely on keeping that fine leakage low, which is why cut size, not just mass efficiency, is the number the best designers watch at this stage.

Scaling: Moving a Design Up the Tower

The most economical way to design a preheater cyclone system is to design one proper cyclone and scale it, and scaling is governed by the ratios of the design file. A cyclone’s proportions, inlet width ratio, inlet height ratio, cylinder height, cone length, vortex finder diameter and length, dust outlet diameter, are all expressed as fractions of the body diameter D, so that if the ratios are held constant, the entire geometry is defined by D alone. When the cyclone is scaled to a new volume flow, the designer holds the inlet velocity, and therefore the Euler number, pressure drop, and the physics of the flow, constant, and the body diameter grows with the square root of the flow: double the flow and the diameter grows by about the square root of two. Collection, however, does not scale neutrally: because the cut size grows with the square root of D at constant velocity, a bigger cyclone collects coarser, which is precisely why the tower uses multiple cyclones in parallel rather than one gigantic one per stage, so that each cyclone keeps a modest diameter and a fine cut while the total flow is partitioned among them.

Scaling from one stage to the next is not a straight copy, because the temperature, and hence the gas density and viscosity, differ, and because the dust load differs. The proper procedure, which the design file automates, is: take the validated design at its home stage; compute the gas properties at the target stage temperature; recompute the volumetric flow at that stage from the mass flow and the new density; scale the body diameter to restore the chosen inlet velocity; and then recompute the cut size and pressure drop, checking that the efficiency and the draft are still acceptable. This cascade of calculations is where a stage-by-stage design actually lives, because each stage is a little different even though all share the same geometry family. The file lays this out as a sequence of linked sheets, so the user can pull the second stage down, change the temperature, and watch the diameter, the cut, and the pressure all propagate, exactly as the plant’s own data will do at commissioning.

The Numbers of the Second Stage Worked by Hand

To make the method concrete, work one design in round numbers that represent a realistic second-stage duty. Suppose the mass flow of gas to the stage is 150,000 standard cubic meters per hour and the inlet temperature is 850 degrees Celsius, giving a gas density of about 0.38 kilograms per cubic meter and a volumetric flow of about 105 cubic meters per second. Choosing an inlet velocity of 20 meters per second, the required inlet area is 105 divided by 20, about 5.25 square meters. With the family ratios fixing the inlet width at 0.2 times D and the inlet height at 0.5 times D, the inlet area is 0.1 times D-squared, so D-squared is about 52.5 and D comes to roughly 7.2 meters, large enough that the stage is almost certainly split into two parallel cyclones, each about 5.1 meters in diameter, each drawing about 52.5 cubic meters per second through an inlet area of about 2.6 square meters. With the body length ratios of the family, the effective number of turns is of the order of 3 to 4, and with a particle density of 2,700 kilograms per cubic meter and a gas viscosity of about 3.6 times 10 to the minus 5 Pascal-seconds, the Lapple-type cut size resolves to the order of a few microns at the family’s conditions, small enough that the mass efficiency of the kiln feed with its broad distribution sits comfortably above 97 percent.

The pressure side falls out with the Euler number: with an Eu of about 5 and an inlet velocity squared of 400, the pressure drop is one half times 5 times 0.38 times 400, about 380 Pascals per cyclone, and with two cyclones in parallel and the associated ductwork, the stage total is in the low hundreds of Pascals above the duct losses. The fan power for the whole stage gas is the 105 cubic meters per second times the total pressure rise divided by efficiencies, which lands in the region of 80 to 120 kilowatts of gas power, and across the tariff that becomes the annual sum the file prints. These numbers are presented as engineering illustrations, not as the answer for any specific plant, because the exact values depend on the real geometry and the real family constants, but they show the shape of the problem: modest diameters, moderate pressures, high efficiencies, and a design that is driven by one decision, the inlet velocity, and audited by two numbers, the cut size and the pressure drop. Reproducing this calculation in the file, with the plant’s own data, is the entire discipline of the exercise.

