Choosing between hollow-fibre and tubular UF by feed solids
The configuration question is usually presented as a technical one. It is mostly a commercial one. Hollow fibre puts far more membrane area into the same volume, and membrane area is what a plant pays for. So the real question is not which configuration is better — it is whether your feed lets you use the cheaper one.
We manufacture tubular. This note is written so that the answer comes out right even when it is not us.
The short answer
Hollow-fibre UF is typically specified up to about 1% suspended solids (roughly 10 g/L). Between 1% and 3%, either configuration can work, and the deciding factors are viscosity, particle size distribution and oil content. Above roughly 3% solids, tubular configurations are the practical choice: the open channel tolerates the solids load and can be cleaned mechanically rather than by backwash alone.
Rule of thumb from field experience. Confirm against your own feed data before sizing.
| Feed solids | Configuration to look at first | Why |
|---|---|---|
| Up to ~1% (≈10 g/L) | Hollow fibre | Highest packing density, lowest cost per m², smallest footprint |
| ~1% to ~3% | Either — the feed decides | Both can work; viscosity, particle size and oil break the tie |
| Above ~3% | Tubular | Open channel tolerates the load and can be cleaned mechanically |
Starting points for discussion, not a specification.
Why area is the deciding commercial variable
A membrane train is priced mostly by membrane area and by the number of pressure housings, pumps and valves that come with it. Hollow fibre wins on both: a fibre bundle delivers several times the membrane area of the same volume of tubes, so a duty that needs one rack of hollow fibre can need a considerably larger installation in tubular.
What hollow fibre buys you
- Higher membrane area per unit volume
- Lower cost per m² of membrane area
- Smaller footprint, fewer housings
- Lower pumping energy at low solids
What tubular buys you
- An open channel that tolerates high solids
- Mechanical cleaning, not backwash-only
- Wall shear that keeps oil moving
- Tolerance of coarse and fibrous material
Both configurations separate on the same principle — a size cut in the ultrafiltration range. The difference is not what they remove, it is what they can be fed. That is why the selection argument is really an argument about the feed, and why the solids number matters more than the membrane data sheet.
What actually blocks a channel
Solids concentration is a proxy, not the mechanism. What blocks a membrane channel is the combination of concentration, particle size distribution and viscosity. A 2% stream of coarse, free-settling particles behaves very differently from a 2% stream of fine colloidal material, even though the gravimetric result on the lab bench is the same number.
Particle size distribution. Fines and colloids pack into a cake that resists backwash. Coarse, free-settling solids are much more forgiving at the same TSS.
Viscosity. Viscosity sets how much shear the channel can develop at a given pressure. A warm, thin stream scours its own channel; a cold, thick one does not, and the same solids figure becomes a much harder duty.
Oil. Oil blinds a membrane before solids do, and it does not come off with a normal backwash. Where the feed can carry oil, the configuration question is usually settled by the oil number rather than the TSS number.
This is why we ask for peak solids rather than average, and for the worst hour rather than the daily mean.
The four numbers that move the boundary
The 1% and 3% figures are a starting frame, not a verdict. Four measurements can move a project from one band to the other:
| Measurement | Direction it pushes | Note |
|---|---|---|
| Peak TSS, not average | Upward → tubular | Excursions foul a membrane; the daily mean does not describe them |
| Oil and grease | Upward → tubular | Emulsified oil is worse than free oil; gravity separation upstream stops working once it is dispersed |
| Fines and colloid fraction | Upward → tubular | A high fines fraction at moderate TSS can be harder than a high TSS of coarse solids |
| Temperature | Downward → tubular | Cold feed raises viscosity and reduces the shear the channel can develop |
Salinity does not decide between hollow fibre and tubular — both are pressure-driven and both concentrate salts. High salinity with high recovery pushes the train towards STRO / DTRO, which is a different question with a different answer. It is worth separating the two decisions rather than treating "difficult feed" as one problem.
The most common place a project gets this wrong
Where screening, settling, dissolved air flotation or oil separation sits upstream, the membrane never sees the raw stream. It sees what comes out of the pre-treatment step — and that stream can be well inside the hollow-fibre band even when the raw feed is not.
Selecting on raw feed data when pre-treatment is planned tends to over-specify the membrane and under-specify the pre-treatment. The result is a tubular train doing a job hollow fibre could have done, while the upstream step that would have protected either one is left undersized.
Ask what the membrane inlet actually looks like at the worst hour of the worst day. That is the design feed.
The same logic applies in reverse. If pre-treatment is being added later, the configuration chosen today may be more conservative than the plant will need once it is in place — worth knowing before the housings are ordered, because the membrane area cannot be quietly un-bought.
Side by side
| Consideration | Hollow fibre | Tubular |
|---|---|---|
| Solids tolerance | Typically up to ~1% | Above ~3%, and the practical choice in between |
| Oily or emulsified feed | Oil blinds the fibre and resists backwash | Wall shear keeps oil moving |
| Cleaning | Backwash-based | Mechanically cleanable |
| Membrane area per unit volume | Higher | Lower |
| Cost per m² of membrane area | Lower | Higher |
| Footprint for a given duty | Smaller | Larger |
| Tolerance of coarse or fibrous material | Limited by channel size | Open channel, more tolerant |
| Separation principle | Size cut in the UF range | Size cut in the UF range |
How the decision is actually made
On paper the boundary is a solids figure. In practice it is a pilot. Flux, operating pressure and cleaning interval all depend on how one specific feed interacts with one specific membrane, which is why these figures are confirmed per project rather than published as fixed values — see tubular UF for the parameters we do publish and the ones we do not.
A short pilot on the actual feed settles the configuration question, the flux, and the cleaning interval in one pass. It is also the cheapest point in a project to discover that hollow fibre would have done the job.
Related: the full selection guide · what ultrafiltration does not remove
Send us three numbers
Peak TSS, oil content and salinity — plus the temperature, if you have it. That is enough for us to tell you which configuration to look at first, including whether hollow fibre would do the job for less money.
Send your feed data