65–85 %
Realistic collection, biomass fly ash
Technology · Mechanical collection
Multicyclone (multi cyclone dust collector)
A multi cyclone dust collector is a bank of small parallel cyclone tubes separating particulate by centrifugal force alone. On biomass fly ash it realistically collects 65–85 %, not the 90 %+ quoted on coarse test dust, at 80–120 mmWC pressure drop paid in fan power. Right tool for coarse-ash and ember knock-out; wrong tool for meeting a stack limit.
80–120 mmWC
Pressure drop
≈ 7 kW per 10 mmWC
Fan power cost at design-basis flow
≈ 5–10 µm
Typical tube cut diameter d₅₀
01 — Principle
What a multicyclone is and how it separates dust
Many small cyclones beat one big one — up to a hard physical limit.
A multi cyclone dust collector packs tens to hundreds of small cyclone tubes — typically 150–300 mm in diameter — into a single casing sharing one gas inlet, one clean-gas outlet and one hopper. In each tube, axial swirl vanes spin the incoming gas; centrifugal acceleration drives particles outward to the tube wall, where they slide down into the hopper while cleaned gas reverses up the vortex core and leaves through the central outlet tube. The reason for many small tubes instead of one large cyclone is geometry: centrifugal acceleration scales inversely with the radius of the spin, so a 200 mm tube separates far finer dust than a 2 m cyclone at the same pressure drop.
Everything a multicyclone does well and badly follows from one fact: the only separating force is particle inertia. No electrical field, no filter media, no moving parts — which makes it cheap, temperature-tolerant and nearly indestructible — and no mechanism at all that acts on particles too small for inertia to matter. In Arrow Energy's gas-cleaning technology range the multicyclone is the entry point of the mechanical family whose limits led to the Electrocyclone, the proprietary hybrid that adds an electrostatic force where the centrifugal one runs out.
On boiler duty a multicyclone almost always sits as a first stage — coarse-ash knock-out and ember quench ahead of an electrostatic precipitator or bag filter — or as the sole collector on legacy plants permitted before modern particulate limits. Understanding exactly what it can and cannot collect is the point of this page.
02 — Cut diameter
The d₅₀ cut diameter: where cyclone physics draws the line
One number predicts everything the efficiency brochure leaves out.
The performance of any cyclone tube is summarised by its cut diameter d₅₀ — the particle size collected with exactly 50 % probability. The classical Lapple relation gives it as d₅₀ = √( 9 µ W / ( 2π N vᵢ (ρₚ − ρ𝗀) ) ), where µ is gas viscosity, W the inlet width, N the effective number of vortex turns, vᵢ the inlet velocity and ρₚ the particle density. For well-designed multicyclone tubes on hot flue gas, d₅₀ lands around 5–10 µm. Above the cut, collection climbs steeply; below it, collection collapses — and no adjustment of tube count changes that, because d₅₀ is a per-tube property.
Grade efficiency around the cut follows the familiar Lapple curve, η(d) = 1 / (1 + (d₅₀/d)²). The table shows what that means for a tube with d₅₀ = 5 µm:
| Particle size | Collection efficiency | Relevance on biomass fly ash |
|---|---|---|
| 40 µm | ≈ 98 % | Coarse ash and sand — collected reliably |
| 20 µm | ≈ 94 % | Upper end of char carryover — collected |
| 10 µm | ≈ 80 % | Transition zone — partial collection |
| 5 µm | 50 % | The cut itself — a coin toss per particle |
| 2.5 µm | ≈ 20 % | Fine char and ash — mostly passes |
| 1 µm | ≈ 4 % | Effectively uncontrolled |
Why do quoted multicyclone efficiencies of 90 %+ mislead?
Because overall efficiency is the grade-efficiency curve weighted by the particle-size distribution of the test dust — and quoted figures are typically measured on coarse dust with a mass-median size of 20 µm or more, where the curve sits above 94 %. Real biomass fly ash carries a heavy fine-char fraction below the cut, dragging true overall collection down to 65–85 %.
The practical rule: an efficiency figure without a stated particle-size basis is not a performance claim, it is an advertisement. When comparing offers, ask for the grade-efficiency curve and compute overall collection against your own ash analysis — or at minimum insist the quoted figure states the mass-median diameter and the fines fraction of the dust it was measured on. Guaranteed figures are stated per project after the technical assessment, on a stated basis: mg/Nm³, reference O₂, dry or wet, and load range.
03 — Service reality
Wear, plugging and re-entrainment: how field performance decays
A multicyclone rarely fails outright; it quietly stops collecting.
