Alumina and Red Mud Filter Cloth for Bayer Refineries
Red mud is the insoluble bauxite residue left after Bayer digestion; alumina hydrate is the product filtered before calcination. Both are filtered — but on a Bayer refinery the usual kit is drum, disc, leaf/Kelly and chamber press filters, and a horizontal vacuum belt is not the default machine of the circuit.
Anyone selling you a belt for red mud has not asked what you run
The default kit across the industry is drum, disc, leaf/Kelly and press filtration. Another cloth supplier states on its own site that belts have been tried on some red-mud washing duties and that the effect was not ideal. We think that statement is worth repeating rather than burying, because it is the single most useful thing an alumina engineer can know before they start shopping for cloth.
There is a counter to it, and it belongs on the page too. Belts do exist on alumina side streams, and at least one third-party compatibility listing includes horizontal vacuum belt filters among alumina options. Existence is not the same thing as default. If your site genuinely runs a belt on an alumina duty, that is a machine family we have registered and can talk about properly — but it is not what we would assume from the word "alumina".
So the first question on this page is not which cloth. It is which machine, and then which stage of the circuit. Everything below is organised that way.
Where filtration happens in the Bayer circuit
The Bayer process digests bauxite in a hot caustic soda liquor. The alumina goes into solution as sodium aluminate; everything that does not dissolve — iron oxides, titania, undigested silica, the desilication product formed in the digester — stays behind as bauxite residue, which the industry calls red mud. One refinery therefore has two very different solids to handle: a residue nobody wants, and a hydrate product that has to come out clean.
That is why "alumina filter cloth" is not one duty. Solid-liquid separation shows up at least three times in the circuit, on three different machines, with three different reasons for the cloth to be there — recovering caustic from the residue, protecting the product from carry-over solids, and washing the hydrate before it is calcined.
The other thing that separates this from an acid-side duty is the liquor itself. It is alkaline and it is hot, and both of those act on the fabric continuously, whether the machine is running or standing full.
If you have landed here from a gypsum page: the chemistry is the opposite end of the pH scale, and the machine population is different. Nothing written for a phosphogypsum belt transfers to a Bayer residue filter.
Phosphogypsum filtration →The circuit, and which steps are filtration steps
- 01 Digestion Bauxite is attacked with hot caustic soda liquor; the alumina dissolves as sodium aluminate. Not a filtration step
- 02 Clarification The insoluble residue — red mud — is settled and counter-currently washed to recover soda. Filtration step
- 03 Security filtration The pregnant liquor is polished before precipitation so that residual solids do not report to the product. Filtration step
- 04 Precipitation Alumina hydrate is crystallised out of the cooled, seeded liquor. Not a filtration step
- 05 Hydrate filtration The hydrate is filtered and washed before it goes forward to calcination. Filtration step
This is industry-level process description, not a description of any particular refinery. Operating values — liquor temperature, caustic concentration, solids loading, particle size — are plant-specific and are not published here as typical figures. They go on the confirmation list instead.
Why a hot caustic liquor is hard on ordinary cloth
The mechanisms below are general polymer and process engineering. How severe each one is on your filter depends on numbers we do not have — which is the whole point of the confirmation list further down.
-
Caustic attack on polyester
Bayer liquor is a hot, strongly alkaline caustic soda solution. Polyester is an ester polymer — its ester bonds are hydrolysed by hot alkali, and the fibre loses weight and strength from the surface inwards.
Cloth implication: A cloth that is perfectly sound in an acid duty can be the wrong polymer here. Polymer choice is decided against your liquor, not against a catalogue.
-
Temperature and chemistry pull against each other
Polypropylene stands up to caustic far better than polyester does, but it gives up temperature headroom in exchange.
Cloth implication: Which way that trade lands depends on the actual liquor temperature at your filter — a number we do not have and will not assume.
-
Scaling in the fabric
Bayer circuits throw desilication product and other scale out of solution onto any surface in contact with liquor, including cloth.
Cloth implication: Scale in the fabric closes free area permanently; it is not washed back out the way a soft cake is.
-
Very fine, high-surface-area solids
Red mud carries iron oxide fines down into the sub-micron range, which is exactly the size that lodges in a weave rather than forming a clean cake.
Cloth implication: Blinding behaviour and cake release have to be treated as one question, not two.