Gas Distribution, Short-Circuiting, and Re-entrainment

The design file devotes real attention to the failure modes, because in practice most cyclone problems are not design problems but flow problems. The first is distribution: when several cyclones serve one stage, the gas must divide equally or in the design ratio, but unequal inlet duct geometry, a throttling damper, or a partial plug in one leg will make one cyclone carry too much flow and the other too little. The overloaded cyclone over-speeds, erodes, and its efficiency is disturbed, while the underloaded one under-performs at its own cut; the effect on the stage as a whole is exactly the leakage and recirculation penalty discussed above. The design file therefore includes the pressure-drop equality the manifold must target and, critically, the commissioning discipline of measuring the static pressure and velocity at each cyclone inlet and balancing the distribution before continuous operation. The second failure is short-circuiting: dust that has been collected at the wall can be picked up again by the high-speed gas spiraling near the wall, particularly at the vortex finder entry and in the cone, and dragged straight out; this re-entrainment is why a cyclone must not be run far above its design velocity and why the dust outlet seal matters so much, since a leaking rotary valve at the apex admits gas that locally disturbs the flow and re-entrains the very dust the machine just collected.

The third failure is leakage at the dust outlet, which deserves its own emphasis because it is the most common silent killer in the sequence. The dust outlet sits at a lower pressure than the cyclone inlet region, so any defect in the apex valve, double flap, rotary valve pocket, or sealing of the downpipe creates a gas short-circuit that flows back into the cyclone, disturbs the inner vortex, and drags collected material upward. In hot, erosive service the result is a localized sandblasting jet that can wear a hole in the shell in a single campaign, and many a mystery of a “failing” preheater has been solved by finding a burned-through apex flap. The design file lists the standard isolations, the double-flap gate, the rotary valve with proper pocket geometry and a venting arrangement, and the continuous discharge hoppers, and it treats the apex seal as a design element, not an accessory. The operational rule that emerges is simple and worth engraving: a skilled operator watches the apex seals of the second stage like a hawk, because at this dust-heavy elevation the cost of seal failure is not an inconvenience, it is the efficiency of the tower.

Erosion, Materials, and Thermal Duty at the Second Stage

The second stage is a hot and erosive environment. The gas is high velocity, the dust is abrasive, and much of the work happens behind an outer concrete shell where access is difficult and expensive. The erosion-critical zones are familiar to any maintenance engineer: the inlet transition where the gas slams into the curved wall, the region directly opposite the inlet entry, the lower barrel above the cone, and the cone taper near the apex, all see the highest velocities and the heaviest particle flux. The material strategy follows the duty: mild steel with a sacrificial thickness margin and changeable wear liners in moderate zones, chromium white iron or composite ceramic wear segments at the worst points, and for the very hottest and most erosive positions, the same refractory and oxide-bonded ceramics families the plant already uses in the preheater. The design file carries the erosion allowance guidance and the liner thickness input, and the maintenance sheet provides the thickness gauging points and the inspection schedule so that a thinning shell is detected, in the engineer’s phrase, before it is felt.

Temperature at the second stage, while below the near-calcination heat of the lower stages, is still high enough to matter for material selection and for structural behavior: the shell experiences thermal expansion that the support steelwork and insulation must accommodate, and the hot face of any refractory protection must be matched to the cyclic duty of kiln start-ups and stoppages. The cyclone at this elevation will also see, in the normal variations of the process, excursions in temperature if the kiln or calciner runs hot, and the design must tolerate them without spalling refractory or distorting the shell. Good insulation on the outside not only protects personnel and reduces heat loss to the environment, which is real energy in a plant audited for thermal efficiency, but also keeps the steel above the acid dew point on cold-run days. The second-stage cyclone, in short, is a pressure-retaining, heat-retaining, dust-handling structure as much as it is an aerodynamic device, and the design file treats those three functions together, because failure in any one of them fails the whole stage.

