Forty Seconds Into the Batch: Order of Addition in a 3 m3 Tank

Forty seconds after the first sack went in, the ammeter on the 3 m3 tank jumped and the impeller stopped turning. The operator had poured 15 kg of powder onto the surface of a tank turning at 80 rpm, and the powder had wetted into a raft about a metre across. The raft wrapped the shaft, the motor tripped on overload, and the batch sat for six hours while maintenance pulled the agitator.

The material did nothing wrong. It behaved exactly as a crosslinked polymer should when it meets water in bulk: it swelled where it landed, and the swollen mass was denser than the liquid and stuck to the first solid surface it touched. The sequence was the problem, and sequence is the one variable most plants never write down.

Why high shear does not save you

The instinct when a powder will not disperse is to turn the agitator up. For a swelling polymer, that is the wrong move at the wrong moment. High shear does not break a partly formed gel; it folds it, stretches it, and wraps it around the impeller, which is the most effective way to stop a mixer that exists.

Worse, high shear at the point of addition creates a vortex that pulls the powder down the shaft rather than into the bulk. The powder then hydrates in the low-flow region behind the impeller, which is where the first lump always forms.

The right combination is low to moderate shear during addition, with good bulk turnover, followed by higher shear once the particles are separated and fully wetted. That is the opposite of what most batch sheets say.

A sequence that works at this scale

Fill the tank to 60 percent with water at the process temperature. Start the agitator at the lowest speed that still moves the whole volume, typically 40 to 60 rpm on a pitched-blade turbine in a 3 m3 tank, and confirm there is a gentle vortex rather than a standing wave.

Feed the powder into the vortex, not onto the surface, at 1 kg per 20 to 30 seconds. A vibratory feeder or a screw works; a sack tipped by hand does not, because the flow rate is uncontrolled and the first dump is always too fast. A funnel with a 25 mm outlet held just above the vortex is the cheapest workable rig, and it is what most small plants end up using.

Continue agitation for 10 minutes after the last kilogram before raising the speed. Ten minutes is not a guess; it is roughly the time needed for a 30/60 granule to reach 90 percent of its final swell at 20 C, and anything shorter leaves dry cores that will swell later in the process, usually in a transfer line.

The feed rate is a number, not a habit

For a 3 m3 tank at 60 rpm, the practical feed window is 2 to 3 kg per minute. Faster than that and the local concentration at the feed point rises above the point where particles can separate; slower and the batch time becomes uneconomic. For a 20 m3 tank with proportionally more agitation, 6 to 10 kg per minute is achievable.

Scale by volume and by agitation power, not by the sack size. A plant that doubles the tank and keeps the same feed rate will get a batch that takes twice as long and disperses perfectly; a plant that doubles the feed rate to keep the batch time constant will get lumps.

Fine grades need slower feeding than coarse ones, because they wet faster and have more surface per kilogram. If you switch from 30/60 to 100/200 mesh without changing the feed rate, expect trouble on the first batch.

What to do when it has already gone wrong

Stop adding immediately. Do not raise the speed, and do not add water to thin it; adding water makes the gel swell further and increases the volume of the problem. If the impeller is still turning, hold at low speed and let the mass circulate; some rafts will break up on their own within 20 minutes.

If the impeller has stopped, do not restart it repeatedly. Each attempt compresses the gel further onto the shaft. Isolate the tank, drain the free liquid from the bottom valve if there is one, and remove the mass mechanically. A diaphragm pump will move a soft gel; a centrifugal pump will not.

For recovery, dilution works only if the gel is still particulate. Add water at no more than 10 percent of the batch volume per hour with continued low-speed agitation, and check for particles with a sieve every 15 minutes. A gel that has become a single mass will not redisperse at any dilution, and continuing to try just fills the tank.

Design changes worth making

An eductor fed from the pump discharge disperses powder into the liquid before it can form a raft, and it removes the feed-rate problem almost entirely because the mixing happens in a small, high-turbulence zone rather than in the tank. For plants that dose polymer regularly, this is the single best capital spend.

An off-centre or angled agitator also helps, because it prevents the stable vortex that pulls powder down the shaft. So does a baffle at the addition point.

Write the sequence on the batch sheet with the feed rate, the agitation speed at each stage, and the hold time. Our note on SAP mixing sequence and feed rate gives the numbers for tank sizes from 1 to 20 m3, and the pattern is the same at every scale.