Industries · Palm oil
Palm oil: dust collection for fibre- and shell-fired mill boilers
A palm oil mill boiler co-fires fibre at 35–45 % moisture with dense, high-LHV shell, typically 20–45 t/h steam at 20–30 barg. Visible black smoke is usually unburnt char — a combustion problem to fix before any dust collector is sized. The standard collector is a multicyclone; the upgrade path for larger mills is an Electrocyclone or ESP.
01 — The duty
The palm oil mill boiler: three fuels from one fruit bunch
Fibre, shell and EFB behave like three different fuels, because they are.
A palm oil mill runs on its own residues. Pressing and kernel recovery leave mesocarp fibre at 35–45 % moisture, palm kernel shell — dense, low-moisture, around 16–17 MJ/kg LHV — and empty fruit bunches (EFB) at 55–65 % moisture as received. The mill boiler, typically 20–45 t/h of steam at 20–30 barg serving sterilisers and turbine, co-fires fibre and shell as standard: fibre arrives finely divided and burns substantially in suspension, shell burns dense and slow on the grate. That split personality defines the dust problem — suspension-burning fibre is the source of the fine char carryover that makes a palm mill stack visibly smoky, while shell contributes little to it.
EFB is the fuel the boiler mostly refuses. It carries the highest potassium and chlorine of the three residues; potassium collapses ash-fusion temperatures, slagging the grate and fouling tube banks, and chlorine attacks superheater metal. Most mills hold EFB below 10–20 % of the blend without additive treatment or washing — a combustion constraint set upstream of any dust collector, and one reason fuel-mix discipline is part of emission control here. This page, part of Arrow's industries series, works through the ash and gas properties, the honest engineering sequence for a smoky stack, and the collector upgrade path as limits tighten across Indonesia, Malaysia and Thailand.
02 — Ash & gas
Fuel, ash and gas properties for fibre–shell firing
The numbers behind the smoke.
| Property | Fibre | Shell | EFB | Design consequence |
|---|---|---|---|---|
| Moisture, as fired | 35–45 % | 10–20 % | 55–65 % | Fibre and EFB drive gas volume; shell stabilises the fire |
| LHV | ≈ 11 MJ/kg | ≈ 16–17 MJ/kg | ≈ 5–7 MJ/kg | Blend ratio sets heat release and carryover |
| Ash content, dry | ≈ 3–6 % | ≈ 1–3 % | ≈ 4–8 % | Moderate mineral load; char dominates fly ash |
| Alkali & chlorine | Moderate | Low | Highest of the three | EFB capped at 10–20 % of blend without additives |
| Burning mode | Largely suspension | Grate | Grate (shredded) | Fibre fines are the carryover source |
| Fly-ash character | Char-rich fines from fibre + coarse grate carryover; embers present | Bimodal PSD; fire risk for fabric media | ||
| Flue-gas moisture | ≈ 15–25 % by volume, blend-dependent | Favourable ESP resistivity conditioning | ||
Note what the table implies for collector choice. The fly ash is char-heavy and ember-bearing — hostile to bag filters without serious pre-collection, exactly as on the biomass power duty — while the wet flue gas holds ash resistivity in the window where electrostatic collection works well without conditioning. Palm mills are consequently plate-and-cyclone territory: mechanical collection as the base, electrostatic collection as the upgrade, fabric rarely justified.
03 — Sequence
Fix combustion first, then collect what remains
A dust collector sized against a combustion fault is sized against the wrong number.
Why is black smoke from a palm oil mill boiler a combustion problem first?
Because the visible plume is unburnt char, not mineral ash — fine fibre lifted out of the furnace before it finishes burning. Wet, uneven fuel feed, holes in the grate fire, and poor secondary-air penetration each dump partially burnt fines into the gas path. Correcting them can cut the dust load at source by half or more, before a single collector is bought.
The honest engineering sequence therefore starts in the boiler house, not the catalogue. Fuel feed evened out (fibre–shell ratio held steady, feeder surging eliminated); grate condition and undergrate air distribution restored; secondary air staged so the suspension-burning fraction completes combustion in the furnace; excess air trimmed toward λ ≈ 1.15 — on units running high excess air this is worth +0.7–0.8 percentage points of boiler efficiency and reduces the gas volume every downstream component must handle. Arrow's energy audit service covers exactly this diagnosis. Only then is the collector sized, against a measured post-correction dust load — and the collector that results is one or two steps smaller than the one that would have been sold against the smoke.
04 — The train
The collection train and its upgrade path
Multicyclone as the installed base; Electrocyclone or ESP as limits tighten.
The standard fit on palm mill boilers is the multicyclone: robust, cheap, no electrical load, and collecting roughly 65–85 % on this duty — essentially all of the coarse grate carryover and little of the fine char below about 20 µm. For decades that matched the regulatory environment. It no longer reliably does: particulate limits for biomass-fired boilers are tightening across the region's palm belts — CONFIRM: current particulate limits for palm-residue-fired mill boilers in Indonesia, Malaysia and Thailand, with reference O₂ — and the fine char a multicyclone passes is precisely what the tighter numbers target.