Linking the Second Stage to Tower Control

The second stage links to the wider process through the instruments and controls that read the tower, and this is where the design file connects to the heat balance and process control cultures of the package. The temperature of the gas leaving the second stage, like the temperature profile up the tower, is the signature of how well the stages are transferring heat and separating material. A rise in the differential pressure across the second stage indicates either a rising gas flow or a developing plug; a fall in the stage temperature combined with a rise in the top-stage pressure drop suggests meal recirculating instead of cascading, the classic leakage symptom; and the gas analysis at the tower exit, CO and oxygen, feedbacks through the control logic to the kiln fuel and the draft. Modern plants read these trends continuously, and expert control systems, kiln expert systems built on process models, use the stage temperatures and pressures to nudge the draft, the calciner firing, and the kiln feed rate within protective envelopes.

The quantitative connection to the heat balance is direct. The enthalpy leaving the second stage upward is a function of its exit temperature, and the enthalpy the meal absorbs in its passage through the second stage is a function of the thermal contact; an inefficient or plugged second-stage cyclone changes both, and the kiln must burn more fuel to compensate for the heat the tower failed to transfer. The mass connection is equally direct: the fine dust that escapes the second stage into the first, and from there toward the fan and the filter, is kiln feed that is not reaching the kiln, material that must be re-cycled, and dust that ages in the gas stream. The design file’s input conventions are chosen so that a change in cyclone efficiency, or a retrofit to a more efficient geometry, flows all the way through to the predicted stage temperatures, the exit gas enthalpy, the fuel consumption on the lower heating value basis, and the cost line. This integrated view, cyclone to heat balance to economics, is the package’s great teaching and the reason the cyclones file lives in the archive with the balance sheets rather than alone.

Retrofit, Trouble-Shooting, and the Diagnostic Habit

Very many cyclone installations are not greenfield designs but retrofits and trouble-shooting jobs, and the design file serves these too. The classic diagnostic sequence for a second-stage problem is simple to state but demands discipline to perform: establish the facts, confirm the distribution, inspect the seals, and only then suspect the geometry. The facts are the stage temperatures, the differential pressures, the dust load arriving and the dust escaping to the filter, the particle size distribution of the feed and the overflow; the distribution is confirmed by the inlet velocity or static pressure measurements at each leg; the seals are inspected by the classic probe, comparing the pressure at the apex with that in the downpipe; and the geometry is questioned only after the first three are ruled out, because in the overwhelming majority of real cases the fault is a plug, a leak, or a distribution problem, not the shape of the cone. Re-entrainment at the vortex finder, an over-speed due to a throttle change, or a build-up of sticky meal in the cone are each diagnosed by their own signatures, and the file organizes those signatures into a structured walk-through.

Retrofits follow the same scaling logic as new design: the goal is usually either a finer cut at the same pressure, achieved by going to a more efficient geometry or adding a small increase in velocity, or a lower pressure drop at the same cut, achieved by scroll entry or optimized vortex finder geometry, or a capacity increase, achieved by careful rescale. Each retrofit is evaluated against its effect on the other stages, because changing one cyclone changes the distribution in the whole tower, and the file’s linked-sheet approach forces that evaluation to be made rather than skipped. The diagnostic habits this section teaches, measure, confirm, inspect, then change, are exactly the standard this library holds, because plants that measure their way to the root cause stay cheap, while plants that guess their way through a campaign of part-replacements stay poor. The cyclones design file is, in the end, a curriculum in that accuracy as much as a set of formulas, and the second tower stage is where the lesson bites hardest.

Reference Data for the Second-Stage Design

The table below collects the typical working values and geometry ratios that anchor a second-stage preheater cyclone design. They are presented as starting points for a realistic duty, to be confirmed with the plant’s actual gas analysis, temperature, and the validated family constants of the design file.

