Standing Water Cleared from a Sunken Patio with a Wykamol Land Drain and Pea Shingle

The patio was a 24 square metre run of natural stone, laid maybe eight years earlier. At the lawn edge it sat about 40mm low, and a shallow basin had formed roughly through the middle. After a storm, the shape of the puddle stayed visible for a full day, sometimes two. The slabs were sound. The pointing was mostly intact. The fault sat below the surface and along the edges, where settled ground and a nearly flat original fall had left the water with no proper exit.

Before any digging started, a bucket of water went down on the slabs. Water gives away the route before a spirit level confirms it. It ran toward the north corner, gathered there and stayed there. That corner became the place to intercept it, because lifting the centre of the patio would have turned a local drainage repair into a full relaid surface.

Why the puddle formed where it did

A patio needs a fall that gives surface water a lower place to travel. The usual working figure for landscapers is around 1 in 80, roughly 12mm of drop per metre of run. This patio had been laid closer to 1 in 200. After ground movement, the low point had reversed slightly, so rain reaching the slabs collected in the dip and then disappeared only by evaporation or slow seepage through the joints.

The subsoil made the defect worse. A trial hole beside the fence found heavy clay about 300mm down, the sort of ground that holds water in place. Moisture that passed through the jointing gathered under the slabs, softened the bedding and helped the settlement continue over time. A land drain suited this corner because it gave the trapped water a channel with a continuous fall away from the clay pocket.

Full lifting would have made sense if the whole slab field had lost its fall. Here the settlement had settled down and the worst of the drainage problem was concentrated at one corner, so the edge drain was chosen as a targeted repair.

Cutting the trench and setting the fall

The trench followed the two edges nearest the low corner, along the boundary between patio and lawn. It was dug about 300mm deep and 200mm wide, deep enough to take the pipe, the shingle bed and the geotextile wrap without leaving the top loose.

The trench bottom mattered more than the surface of the slabs above it. A line and level set the bottom to drop steadily at about 15mm per metre toward the discharge point. That point was a gravel-filled dispersal area in a lower part of the garden, well clear of the house.

Wykamol supply perforated coil land drain in 80mm and 100mm bore, commonly sold pre-wrapped in a filter sock or bare for wrapping on site. The 100mm bore was used here. Before the pipe went in, the trench was lined with non-woven geotextile membrane and folded up both sides into a U-shaped liner.

That membrane was there to keep the 20mm pea shingle clean. Fine clay particles work into unprotected shingle and choke the voids that the water is meant to move through. That silting is what kills many amateur land drains inside three or four winters. The membrane was cheap insurance against the trench turning back into a wet clay slot.

About 50mm of 20mm pea shingle went into the bottom first. The perforated coil sat on that bed with the holes facing down and to the sides, then more shingle was poured over it until the trench was nearly full. The geotextile was folded over the top to close the shingle around the pipe before the surface was capped.

The repair did not use a pump or a soakaway crate. It relied on a fall the water could actually follow, from the old low corner to lower ground beyond the patio.

The discharge point

The drain had to empty somewhere legal and lower than its start, so it fed a shingle dispersal pit two metres square in the far corner, where the ground was more permeable. Sending a land drain into a rainwater downpipe or a foul sewer creates the sort of trouble that only appears later, once the system is already buried.

Repointing the stone after the water had a route out

Once the edge drain gave water somewhere to leave, the joints between the slabs could be treated as part of the surface finish instead of the only escape route. Cement mortar pointing had already shown its usual weakness: it cracks, it lets water through unpredictably and it can stain natural stone badly.

GftK VDW resin grout was the better fit discussed for this patio. It is a resin slurry, made in two-part and single-part forms, brushed across the whole patio and worked into the joints before the surface is swept clean ahead of curing. It sets firm with a little flexibility, resists frost heave and avoids the staining associated with cement on natural stone.

The resin needs enough joint to grip. On this type of work, the joints should be at least 5mm wide and 30mm deep. The slabs also have to be bone dry when the grout goes down, because damp stone can leave a haze across the face after curing.

For this natural stone, the disturbed edge pieces were hand levelled off a full mortar bed so adjacent slabs sat dead flush. That mattered as much as the pointing, because proud edges catch water and foot traffic, then begin the same loosening cycle again.

There was one separate issue with the stone itself. Marshalls Symphony and similar riven products can bloom with efflorescence in their first year or two, leaving a white powdery salt haze on the surface. The mark comes from calcium and other salts migrating out as the bedding dries. Dry brushing and normal rain clear most of it. Stubborn patches can be treated with a proprietary efflorescence remover used sparingly and rinsed off, with the mild approach first because acid-based cleaners can etch the wrong stone.

Holding the edges so the fall does not drift again

The original patio had almost certainly weakened at the perimeter. Slabs at an unrestrained boundary slowly creep outward and downward, joints open up and the intended fall drifts out of shape. The repair added block edging bedded on concrete and backed with a concrete haunch.

The haunch was a triangular fillet of concrete along the buried back face of the edge course. It did the structural work by stopping the edging blocks from kicking outward under foot traffic and freeze-thaw movement. Open, water-carrying sub-bases need the same restraint, because loose aggregate can move sideways when the boundary is weak.

The sub-base below the disturbed area was rebuilt instead of being packed back loose. MOT Type 1, the crushed graded aggregate that carries the load, only works properly when compacted in layers no thicker than about 100mm. Each layer needs a plate compactor or roller pass. If 200mm is dumped in one lift and given a token whack, it stays loose and settles under use, which is how a flat patio can turn into a reversed fall.

Here, the disturbed sub-base went back in two 75mm compacted layers before bedding and edge pieces returned above it. That gave the repaired edge something firm to bear on and kept the trench cap from sinking into the drain line.

What the water did afterward

The test after backfilling used the same bucket method. Water tipped into the old low corner tracked to the trench line, dropped into the shingle and disappeared within a minute or two as the pipe carried it away.

Through the following autumn, the outline of the old puddle never reappeared. By the end of the following autumn, the old outline still had not returned, even though the hollow that made it visible had never been lifted out.