4 Drainage Faults Fixed on a Sloping Patio with an ACO HexDrain Channel

Fault one was recorded during the initial lift: the porcelain slabs had been laid on five dabs of mortar per slab, not a full bed. Water tracking under the slabs pooled in the voids between dabs and had nowhere to fall except toward the house. This is the single most common failure on porcelain patios reported by installers working to the ACO HexDrain specification, and it was the reason the channel alone would not have fixed the site.

The existing gradient measured 1:40 across the 22 square metres, feeding water directly at a rendered wall with no cavity tray visible above the damp-proof course. The channel would intercept surface flow. It could do nothing about the sub-slab water already trapped in the dab voids. Both problems had to be addressed in the same excavation.

Why the porcelain came up before the channel went in

Porcelain has near-zero water absorption, typically below 0.5 percent by mass, so any water that reaches the underside of a slab stays there until it finds a fall. On a five-dab bed the underside is mostly air. On this patio the dabs had been placed with no porcelain paving slurry primer on the slab backs, which is the second fault, so even the mortar that was present had bonded poorly and lifted with the slabs in one piece.

The correct method for porcelain over a mortar bed uses a full wet bed at roughly 30 to 40 millimetres, screeded flat, with a slurry primer brushed onto the back of each slab immediately before bedding. The primer is a cement-and-polymer suspension that keys the dense porcelain to the mortar. Without it the bond relies on suction alone, and porcelain has almost none.

Relaying the field meant a full-bed mortar of 4:1 sharp sand to cement, primer applied slab by slab, and a revised fall of 1:60 running away from the wall toward the new channel line. The shallower fall was possible only because the surface water was now being collected mechanically instead of being shed across the whole slope.

Before any slab returned, the sub-base was tested. That is where fault three sat.

The sub-base was 60 mm of unbound sand

Excavation to 250 millimetres exposed the garden path sub base the original crew had relied on: a thin layer of unbound building sand over topsoil, no compacted stone at all. A patio carrying foot traffic and garden furniture needs a bound or well-compacted granular sub-base, normally 100 to 150 millimetres of MOT Type 1 or a Type 3 permeable equivalent, laid in two compacted lifts.

Type 1 is dense-graded and sheds water. On a site draining toward a channel that is acceptable, because the channel does the collecting. Where the design goal is infiltration, a Type 3 open-graded stone lets water pass through to the subgrade. Here the subgrade was clay, so infiltration was ruled out and dense Type 1 at 150 millimetres compacted in two 75-millimetre lifts was specified, with a plate compactor run to refusal on each lift.

The topsoil was dug out and a geotextile separation membrane laid across the formation to stop the clay pumping up into the stone under load. This step is routinely skipped and is the reason many sub-bases turn to slurry within a few years.

The ACO HexDrain run

The channel chosen was the ACO HexDrain Brickslot in the discreet slot-top finish, six metres of one-metre modular runs plus one corner unit and two end caps. HexDrain is a polypropylene channel rated to load class A15, which covers pedestrian and domestic garden use. It is not rated for vehicles, so it stays clear of the driveway crossing at the property boundary.

The channel was bedded on a haunch of C20 concrete, roughly 100 millimetres under the invert and 100 millimetres up each side, with the grating set 3 to 5 millimetres below the finished porcelain surface so water falls into the slot rather than skating over it. The Brickslot top was set flush and level along the run while the channel base followed the built-in fall of the HexDrain sections toward a single outlet.

That outlet fed a 110-millimetre solvent-weld pipe run to an existing surface-water gully 4 metres away, with a rodding access left at the corner unit. The fall inside the pipe was set at 1:80, comfortably within the range for gravity surface-water drainage. The channel intercepts the full 1:60 sheet flow off the porcelain before it reaches the wall, which is the outcome the five-dab patio could never deliver.

A gravel margin 150 millimetres wide was set between the channel and the wall render as a splash-back zone, so wind-driven rain hitting the render drains straight into the gravel and down to the channel invert instead of ponding at the base of the wall.

Edge restraint, or the fault nobody sees until it spreads

Fault four was on the adjoining block paving path, not the patio itself. The blocks had no block paving edge restraint along the garden side. Over two seasons the outer course had migrated outward by up to 20 millimetres, opening the joints and letting the laying course wash out during heavy rain.

Block paving relies on interlock. Every block leans on its neighbour, and the whole field pushes outward at the perimeter under load. Without a restraining edge the perimeter blocks walk, the joints open, sand escapes, and the field loosens progressively from the outside in. The fix was a haunched concrete edge along the exposed side, C20 concrete brought up two-thirds of the block depth and battered away at 45 degrees, with the perimeter blocks re-bedded tight against it.

Jointing: what went back between the slabs

The original patio had been pointed with a sand-and-cement slurry that had cracked and shed weeds within a year. For the relaid porcelain, a two-part resin patio jointing compound was brushed in wet and consolidated with a soft foam float, giving a joint that flexes slightly and resists frost heave.

For the block paving path the choice was Rompox weed resistant compound, a one-part polymer-modified jointing sand that sets firm after wetting and stays permeable. Rompox is swept into dry joints, then activated with a fine spray; the polymer binds the sand grains so they no longer wash out and offer no seed bed. The permeability matters on the path because it drains through the joints to the sub-base, whereas the sealed resin joint on the porcelain sheds every drop to the channel.

Two different jobs, two different jointing products, for the same reason: the porcelain field is a shedding surface feeding a channel, and the block path is a permeable surface draining downward. Matching the joint to the drainage strategy of each surface is what stops the joint failing again.

What the second winter will actually test

The channel, the full mortar bed, the compacted Type 1 and the haunched edge all address faults that were visible on excavation. The one variable still unproven is the clay subgrade under the geotextile. Dense Type 1 over clay keeps water moving to the channel as long as the clay stays stable, but a clay formation that shrinks and swells across a dry summer and a wet winter can still move a rigid mortar bed above it. Whether the 1:60 fall holds its line through a full freeze-thaw cycle over that subgrade is the question the excavation could not answer.