-
The cloth form changes with the machine
A drum needs a sleeve, a leaf needs a bag or a covered element, a chamber press needs plates, a belt needs an endless fabricated loop.
Cloth implication: These are different products with different fabrication. Nobody can answer "which cloth" before the machine is named.
Our own temperature and chemical limits are not on this page. They are on the UNKNOWN list below until the factory confirms them in writing, and they are then stated in the quotation for your duty rather than published as a general figure.
What the circuit runs, and what we have registered
The left column is the machine population of the industry. The right column is our own scope as it stands today — two equipment families, neither of which is the Bayer default. We would rather you read that from a table than discover it in an email.
| Stage or machine | Usual machine type (industry-level) | In our equipment register |
|---|---|---|
| Red mud separation and washing | Thickeners and washers upstream; residue dewatering historically on drum filters, and increasingly on filter presses where the residue is dry-stacked. | No. Neither registered family covers a Bayer residue drum. |
| Security / polishing filtration | Pressure leaf (Kelly-type) filters, normally run with a filter aid so the leaf builds a precoat. | No. Leaf and Kelly cloth is a different product from a belt. |
| Hydrate filtration | Horizontal pan and vacuum drum or disc filters, sized for the wash duty ahead of calcination. | No. |
| Residue dewatering on a pressure filter | A filter press, where the plant is moving residue towards a stackable cake. Published mappings of this duty name the machine only as a press — chamber, membrane or vertical tower is not settled there. | Only where that press is a vertical pressure (tower) filter — that is a registered family. |
| Horizontal vacuum belt filter | Present on some alumina side streams, but it is not the default machine of the circuit. | Yes — this is our registered belt family. |
Whether this mill can make cloth in those forms is exactly what has not been confirmed to us in writing. It reads UNKNOWN in the table below and it will stay UNKNOWN until it is confirmed — not until it becomes convenient.
Machine types above are named only to identify what a cloth would have to fit.
The one plausible overlap
Residue dewatering on a pressure filter is the place where our scope and a Bayer refinery can meet. Where a plant is pushing bauxite residue towards a stackable cake, the route is a press rather than a vacuum machine. Which kind of press is not settled by any public mapping of this duty — they name the machine only as a press — so that is a question for you rather than for this page. If your answer is a vertical pressure (tower) filter, that is a family we have a cloth page for: endless zigzag cloth, low elongation, made to the plate size. That page states its own UNKNOWN list honestly too.
If your residue press is a chamber or membrane press rather than a tower press, say so in the RFQ. That is a different cloth again, and whether we can cover it is on the UNKNOWN list.
How a cloth would be specified for this duty
Selection runs backwards from the liquor and the machine, in that order. The caustic concentration and the temperature at the filter narrow the polymer; the solids — fine, alkaline and prone to scaling — narrow the weave and the surface finish; the machine decides the form of the cloth and its dimensions. On a Bayer duty the polymer question comes first rather than last, because a fabric that is chemically wrong will fail on strength no matter how well it filters.
We are not going to name a polymer for your circuit on a web page. Getting that answer from a published table instead of from your liquor analysis is how a refinery ends up replacing cloth on chemistry it could have designed around.
Specification status
Our factory specification package has not been confirmed in writing yet, so every parameter reads UNKNOWN — including the four capability rows at the top, which ask whether a given cloth form is within scope at all. Each one is fixed during the engineering review and stated in the quotation, not inferred from this page.
| Parameter | Value | Unit | How it is fixed |
|---|---|---|---|
| Drum / disc filter cloth for this duty | UNKNOWN | — | Factory capability not confirmed in writing yet |
| Leaf / Kelly filter cloth for this duty | UNKNOWN | — | Factory capability not confirmed in writing yet |
| Chamber / membrane press cloth for this duty | UNKNOWN | — | Factory capability not confirmed in writing yet |
| Vertical pressure (tower) cloth for this duty | UNKNOWN | — | Family page exists; this duty not yet assessed against it |
| Alkali-resistant polymer range | UNKNOWN | — | Confirmed per polymer and finish |
| Continuous temperature limit | UNKNOWN | °C | Confirmed per polymer and finish |
| Caustic concentration limit | UNKNOWN | — | Confirmed against your liquor analysis |
| Air permeability | UNKNOWN | L/dm²/min | Test method and test pressure will be stated with the value, once the mill confirms them in writing |
| Max fabricated width | UNKNOWN | mm | Confirmed against machine width |
| Seam / fabrication options | UNKNOWN | — | Confirmed against machine and seam photo |
Industry reference — not our specification
| Observation | What is observed | Source |
|---|---|---|
| Default machine kit on a Bayer refinery | Drum, disc, leaf — increasingly press | Industry-to-machine mapping, third-party plants |
| Published third-party view on belts in alumina | Tried on some red-mud washing duties; described as not ideal | Statement published by another cloth supplier on its own site |
Both rows describe the industry and other people's published statements. Neither is a measurement of ours, and neither is a claim about what this mill can supply.
