At pH 9.6 the Gel Skins Over Before the Core Wets

Nine point six. That was the pH of a batch drawn from a cell that had received 4,000 tonnes of cement-stabilised waste the previous month. The operator dosed the usual 2.8 percent and got a batch that looked set on top and was still liquid underneath at four hours. A finger poked into the surface went through a rubbery skin into free leachate.

Alkaline leachate is one of the few conditions where the failure is not about how much polymer, but about how fast the outside of each particle reacts.

What happens at the particle surface

Polyacrylic acid absorbs by uncoiling. In water near neutral pH, the carboxylate groups repel each other and the network opens, which is what lets water flood in. Above about pH 10, two things change at once. The external solution is rich in sodium or calcium hydroxide, so the ionic strength at the particle surface is high and the uncoiling is suppressed. And the surface layer hydrates fast, forms a gel envelope, and becomes a diffusion barrier before the interior has seen water.

The result is a skin. It is mechanically strong enough to survive gentle mixing, it keeps the dry core dry, and it makes the batch appear to thicken when it has only partly reacted. Under a paint filter test the skin survives and the core does not, so the sample fails for free liquid that was never absorbed.

The mixing mistake that makes it worse

When a batch looks slow, the instinct is to add more powder. On alkaline leachate this is the wrong move, because the extra powder forms more skins rather than more gel. Doubling the dose can produce a batch that is harder to mix, contains more unreacted material, and fails the same test.

Equally damaging is slow agitation. A paddle turning at 60 rpm in a 10 m3 tank does not provide the local shear needed to break a forming skin. Operators see a lumpy batch, assume the powder is poor quality, and switch suppliers, which costs a month and changes nothing.

Pre-acidification, in the right place

The clean fix is to bring the pH down before the powder goes in. Muriatic acid or a ferric or aluminium coagulant will do it; the choice depends on what the site already handles. Target pH 7.5 to 8.5, measured on a mixed sample rather than a grab, and add the acid with agitation, because a strong base neutralised locally can spike and then rebound.

The order matters. Acid first, mix for ten minutes, confirm pH with a calibrated meter, then powder. Dosing powder and acid together gives the polymer a window of high pH exactly when it is trying to wet, which is the worst possible sequence.

Where acid handling is unacceptable on site, an alternative is to dilute the alkaline stream into a larger volume of neutral leachate before dosing. Dilution reduces the surface concentration even though it does not change the total alkalinity. It works, but it means the batch volume is set by the alkalinity rather than by the disposal plan, which is a planning constraint some sites cannot live with.

Reading a batch that has skinned

Cut a set sample in half with a trowel after thirty minutes. A healthy batch is uniform from surface to centre, with a slight sheen. A skinned batch shows a distinct rind, two to five millimetres thick, and a softer, wetter interior. If the rind is there, the dose is not the problem, and adding powder will not fix it.

The second test is a simple dispersion check. Take 50 g of set material, place it in a litre of clean water, and stir gently for a minute. A well-formed gel disperses into cloudy fines. Skinned material leaves intact blobs with dry powder visible when you break them open.

One plant on a lime-stabilised waste cell solved this by adding a 15-minute pre-mix of the batch with recirculated treated effluent at pH 7.2 before powder dosing. Consumption dropped 18 percent and the paint filter pass rate went from 70 to 96 percent. No new equipment: one valve and one tank.

The number to put on the batch sheet is pH before dosing, not pH of the raw sump a week earlier. Alkaline leachate drifts as carbonation proceeds, so a value from the previous batch is often half a unit out. Where the site expects regular alkaline batches, a grade formulated for a wider pH band will tolerate more drift, and the technical data for each grade lists the working range explicitly.