Settler Flooding: Why the Dispersion Band Grows and How to Bring It Back
Flooding means the dispersion band has grown until it reaches the weir, so organic leaves with the aqueous (or the reverse). The plant response is usually to cut flow, which hides the symptom and loses production. Diagnose first — the cause is almost always one of five things.
Diagnostic order
| Check | Reading that points to it | Action |
|---|---|---|
| Total flow vs design | Combined O+A above design, or one phase above its share | Restore design O/A; do not raise both |
| Temperature | Drop of 5 °C or more vs commissioning | Reheat; viscosity rises sharply when cold |
| Crud / rag layer | Visible pad at the interface, rising over days | Desludge, clay-treat organic, find the solids source |
| Organic quality | Dark colour, smell, slow phase separation in a jar test | Caustic wash or clay treatment; consider a bleed |
| Mechanical | Baffle displaced, weir out of level, mixer speed changed | Re-level, restore mixer tip speed |
The jar test is the fastest discriminator
Take a sample from the mixer outlet, let it stand in a graduated cylinder, and time the separation. If it separates cleanly in the same time as commissioning, the problem is hydraulic — flow, weirs, baffles. If it separates slowly or leaves a stable haze, the problem is chemical or organic degradation.
What usually turns out to be true
Crud is the most common cause in leach circuits. Silica, gypsum, flotation organics and fine solids accumulate at the interface. The pad restricts flow area and accelerates flooding. Desludging buys time; the real fix is upstream — improve filtration or add a coalescer.
The second most common is temperature. Viscosity roughly doubles for a 20 °C drop in many diluents, and the settling rate falls with it. Plants that run well in summer flood in winter without any change in chemistry.
Do not fix flooding by raising mixer speed. More shear makes smaller droplets and slows separation. If anything, reduce tip speed and let the band recover.
Recovery procedure
Reduce total flow to 60%, maintain design O/A, raise temperature to set point, desludge the first stage, and hold for one shift. If the band returns to normal within a shift, the cause was hydraulic or crud. If it does not, take organic and aqueous samples for a jar test and a degradation check before restarting at full rate.
Send us the stage volume, design flow, current O/A and a photo of the interface, and we will tell you whether the settler is undersized for the duty or the problem is operational.