Tube erosion. Separation happens by driving abrasive particles against the tube wall at high tangential velocity — the mechanism is the wear. Bagasse fly ash is silica-rich; rice-husk ash is the extreme case at 85–90 % amorphous silica, and rice-husk-fired multicyclones can wear through tube walls and swirl vanes in a few seasons. Worn tubes and opened clearances degrade the vortex, the effective d₅₀ coarsens, and efficiency drops with no visible external symptom until the stack shows it.
Hopper in-leakage and re-entrainment. The vortex core of each tube extends down toward the hopper, and the static pressure there sits below ambient. Any leak — a worn rotary valve, a warped access door, a flange gasket — draws air up through the collected dust, and the vortex lifts already-captured material straight back out of the outlet tube. A modest hopper leak can erase 10–20 percentage points of collection while every tube remains mechanically perfect. Dust-valve integrity and hopper-level discipline are worth more than any internal refinement.
The monitoring answer is simple and cheap: trend outlet dust against pressure drop. A healthy multicyclone holds a stable ΔP in its 80–120 mmWC band with a stable outlet; falling ΔP with rising emission points to worn or plugged tubes forcing flow maldistribution, while stable ΔP with rising emission points to hopper in-leakage. An annual isokinetic outlet measurement at stated reference conditions — mg/Nm³ at 6 % O₂ dry — costs a fraction of one season of quietly lost collection.
Plugging. Small tubes plug: sticky ash bridges the dust outlet at the tube tip, moist start-up conditions cake the vanes, and oversize tramp material lodges in individual tubes. Each plugged tube forces its gas share through its neighbours, raising their velocity and wear while contributing nothing. Periodic internal inspection — tube by tube, not just a hopper glance — is the only way plugging is caught before it cascades.
04 — Fan arithmetic
Pressure drop is the real price tag
80–120 mmWC, converted honestly into kilowatts.
How much fan power does a multicyclone consume?
Fan shaft power equals gas volume flow times pressure drop, divided by fan efficiency. At the design-basis flow of about 51 m³/s actual, each 10 mmWC (98 Pa) costs roughly 7 kW at a combined fan-and-motor efficiency of 72 % — so a multicyclone at 100 mmWC draws about 70 kW continuously, comparable to hundreds of MWh per year.
DESIGN BASIS — ILLUSTRATIVE CALCULATION, NOT A GUARANTEE · all concentrations @ 6 % O₂ dry · gas flow ≈ 51 m³/s actual, derived from the design-basis ESP (collecting area 3,888 m² ÷ SCA 75.7 s/m) on the 60 t/h bagasse-fired boiler
Work the numbers through: 51 m³/s × 981 Pa (100 mmWC) = 50 kW of ideal flow work; at 72 % combined fan-motor efficiency, about 70 kW at the motor terminals. Over an 8,000-hour operating year that is roughly 560 MWh — power generated by the plant and consumed by its own draught system. This is the honest frame for comparing collectors: a multicyclone has no electrical absorbed power, but its 80–120 mmWC is not free; it is fan power in different clothing. The same arithmetic applied to a bag filter at 150–200 mmWC end-of-life explains why pressure drop, not capital price, dominates lifetime cost comparisons between particulate technologies.
It also explains why "improving" a multicyclone by raising velocity is a dead end. Cyclone pressure drop rises roughly with the square of inlet velocity while d₅₀ improves only with its square root — so halving the cut diameter mechanically means four times the velocity and roughly sixteen times the pressure drop — an order of magnitude on the fan bill, and accelerated tube wear in the bargain. When finer collection is needed, the answer is a different force, not a faster vortex.
05 — Selection
Where a multicyclone is the right answer — and where it is not
A good tool for exactly one job description.
When is a multicyclone the right choice?
When the duty is coarse: knocking out ash and sand above roughly 10 µm, quenching glowing embers ahead of a fabric filter, or pre-separating the bulk of a heavy dust load at temperatures beyond what other collectors tolerate. It is the wrong choice whenever a stack limit in tens of mg/Nm³ — or even hundreds — must be met at the multicyclone outlet.
The legitimate roles: ember knock-out ahead of a bag filter, where capturing and quenching glowing char protects fabric rated no higher than 260 °C; coarse pre-separation that unloads a downstream collector and takes the most abrasive fraction out of the gas first; and rugged service on very hot or dirty streams where its refractory-lined, zero-moving-part construction survives conditions that would exclude anything else. In these roles the 65–85 % efficiency is not a weakness — the fines it passes are exactly what the final collector is sized for.
The illegitimate role is final collector on a modern permit. At a bagasse-boiler inlet of 6,000 mg/Nm³, even the top of the realistic band leaves 900 mg/Nm³ at 6 % O₂ dry at the outlet. No maintenance campaign, tube redesign or vendor promise moves a mechanical cyclone bank into the tens of mg/Nm³; the physics of Section 02 forbids it. Plants in that position are not maintaining a collector — they are deferring a replacement.