Arrow's upgrade path keeps the staging logic and upgrades the physics. Step one: retain or refurbish the multicyclone as coarse knock-out and ember quench. Step two, chosen against the applicable limit: an Electrocyclone — cyclonic separation with electrostatic augmentation, 85–95 % stage collection, 50–80 mg/Nm³ standalone outlet at 6 % O₂ dry, near-zero moving-part maintenance and a footprint 60–65 % of an equivalent ESP — where that outlet satisfies the permit; or a compact electrostatic precipitator behind pre-collection where the limit sits in the tens of mg/Nm³. Because the pre-collector removes most of the load, the ESP that finishes the job is a small one — the same collecting-area arithmetic worked on the sugar & bagasse page, scaled down to mill size.
The threshold where this investment typically becomes compelling is around the 21 t/h boiler class and above: large enough that the mill exports power or plans to, large enough to attract regulatory attention, and large enough that the collector cost is small against the boiler it protects. Below that class the calculus is done mill by mill.
Sizing on this duty follows the same arithmetic as the larger biomass plants, scaled down. A 30 t/h mill boiler moves on the order of 60,000–80,000 Nm³/h of wet flue gas depending on blend and excess air; at raw dust loads of a few g/Nm³, a multicyclone at 75 % leaves several hundred mg/Nm³, an added Electrocyclone stage takes the train into its 50–80 mg/Nm³ envelope, and a compact ESP behind pre-collection reaches the tens. Because the pre-collector removes 85–95 % of the mass, the ESP is sized against a load an order of magnitude lighter than raw gas — the same [ln(1/P)]² collecting-area saving worked on the design basis, at mill scale.
05 — Practicalities
Mill realities: space, power and the outage window
Designing for a working mill, not a green field.
Can a dust collector be retrofitted to an existing palm oil mill boiler?
Yes — the multicyclone-to-Electrocyclone or pre-collector-plus-ESP conversion is designed as a retrofit: the new stage installs in or beside the existing collector's plot, connections are prefabricated to suit the mill's low-season shutdown, and induced-draught fan margins are checked against the changed pressure drop before anything is cut.
Three practicalities decide retrofit success. Space: mill back-ends are congested, which is where the Electrocyclone's compact footprint and vertical gas path earn their keep. Power: an Electrocyclone draws 120–140 kW and a mill-scale ESP less than that on this duty — parasitic load the mill's own generation must carry, checked against turbine margin at crushing peak. Ash: the collected char-rich fines are dry and conveyable, and mills commonly return them to plantation as soil amendment; hopper and conveyor sizing anticipates the roughly doubled catch that better collection brings. Anonymised summaries of comparable biomass and mill-boiler projects are collected under references. Guaranteed figures are stated per project after the technical assessment, on a stated basis — mg/Nm³, reference O₂, dry or wet, and load range.
FAQ
Engineering questions, answered
Why does a palm oil mill boiler make black smoke?
Black smoke is almost always unburnt char — fine fibre burning in suspension is lifted out of the furnace before combustion finishes, driven by wet fuel, uneven feeding or poor air distribution. It is a combustion problem first: fixing fuel feed, grate condition and air staging cuts the dust load at source before any collector is sized against what remains.
Is a multicyclone enough for a palm oil mill boiler?
A multicyclone collects roughly 65–85 % on this duty and remains the standard fit on mill boilers. It misses the fine char fraction below about 20 µm, which is what makes smoke visible and what tightening limits target. Whether it is enough depends on the applicable limit; where tens of mg/Nm³ are required it is not, and an Electrocyclone or ESP stage follows.
What is the upgrade path when emission limits tighten?
Keep or refurbish the multicyclone as first stage, then add collection for the fine fraction: an Electrocyclone at 85–95 % stage efficiency where 50–80 mg/Nm³ satisfies the permit, or a compact ESP behind pre-collection where the limit sits in the tens of mg/Nm³. Mills of roughly 21 t/h and above are where the economics of the upgrade typically close.
Why do palm oil mills limit EFB co-firing to 10–20 %?
Empty fruit bunches carry the highest potassium and chlorine of the palm residues. Potassium drops ash fusion temperatures, causing slagging and fouling on grate and tubes, and chlorine drives superheater corrosion. Most mills hold EFB below 10–20 % of the fuel blend unless additives or fuel washing are applied — the boiler sets this limit before the dust collector does.
What dust emission can a palm oil mill boiler achieve?
With combustion corrected and a multicyclone alone, outlet dust commonly remains in the hundreds of mg/Nm³. Adding an Electrocyclone brings the design envelope to 50–80 mg/Nm³ at 6 % O₂ dry standalone, and a pre-collector plus ESP train reaches the tens of mg/Nm³. Guaranteed figures are stated per project on a stated basis and load range.
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.