What we can build for this duty
The parameter table above reads UNKNOWN because our own values are not confirmed in writing yet, and we will not fill that table with industry averages. This section is the other half of the answer: the options this cloth is actually built from, what each one changes, and which piece of your data decides it. Nothing below is a specification — it is the shape of the decision.
Cloth form On a Bayer refinery the usual kit is drum, disc, leaf or Kelly and chamber press. A horizontal vacuum belt is not the default here, and nothing in this section changes that — it describes what can be built once the machine is identified, not a case for putting a belt on a Bayer circuit.
Recommended material direction
Bayer liquor is hot and strongly alkaline, and that single fact removes the polymer that dominates every other duty on this site. Polyester is an ester polymer: hot alkali hydrolyses its ester bonds and the fibre loses weight and strength from the surface inwards. The failure is chemical, it is progressive, and it does not look like wear.
This narrows the field; it does not pick. Which polymer goes on your machine is settled against your measured numbers and not on a web page — the same position the selection section above takes. What follows is the reasoning behind that position, not a departure from it.
- Polypropylene PP
Continuous reference up to ~90 °C — a published property of the polymer, not a value of ours
- Acid Excellent
- Alkali Excellent
- Hydrolysis Excellent
- Abrasion Good
It stands up to caustic far better than polyester does. The constraint is the reference continuous temperature above, and a Bayer filter can sit beyond it — which is precisely the number this page does not have.
- Polyphenylene sulfide (PPS) PPS
Continuous reference up to ~190 °C — a published property of the polymer, not a value of ours
- Acid Excellent
- Alkali Excellent
- Hydrolysis Excellent
- Abrasion Good
Excellent in alkali with the highest reference temperature of the realistic candidates, so it is the direction where the liquor is both hot and caustic. The cost step over polypropylene is large and has to be earned by the duty.
- PET Hot caustic is the published failure mode for polyester. Not a direction on Bayer liquor unless the temperature at your filter is genuinely low — and that is a measurement, not an assumption.
- PVDF Its weak side is strong alkali, which is the opposite of the usual assumption about fluoropolymers. Cost and chemistry both point away from it here.
What actually decides it The liquor temperature at your filter, and the machine family. Both are measurements this page does not have and neither can be inferred from the word "Bayer".
Everything in this block is reasoning from published polymer behaviour applied to the chemistry described further up this page. It is a direction, not a selection — and the selection is made against your measured numbers during the engineering review, then written into the quotation. The full polymer envelope is on the materials page.
Weave options
Red mud carries iron oxide fines down into the sub-micron range, which is the size that lodges in a fabric rather than building a cake on it.
- Double layer / multi-layer Two or more interlaced fabric layers woven as one. A fine filtering face carried on a coarse structural back — high strength with fine retention, and drainage channels inside the fabric. Watch: Thicker and heavier; more expensive to weave. The usual answer on heavy belt duties.
- Twill The crossing point steps sideways each pick, giving a diagonal rib. More flexible and more permeable than plain at the same yarn count, with better cake release. Watch: Slightly less dimensionally stable than plain.
- Satin Long floats, few crossing points. The smoothest surface of the three, so the best cake release and the easiest to wash back. Watch: Floats snag and abrade; retention is coarser at a given yarn count.
Where blinding is the dominant failure the surface matters more than the rating: what the cloth recovers after a wash decides the cycle, not what it passed when it was new.
Yarn form. Multifilament and staple yarns retain the fines better and release them worse. On a sub-micron feed that trade is the specification, not a detail of it.
Yarn form, weave and surface finish are settled as one decision. The four yarn forms available are monofilament, multifilament, mono + multi combination and staple / spun.
