Pretreatment

Why pretreatment decides everything downstream

The membrane is the expensive, fragile part of the plant.

A reverse-osmosis membrane is a thin, physically unforgiving barrier. It is specified to a tight particle-size limit and a defined scaling envelope, and it is expensive to replace. Everything upstream of it exists to keep the feed inside those limits for long enough to be economic.

That inverts the usual intuition. Pretreatment is not the boring part of the design; it is the part that determines whether the rest of the plant is worth having. An under-designed pretreatment stage does not fail loudly — it fails as a slow loss of performance, and then as membrane replacement.

It is also where most of the engineering judgement sits, because the right barrier depends on what the water actually carries. A raw surface water needs solids removal and disinfection control; a groundwater source may need only conditioning and iron removal; a wellfield source may need very little at all.

The barriers, in the order a train usually builds them

Pretreatment is not one process but a sequence of barriers, each removing something the next one cannot tolerate. The sequence usually starts with physical removal of large material, moves to progressively finer solids removal, and ends with whatever chemical or biological control the source needs.

Screening and similar physical steps stop large debris and organisms reaching the rest of the plant. Media filtration removes suspended solids to a design filtration rate rather than to a particle size. Cartridge filtration is the fine, disposable barrier in front of the membranes themselves. Ultrafiltration pushes that barrier to a genuinely tight pore size and is often used where the design needs a defined integrity test. Chemical conditioning handles what filtration cannot — coagulation, antiscalant, biocides — and disinfection control matters wherever biological growth threatens the train.

The technology cards below are the reference for each of these. Where the platform genuinely calculates a barrier — media filtration area, cartridge rating, ultrafiltration flux — the card says so, and the matching sizing tool is listed with it.

What pretreatment has to achieve before RO

For reverse osmosis specifically, pretreatment has two distinct jobs that are easy to conflate. The first is particulate control: keeping suspended solids and colloids below the limit the membranes were designed for. The second is scaling control: keeping the concentrate below the solubility limits of the sparingly soluble salts — calcium sulfate, barium sulfate, silica among them — at the recovery you intend to run.

These are different problems with different timescales. Fouling shows up as a fall in normalised permeability over weeks or months and is usually managed by pretreatment and cleaning. Scaling can precipitate inside the membrane element within hours at the wrong recovery, and it is not recoverable by cleaning at all. The recovery you can defend is therefore set by the scaling envelope, not by the membrane’s rated flow.

This is why the design basis and the scaling tools sit on this hub alongside the filtration ones. Choosing a recovery before establishing what the concentrate will do is the most common way an RO design fails.

Pretreatment & membrane technologies

Each entry states its own modelling level. Anything EngiMetric does not calculate is marked, rather than implied to be engineered.

• Media Filtration Quantitatively modelled

  Granular-bed filtration that removes suspended solids ahead of a membrane, sized here from design flow, filtration rate and bed geometry.

• Feedwater Screening & Treatment Selection Quantitatively modelled

  Compares a stated or measured feedwater analysis against the barriers a desalination pretreatment train can actually contain, raising a candidate only from a fact you supplied or the platform derived — and reporting its own confidence rather than choosing for you.

• Ultrafiltration Quantitatively modelled

  Low-pressure membrane barrier that removes colloids, bacteria and suspended solids ahead of RO or NF, quantified by membrane area, TMP and flux condition.

• Nanofiltration Quantitatively modelled

  Loose-rejection membrane separation between UF and RO, used for hardness reduction and partial demineralisation before an RO stage.

• Ion Exchange Quantitatively modelled

  Exchanger-based demineralisation and softening, sized by resin volume, bed area, service flow and regenerant demand.

• Cartridge Filtration Reference only

  Disposable or washable depth filters that protect downstream membranes from particles, ahead of any high-pressure stage.

Pretreatment sizing tools

Each tool calculates one thing and shows its working. They are the right next step once the principle makes sense.

• RO Scaling Risk & Pretreatment Analysis

  Screen CaCO3 scaling risk in RO feed and concentrate using the Langelier Saturation Index (ASTM D3739) and the Stiff & Davis Stability Index (S&DSI) for high-TDS and seawater feeds.

• Filter Sizing Calculator

  Size a conventional rapid-gravity multimedia filter from design flow, filtration rate, bed depth and media effective size: required filter area, design area after the engineer's safety factor, and media volume. The unit count needs a stated catalogue unit area, and no clean-water head loss is invented.

• Cartridge Filter Sizing Calculator

  Size cartridge and bag filter elements and housings from design flow, feed solids and the manufacturer's element flow rating: hourly solids load, element count and housing count. No element size is assumed and no differential pressure is invented.

• Disinfection CT Value Calculator

  Compute the achieved CT value from residual disinfectant concentration and contact time, compare it against a required CT you state from your disinfection reference, and get the contact time the design needs to reach it.

• Ultrafiltration (UF) Membrane Engineering

  Size UF membrane area from permeate flow and design flux, compute transmembrane pressure from the feed/filtrate/concentrate pressures, and read normalised flux against the reference condition to expose fouling.

• Nanofiltration (NF) Membrane Engineering

  Close the NF water and solute balance from a stated recovery and rejection to get permeate and concentrate flow and TDS, then size membrane area from the permeate flux.

• Ion Exchange (IX / IER) Bed Sizing

  Size an ion exchange bed from service flow, resin volume and empty-bed contact time, and compute stoichiometric regenerant demand from the ionic loading.

• Sodium Metabisulphite Dosing Calculator

  Dose sodium metabisulphite from the chlorine residual it must neutralise or the sulphite level it must hold, using the stoichiometric ratio and your product strength. No chlorine residual and no product strength are assumed.

• Coagulant Dosing Calculator

  Turn a jar-test coagulant dose into active metal and product mass per day, from the dose basis you measured and the product active content. The dose is your jar test and the pH outcome is not predicted.

• Clarifier Sizing Calculator

  Size a circular or rectangular clarifier from the surface overflow rate, then check it against the solids loading when you state both. Reports which criterion governs and never predicts a solids removal.

• UV Disinfection Sizing Calculator

  Size a UV reactor bank from the required dose and design flow, then count parallel reactors against the out-of-service allowance. The ideal power is stated separately from the validated reactor power.

Demo cases with pretreatment in the train

Each case opens in the Plant Designer as a working, editable design. They are synthetic demonstration bases — not project data.

• Seawater — open intake, single RO stage Quantitative case
• Seawater — beach wellfield, lower design flux Quantitative case
• Brackish groundwater — single RO, higher recovery Quantitative case
• Seawater — UF pre-treatment ahead of RO Quantitative case
• Municipal seawater — larger flow, multi-vessel RO Quantitative case
• UF Prologue + Brackish RO Quantitative case