POME: what a wide-channel tubular UF can and cannot do
POME is the liquid discharge from palm oil extraction — hot, acidic, oily and colloidal, with most of its organic load in solution. That combination is what fouls a fine-channel membrane, and it is why channel width, not pore size, is the first decision on this stream.
Four properties decide the outcome
Oil that emulsifies
Free oil floats and can be skimmed off. Emulsified oil does not, and it is the single most common reason a membrane fouls in a mill. Whether the oil in a given POME is emulsified depends on the fruit, the mill and what happens upstream.
Solids that are not just silt
POME carries fibre, grit and shell fragments as well as biological solids. Coarse material is exactly what an open channel tolerates and a narrow one does not.
Temperature
It leaves the mill hot. Elements have a temperature rating, so either the stream is cooled before the membrane or the element is specified against the real inlet temperature — measured at the membrane, not at the mill outlet.
Acidity
POME is acidic. That matters twice: for the wetted materials, and for whether an acid clean is available when the membrane eventually fouls.
Ultrafiltration separates on size. The dissolved organic load that drives BOD and COD passes straight through it. If the target is BOD or COD, that is the biological stage's job and a membrane only polishes what comes out of it. We would rather say this at the start than have it discovered at commissioning.
Where a tubular UF earns its place in a mill
1 · Polishing after the biological stage
When the limit being missed is on suspended solids or oil rather than on BOD, a membrane downstream of the pond or the tank is often the shortest route back to compliance.
2 · Ahead of reverse osmosis
Where the mill wants reuse rather than discharge, the RO stage needs solids and oil taken out first. This is the application we know best — RO pre-treatment is most of what our reference installations do.
3 · Concentrating a sidestream
Where the value is in recovering oil or in cutting the volume going to the pond, the membrane is a concentrator rather than a polisher. Different duty, different sizing, different cleaning regime.
Replace the anaerobic stage. A membrane generates no biogas and removes no dissolved organics. A mill with no biological treatment does not need a membrane first — it needs the biology, and then possibly a membrane.
Why channel width comes before pore size here
The element range covers five open channel diameters. On an oily, fibrous stream the choice between them decides whether the plant runs or plugs.
| Channel | When it is the right choice |
|---|---|
| 6 mm | The cleanest streams, where packing density matters more than solids tolerance. Not the choice where fibre and grit are present. |
| 8 / 10 mm | The common middle. Most polishing duties land here, and it is where we would start on a POME stream. |
| 12.7 mm | Where the solids or oil load is above what the middle range tolerates. |
| 25 mm | For streams that plug anything smaller. The widest channel we make; it belongs in the specification, not in the headline. |
What "mechanically cleanable" means in a mill
Cleaning is done by pushing sponge balls through the channel, not by backwashing alone. On an oily stream that is the difference between a clean that works and one that merely redistributes the oil over the membrane surface.
Three numbers decide whether it works — none is on a datasheet
Ask these of the mill before anything is sized:
- Oil content, and specifically whether it is emulsified
- Peak suspended solids, including fibre and grit
- Temperature at the point the membrane would sit
- What the existing biological stage already achieves on BOD and COD — that sets what the membrane still has to do
These four are why we do not publish a design flux for POME. A number that is right for one mill is wrong for the next one.
How a first POME project would run
We will be straight about where we stand. Our operating references are in landfill leachate, RO pre-treatment, plant extraction and chemical effluent — not in a palm oil mill. What we have is the element, the channel range, and a bench test that answers the question in two weeks rather than two years.
1 · Bench test
5–20 L of POME, chilled or preserved. We run it across the candidate channel at your operating temperature and report a flux curve, a cleaning trial, and a plain answer on whether the element is the right tool.
2 · Pilot at the mill
The skid ships to site and is plumbed into the stream, and runs until the cleaning interval repeats — typically four to eight weeks.
3 · Sizing
Element count, tank volumes and chemical consumption come out of the pilot numbers, not out of a catalogue.
If the bench test says the membrane is the wrong tool for your stream, we say so before you spend money on a pilot.
Send a POME sample
The bench test is the cheapest part of the project and the only part that turns these four questions into numbers. Tell us the mill, the flow and what you are trying to meet.
Common questions
No. Ultrafiltration does not generate biogas and does not remove dissolved organics. It polishes what the biological stage produces; it does not stand in for it.
Not on its own. BOD is largely dissolved, and ultrafiltration separates on size. Where the limit being missed is on suspended solids or oil, a membrane helps directly. Where it is on BOD, the biological stage has to carry it.
Measure it where the membrane would sit, not at the mill outlet. If it is above the element rating, either cool the stream or tell us the real number and the element is specified against it.
That is the case where a membrane is most likely to be the wrong answer, or to need de-emulsification upstream first. It is also the first thing a bench test shows, which is why the test comes before the pilot.
More on this: Selection guide · Tubular UF · Pilot testing