EPB Face Conditioning: Where Polymer Foam Ends and SAP Begins

On a 6.4 m earth pressure balance machine cutting through mixed gravel and silty clay, the foam injection ratio had crept from 40 to 65 percent over a month because the belt was slipping and the muck was running off the conveyor. More foam made the muck wetter, not stiffer, and the screw conveyor started to lose face pressure. The fix was not more foam. It was 0.6 kg of coarse polyacrylate granules per cubic metre of muck, added at the belt transfer point.

Foam and granules are asked to solve different problems, and confusing the two is the most common conditioning mistake on EPB drives. This article separates the jobs, gives the slump targets that decide the dose, and covers the compatibility question nobody asks until the anti-clay agent starts fighting the polymer.

Foam Does Three Jobs, Granules Do One

Foam lubricates the screw and the cutterhead, reduces friction between particles so the muck flows, and provides a compressible medium that helps hold face pressure. It also adds water, typically 30 to 50 litres per cubic metre of muck at a 5 percent expansion ratio. That water is the problem when the ground already delivers 18 to 22 percent moisture and the belt cannot grip.

Granules do one job: they take free water out of the muck matrix and lock it into a gel so the material behaves like damp soil rather than wet gravel. They do not lubricate, they do not help face pressure, and they add 0.5 to 1.2 kg of dry mass per cubic metre. Treat them as a water-management tool, not a conditioning agent.

The Slump Test That Decides the Dose

Fill a standard slump cone with muck straight off the belt and lift it. A 150 mm slump that holds its shape and shows no free water ring is the working target for a conveyor transfer. Below 100 mm the muck is too stiff and will bridge in the hopper. Above 200 mm, or any visible water ring, and the belt will slip on the incline.

Run the test at three points: at the screw discharge, at the transfer point before dosing, and 30 m along the conveyor. If the slump drops more than 40 mm between the first two points, the muck is losing water to the belt and the dose needs to move upstream, not up. Dosing at the wrong point is why some sites report that granules do nothing.

Dose per Ring: 0.4 to 1.2 kg per Cubic Metre

A 6.4 m machine at a 1.5 m ring advance excavates about 48 cubic metres per ring. At 0.6 kg per cubic metre that is 29 kg per ring, or roughly one 25 kg bag and a bit. A wet gravel face at 22 percent moisture may need 1.0 to 1.2 kg per cubic metre, while a dry clay at 14 percent needs 0.2 to 0.4.

Do the moisture check weekly, because the number moves with the geology and dosing on last week’s setting is how sites end up with 8 percent overconsumption over a drive. A moisture meter on the belt, calibrated against oven drying once a fortnight, pays for itself in a month on a drive of this size.

Compatibility With Bentonite and Anti-Clay Agents

Two additives can neutralise an anionic polyacrylate. Anti-clay agents are often cationic surfactants or polymers that bind directly to the granule surface and stop it wetting. If both are in use, dose them at different points, at least 3 m apart, and never premix them in the same tank. Bentonite is less of a problem but consumes water: a 6 percent bentonite slurry holds water tightly and will cut granule efficiency by 30 to 40 percent.

Calcium in groundwater is the third factor. In a gypsiferous ground with 600 mg/L calcium, a standard grade drops from 280 g/g to about 90 g/g, and the dose has to rise accordingly. Ask for an absorption figure in the site water before you set the rate, and keep a 25 kg trial bag on site to re-check when the face changes.

Coarse cuts between 0.5 and 2 mm work best at the belt because they resist attrition and do not blow off in the ventilation airflow. The granular SAP products page sets out the mesh options and the brine figures that matter underground.