Permeability considerations
Air permeability is the figure everyone compares and the easiest one to quote meaninglessly. It is measured per ISO 9237 at a stated pressure drop — often 200 Pa — or per ASTM D737 (Frazier). A value with no stated pressure and no stated unit is not comparable to anything, including ours. The same applies to a pore or retention rating: on a woven cloth that is a nominal figure and not an absolute cut-off, because woven media do not have a single pore size and the cake quickly becomes the real filter.
Comparing permeability figures across machine families on this circuit is close to meaningless — a drum cloth, a leaf cloth and a press cloth are asked to do different jobs. Settle the machine first, then ask for the method and the test pressure.
How each property is measured, and what to ask any supplier for alongside the number, is set out on the testing and inspection page. Naming the method is a fact about a public standard, not a claim about our laboratory.
Joint options
The descriptions below are written from vacuum belt and endless-loop practice, because that is where this vocabulary comes from. On a drum, disc, leaf or chamber press the equivalent question is how the cloth is made up and secured, not which joint it takes.
- PAD / welded overlap The two ends are overlapped and thermally fused. A smoother, quieter joint that seals better than a clipper and carries load more evenly. Watch: Made off the machine or with a portable press; a change-out takes longer.
- Clipper (metal hook) Interlocking metal hooks crimped onto both cloth ends, joined by a pintle wire. Fitted and replaced on the machine without dismantling it. The fastest change-out. Watch: The joint is a discontinuity: it leaks vacuum as it passes the box, and it is where most belt failures start. Hook material has to suit the liquor.
- Spiral Spiral coils on each end interlaced and pinned. Flexible, drains through the joint, distributes load along the full width. Watch: Adds an open line across the belt — not appropriate where the joint must seal.
- Sewn overlap The ends are overlapped and stitched. Simple and repairable in the field. Watch: The stitch line is the weak point and a leak path. Least common on vacuum belts.
Which of these is relevant depends entirely on which machine you run. On a chamber press the question is not a joint at all but how the panel is made up, and that is settled from the machine.
The four constructions drawn in cross-section: how joints fail and what each one costs you.
Edge treatment
- Cut and sealed The cut edge is thermally sealed against fraying. The baseline. Adequate where the machine edge seal is in good condition.
- Folded / hemmed The selvedge is folded back and secured. Adds thickness and stiffness at the edge where the seal rubs.
- Reinforced edge band A separate band is bonded along the selvedge. For machines that chew edges — the usual answer when a belt keeps failing at the edge while its body is sound.
- Coated / sealed band A polymer band is applied to the edge. Seals the edge against vacuum bypass as well as protecting it mechanically.
- Guide strip / tracking profile A profile bonded to the underside to run in the machine guide. Only where the machine is built for it. Fitting one to a machine that is not is a common and expensive mistake.
These are belt and loop edge treatments. On drum, disc and leaf cloth the equivalent question is the made-up edge and how it is secured, which follows the machine.
Relevant equipment
Whatever cloth form this duty turns out to need, it is decided at the loom before it is decided anywhere else: what a weaving platform can hold — pick density, tension, width — sets what the finished cloth can be asked to do. Our weaving runs on DORNIER (Germany) and Sulzer (Switzerland) platforms, machines built for technical textiles rather than for apparel.
What the weaving platform decides here is the same as anywhere else — construction held consistently across the width — but on this circuit the prior question is whether the cloth form is ours to build at all.
Model designations, working widths and installed quantities are being confirmed against the machine nameplates and are not published until they are. A weaving platform is easy to name and hard to fake, and we would rather you were able to check.
The full platform description, the eleven process stations from yarn to crate, and the nine laboratory checks: manufacturing.
Factory capability for this duty
The rows below are the part of our capability envelope that this duty touches. The last group is the one that makes the rest of it worth reading — a supplier who can do everything has told you nothing.
Machine families
- Rubber-belt horizontal vacuum belt filter Built to your spec
Core family. Endless belt with clipper, PAD or spiral joint and a finished edge.
What decides it Brand and model, measured width and developed length, a photograph of the joint.
- Drum, disc, leaf or chamber press Engineering review
Not our default family. We ask what machine you run before saying anything about cloth.
What decides it The machine family, confirmed first.
- Indexing / tray vacuum filters Not offered
Different cloth mechanics. We will tell you so rather than sell you a belt that does not fit.
