Two diseases, two addresses
Fouling is deposition of suspended and organic material — silt, colloids, biofilm, organics — and it happens where the dirt arrives first: the lead elements of the first stage, which act as expensive depth filters for whatever pretreatment let through.
Scaling is precipitation of dissolved salts — calcium carbonate, calcium sulphate, silica — and it happens where concentration is highest: the tail elements of the last stage, where the concentrate approaches its solubility limits on the way out.
The geography is the first diagnostic: probe or stage-wise data that points to the front of the train says fouling; the back says scaling.
Reading the symptoms
The three plant-level trends to watch (normalised for temperature and pressure — raw numbers mislead seasonally):
| Symptom | Fouling (front end) | Scaling (tail end) |
|---|---|---|
| Differential pressure (feed–concentrate ΔP) | Rises early and fastest, stage 1 | Rises late, last stage |
| Normalised permeate flow | Falls | Falls |
| Normalised salt passage | Roughly steady, or slow rise | Rises distinctly |
Salt passage is the sharpest discriminator: a scaled surface disrupts the membrane's rejection locally, so permeate quality deteriorates in a way plain dirt usually does not cause early on. Biofouling adds its own signatures — ΔP that returns within weeks of cleaning, and slime on the cartridge filters and probe samples.
The cures are opposite
Cleaning chemistry must match the disease. Alkaline cleaners (high pH) lift organics, biofilm and particulate fouling. Acid cleaners (low pH) dissolve carbonate scale and metal oxides. Order matters when both are present: alkaline first, then acid — acid first can set organic foulants like cooking an egg, and acid on a biofilm-loaded element does very little at all. Sulphate scales barely respond to either, and silica scarcely to anything practical — those two you prevent, because you cannot reliably clean them off.
Clean at 10 – 15% loss of normalised flow (or equivalent ΔP/salt-passage triggers) — waiting for 30% turns a routine CIP into a partial recovery at best, because compacted deposits shield what is under them.
Prevention, briefly
Fouling prevention is the pretreatment article's territory: SDI below 3, carbon integrity, biology managed, and honest flux — running elements above design flux drags foulant onto the surface faster than crossflow removes it. Scaling prevention is arithmetic: choose recovery from the concentrate-side LSI and sulphate/silica projections, then hold it — the plant that "just turned up the recovery" is the plant that scaled a quarter's membrane budget into the tail vessels. Antiscalant only works at the dose and recovery it was projected for; changing either without re-projection quietly voids the protection.
And log from day one: normalised flow, salt passage and per-stage ΔP weekly. Trends diagnose; snapshots guess.
Rule of thumb: front of the plant, ΔP first, quality holding — think fouling, clean alkaline. Back of the plant, salt passage rising — think scaling, clean acid, and re-examine the recovery you are really running. If the same symptom returns within a month, the cause is still upstream.