Design Variable	Symbol	Typical Value / Range	Notes
Stage inlet gas temperature	T	~700-900 °C	Sets density, viscosity, volume
Inlet gas density	ρ	~0.30-0.40 kg/m³	Ideal gas at stage temperature
Inlet velocity	v	15-25 m/s; design ~20 m/s	Master dial: efficiency vs pressure
Euler number	Eu	~4-6 (tangential), 6-9 (high-efficiency)	Delta-P = 0.5 · Eu · ρ · v²
Inlet width ratio	b/D	~0.18-0.22	Family ratio
Inlet height ratio	a/D	~0.4-0.6	Family ratio
Vortex finder diameter ratio	Dv/D	~0.4-0.55	Sets inner vortex and pressure
Effective number of turns	N	~3-5	Geometry signature; longer body = more turns
Particle density (kiln feed)	ρp	~2,600-2,800 kg/m³	Denominator of cut-size model
Stage collection efficiency	η	>97-98% mass basis	Convolved with feed PSD

Frequently Asked Questions

Why does the second tower stage deserve its own design treatment?

Because it is the transition stage where the gas is still hot, the dust load is at its heaviest, and the temperature and flow regime are neither the cold top nor the near-calcination bottom of the tower. Its efficiency and pressure drop set the top-of-tower dust cycle and feed directly into the fan power bill, so its design is the most consequential single decision at the upper end of the preheater.

What is the difference between the outer and inner vortex in a cyclone?

The outer vortex is the descending spiral of gas near the wall that carries the particles to the dust outlet, and the inner vortex is the ascending central spiral that exits through the vortex finder. Separation happens when particles are flung to the wall by the outer vortex; any particle that reaches the up-flowing inner core escapes with the cleaned gas, which is why the vortex finder and the central flow govern the finest cut.

Why does the cut size grow when the cyclone gets bigger?

Because at constant inlet velocity the particle in a larger cyclone must travel a longer radial distance to reach the wall, so the same centrifugal force collects coarser material. This is why preheater stages use several cyclones in parallel instead of one giant cyclone, and why scaling a validated design up and down is done by holding the velocity and ratios constant.

What is the first thing to check when a second-stage cyclone is underperforming?

Not the geometry. The first checks are the hard facts: the gas distribution between parallel legs, the condition of the dust outlet seals, and the differential pressures and temperatures, because most underperformance is caused by a plug, a gas leak at the apex, or an unequal flow split, not by the shape of the cone. Inspect the geometry only after ruling out the flow faults.

Why is the pressure drop so expensive on a hot stage like the second one?

Because the gas at 850 degrees Celsius has expanded to roughly four times its cold volume, so every Pascal of pressure drop is paid on a much larger volume flow, and because fan power is volume times pressure divided by efficiency. The Euler number, quadratic in inlet velocity, means a small velocity increase quadruples the pressure cost, which is why the inlet velocity band is respected so carefully.

Summary

This article has developed the complete engineering of the preheater cyclone with the emphasis on the second tower stage, as delivered in the cyclones design file of the cementequipment.org package. It placed the second stage in the thermal cascade and showed why its temperature, dust load, and duty make it the referee of the tower. It walked the aerodynamics of the outer and inner vortex, the geometry families and their ratios, the volumetric flow and gas properties that anchor the design, and the inlet velocity as the master dial. It derived the pressure drop through the Euler number and translated it to fan power and money; it established the collection through the cut-size model and the grade efficiency curve; it set out the scaling rules that let one validated design populate the whole tower; and it worked a realistic second-stage design in numbers from temperature to cut size to dollars. It then examined the operational realities, the gas distribution, the apex seal, short-circuiting and re-entrainment, the erosion and material strategy, the link to tower control and the heat balance, and finally the disciplined diagnostic habit that keeps a preheater healthy for decades.

The enduring lesson of the cyclone, at the second stage as everywhere, is that nothing is free and everything is connected. Every Pascal of pressure is a kilowatt on the fan bill; every micron of cut is a gram of dust recirculating around the tower; every degree of temperature is a line in the heat balance and a conversion in the cost account. The engineer who holds the flow, honors the Euler number, respects the cut-size model, seals the apex, balances the distribution, and verifies the assumptions against measured plant data is the engineer who gives the tower its efficiency and the plant its margin. That is the standard this library teaches, and the cyclones design second edition 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.