Cloth construction
- PA, PPS, PVDF, PTFE Engineering review
Available as a direction where the duty genuinely needs them. Cost and lead time are the constraint.
What decides it Whether the duty actually needs them rather than PET or PP.
What we do not do
- Quoting from an industry name alone Not offered
Two plants in the same industry run different slurries through different machines. We ask which.
The full envelope, including the machine families and services we decline: what we build to order and what we will not take on. Fifteen pages on how a specification like this is built and read: the filter cloth specification guide.
How alumina-side cloth comes off the machine
A failed cloth is usually three-way: something in the cloth, something in the machine, and something in the process. On a caustic duty the first question is always whether the fabric failed chemically or mechanically, because the two look different and lead to different fixes.
Chemical degradation of the fabric
The cloth goes soft, loses tensile strength and tears rather than blinds. A polymer-selection question, or a liquor-temperature excursion.
Diagnose this →Scale blinding
Throughput falls away and never comes back after a wash cycle. Process-side scaling interacting with weave choice.
Diagnose this →Seam or joint failure
Cloth-side and fabrication-side: seam type, squareness, and how the joint was closed on site. Applies to any fabricated cloth, not only belts.
Diagnose this →Belt tracking and wrinkling — only if you run a belt
Often machine-side (roller alignment, edge guides) before it is cloth-side. Irrelevant on a drum, a leaf or a press.
Diagnose this →We do not quote a life figure for this duty, and we have no measurement of our own to quote from. Anyone who offers you a number of months for a Bayer circuit without asking the liquor temperature and the caustic concentration is selling, not engineering.
Where the replacement is a fabricated endless cloth, measuring the old one and photographing where it failed is the first step in telling the three causes apart. How to measure your old belt →
What we need from you before we can even say yes or no
On most duties this list is what turns an enquiry into a quotation. On this one the first two items decide something earlier than that: whether the answer is a quotation or an honest no.
| What to confirm | How to get it |
|---|---|
| Which machine — type first, then brand and model | Drum, disc, leaf/Kelly, chamber press, tower press or belt. This one question decides whether we can help at all. |
| Which stage of the circuit | Red mud washing, security filtration, or hydrate filtration. They are three different duties on one site. |
| Liquor temperature at the filter | From the process sheet. Caustic plus heat is the combination that decides the polymer. |
| Caustic concentration and free soda | From a recent liquor analysis rather than from the design basis. |
| Cloth form and dimensions in use now | Sleeve, bag, plate set or endless loop — with the measurements of the one you are replacing. |
| Photos of the old cloth | Face and back, plus wherever it failed. This is the single most useful thing you can send. |
| What failed | Strength loss, blinding, cake moisture, or seam. Compare against the failure modes above. |
| Whether a filter aid or precoat is used | Relevant on security filtration especially — it changes what the cloth is actually asked to retain. |
Before you enquire
Can you supply red mud filter cloth?
It depends entirely on the machine, and we cannot answer it from the words "red mud" alone. If the duty is on a vertical pressure filter, that is a registered family and there is a product page for it. If it is a drum, a disc or a leaf/Kelly filter — which is the more likely answer on a Bayer circuit — then whether this mill can cover that cloth form is on the UNKNOWN list above, and we will say so plainly rather than take the enquiry and work it out later.
Why does this page argue against the machine you sell cloth for?
Because the machine population on a Bayer circuit is drum, disc, leaf and press, and printing the opposite would cost an engineer a purchase order and a shutdown before it cost us anything. If a belt is genuinely what you run on an alumina duty, tell us and we will engage with it properly. If it is not, you have saved yourself an email.
Get an honest answer on an alumina duty
Three steps after you submit: an engineer checks the machine and the stage you named, we come back for anything missing, and you get either a quotation, a sample plan, or a clear statement that this is outside what we can confirm.
Every cloth is made to your machine dimensions — we quote per RFQ. No online checkout.
Samples available — freight terms confirmed per request.
Where to go next
Vertical Pressure Filter Cloth
Tower-press cloth: endless zigzag, low elongation, made to the plate size. The plausible overlap with this circuit.
Vertical Pressure Filters
The machine family behind that cloth, and what we need from each machine.
Phosphogypsum Filtration
An acid-side duty where the belt genuinely is the standard machine — the contrast is instructive.
What Is Confirmed About Us
What this mill has and has not confirmed in writing, stated as it stands rather than as we would like it.