The question
What are the different things that accumulate on an RO membrane, and how do you tell them apart from the control room?
Fouling and scaling are not the same thing
The terms are often used interchangeably on site, but the distinction decides the response. Fouling is accumulation of material that arrived at the membrane in the feed. Scaling is precipitation of dissolved salts that exceeded their solubility inside the system.
That difference matters because the remedies are opposite in direction. Fouling is fixed upstream, by improving pretreatment. Scaling is fixed by changing the antiscalant or lowering the recovery — nothing upstream will help once the concentrate chemistry is beyond its limit.
The five things that foul a membrane
- Particulate fouling — suspended solids that pretreatment did not remove. Usually shows first as rising differential pressure across the lead elements.
- Biological fouling — biofilm establishing on the membrane surface. Self-propagating, hardest to reverse, and often traceable to a dechlorination step that removed the residual too early.
- Organic fouling — natural organic matter and hydrocarbons adsorbing onto the membrane. Common on surface water sources and where any oil is present.
- Colloidal fouling — very fine particles that pass a conventional filter but agglomerate at the membrane surface. This is what SDI is measuring.
- Metal oxide fouling — iron, manganese and aluminium oxidising and depositing. Frequently the result of oxidation between the well and the membrane rather than of high raw levels.
Reading the signature in the data
None of this is visible in raw readings, because raw flow and pressure move with temperature and feed conditions. The data has to be normalised to reference conditions first. Until that is done, seasonal variation and genuine decline look identical.
| What you observe | Most likely mechanism | Where to look |
|---|---|---|
| Differential pressure rising, salt passage steady | Particulate or biological fouling | Lead elements, pretreatment, SDI |
| Normalised flow falling, salt passage rising | Scaling | Tail elements, recovery, antiscalant selection |
| Salt passage rising sharply, differential pressure flat | Membrane damage or o-ring leak | Probe the vessels to locate it |
| Differential pressure rising after every clean, faster each time | Biological fouling | Dechlorination point, stagnant sections, feed biology |
What to do about it
Cleaning chemistry follows the diagnosis. High-pH cleaners address organic and biological fouling; low-pH cleaners address scale and metal oxides. Using the wrong one wastes a cleaning cycle and can set the deposit harder.
But cleaning is the symptom-level response. If cleaning frequency is rising, the system is telling you that something upstream has changed — the source water, the pretreatment performance, or a step that was bypassed and never restored.
The most valuable thing you can do is clean early. A membrane cleaned at a 10–15% change in normalised performance recovers far better than one left until the system can no longer meet demand.
What to check
- 01Normalise operating data before drawing any conclusion from it.
- 02Trend differential pressure per stage, not just across the whole system.
- 03Measure SDI and turbidity at the membrane feed point, not at the plant inlet.
- 04Record cartridge filter element life — a shortening life is a pretreatment message.
- 05Verify antiscalant is actually being delivered, and at the correct rate for the current recovery.
- 06Check the dechlorination point and confirm there is no oxidant reaching the membrane.
- 07Confirm no section of the plant stands stagnant between production runs.
When to get technical support
Where cleaning no longer restores normalised performance, an element autopsy identifies the deposit directly and removes the guesswork. It is a destructive test on one element, but it is far cheaper than replacing a full set on the wrong diagnosis. Where salt passage has risen sharply without a differential pressure change, probing the vessels locates the fault before anything is replaced.
Talk to an EngineerFrequently asked
There is no universal interval — it depends on feed quality, recovery and pretreatment. What matters more than the absolute number is the trend. A stable interval is a healthy system; a shortening interval is the system telling you something upstream has changed.