06 — Upgrade path
From multicyclone to Electrocyclone: same slot, different physics
The retrofit that changes the train without rebuilding it.
For plants whose multicyclone no longer earns its pressure drop, the direct upgrade is Arrow Energy's proprietary Electrocyclone: a hybrid pre-collector that keeps cyclonic separation for the coarse fraction and adds DC corona charging so the 5–20 µm fines migrate electrostatically to the same collecting wall. It occupies the same position in the train and a comparable plot — 60–65 % of an equivalent-duty ESP footprint — and shifts stage collection from the mechanical 65–85 % to 85–95 %.
| Parameter | Multicyclone | Electrocyclone |
|---|---|---|
| Separation mechanism | Centrifugal only | Centrifugal + electrostatic augmentation |
| Stage collection | 65–85 % | 85–95 % (design basis 88 %) |
| Outlet to final collector | 900–2,100 mg/Nm³ | 720 mg/Nm³ (design basis) |
| Downstream ESP collection needed for 24 mg/Nm³ | 97.3–98.9 % | 96.67 % |
| Pressure drop | 80–120 mmWC | CONFIRM: Electrocyclone pressure drop range, mmWC |
| Electrical absorbed power | ≈ 0 kW | 120–140 kW |
| Service temperature | Material-limited; very high | ≤ 400 °C |
| Wear on silica-rich ash | High — separation velocity is the wear mechanism | Lower velocity for the same duty; reduced erosion |
| Maintenance | Tube replacement, plug clearing, hopper sealing | Near-zero moving parts; HV supply and electrode checks |
DESIGN BASIS — ILLUSTRATIVE CALCULATION, NOT A GUARANTEE · all concentrations @ 6 % O₂ dry · 60 t/h bagasse-fired boiler · 6,000 mg/Nm³ raw dust · Electrocyclone 88 % → 720 mg/Nm³ · four-field ESP 96.67 % → 24 mg/Nm³ · train total 99.60 %
The downstream consequences are where the upgrade pays. Under the Matts-Öhnfeldt sizing relation (k ≈ 0.5) used for real polydisperse dust, required ESP collecting area scales with the square of ln(1/P); relaxing the ESP's duty from ~98–99 % to 96.67 % is the difference between adding a field to an existing precipitator and leaving it alone. For a bag filter, cutting inlet load by a further factor of two to three extends media life and drops cleaning frequency in proportion. Anonymised reference classes for this architecture — a 170 t/h bagasse sugar mill in Thailand, a 250 t/h biomass power plant in Thailand, a 230 t/h sugar mill in Colombia — are summarised under references.
FAQ
Engineering questions, answered
What is a multi cyclone dust collector?
A multi cyclone dust collector, or multicyclone, is a bank of small parallel cyclone tubes — typically 150–300 mm diameter — inside one casing with a common inlet, outlet and hopper. Each tube spins the gas so centrifugal force throws particles to the wall. It is purely mechanical: no bags, no electrical field, no moving parts.
How efficient is a multicyclone really?
On biomass-boiler fly ash, expect 65–85 % overall collection. Quoted figures of 90 % and above are usually measured on coarse test dust with a mass-median size of 20 µm or more; real bagasse and wood ash carries a heavy fine-char fraction below the tube cut diameter, which passes straight through.
What pressure drop does a multicyclone impose?
Typically 80–120 mmWC across the unit, paid continuously in induced-draught fan power. At the 60 t/h design-basis gas flow of roughly 51 m³/s actual, each 10 mmWC costs about 7 kW of fan shaft power — so a 100 mmWC multicyclone consumes on the order of 70 kW around the clock.
Why is my multicyclone losing efficiency?
Two mechanisms dominate. Tube and vane erosion from silica-rich ash — worst on rice husk, whose ash is 85–90 % amorphous silica — opens clearances and degrades the vortex. And air in-leakage at the hopper or dust valve pulls collected dust back up the vortex core, re-entraining material that was already caught.
Can a multicyclone meet modern particulate emission limits?
No. At a bagasse-boiler inlet load of 6,000 mg/Nm³, a multicyclone at 65–85 % collection leaves 900–2,100 mg/Nm³ at 6 % O₂ dry — one to two orders of magnitude above limits set in tens of mg/Nm³. It serves as a pre-separator or ember knock-out, never as the final collector.
What is the upgrade path from a multicyclone?
Replace it with an Electrocyclone in the same position: Arrow Energy's proprietary hybrid adds electrostatic augmentation to cyclonic separation, lifting stage collection to 85–95 % against 65–85 % mechanical, for 120–140 kW of absorbed power. On the design basis, that cuts the load reaching the final collector to 720 mg/Nm³.
Related engineering pages
Send us your plant data
Fuel, boiler capacity, gas flow, current emission and the limit you must meet. An Arrow engineer replies with a technical assessment basis — not a brochure.