Edge Restraint Failure Corrected along a 14-Metre Block Paving Border with a Concrete Haunch

Reading the Movement at the Edge

The outer edge had crept outward noticeably across a single season. Along the run of Marshalls Drivesett Argent, the perimeter course had rotated and lipped, and the joints between that course and the second course had opened into visible gaps. Rainwater then tracked through the widened joints and began carrying fines out of the bedding sand, which made the spread accelerate.

A trial pit at the worst point, roughly midway along the run, gave the answer below the blocks. The edge course sat on a thin mortar bed, barely more than skim thickness, with no concrete behind or beneath it. Lateral thrust from foot traffic and the occasional wheelbarrow had nowhere to go except into that exposed outer line.

The field paving over the compacted MOT Type 1 sub-base remained firm. That separated the repair into a narrow edge rebuild instead of a full lift. The outer course and second course had to come up, the failed bedding had to be removed, and a continuous concrete haunch had to be formed along the border.

Setting the Line Before Lifting

String lines went in before any blocks came out. Steel pins at regular centres held the line at the intended finished edge, pulled back to the original design position by referencing two undisturbed corners that had stayed square. The gap between the string and the face of the moved blocks confirmed the average displacement before the lift began.

The first unit came out at the opened joint with a block extractor and a bolster. Once that block was removed, the rest of the outer course lifted with far less resistance. The second course was lifted as well, since both courses needed to be reset against the new restraint.

Each block was stacked by course order on a board. Keeping the order preserved the running bond and avoided unnecessary recutting when the courses went back down. The old mortar bed underneath broke up in handfuls, matching the trial-pit finding and showing that the edge had lost its bearing as well as its restraint.

The exposed sub-base was scraped back to clean MOT Type 1. A straightedge and spirit level checked the line and level, and the surface was tested under the plate. Two soft spots appeared where fines had washed out. Those areas were dug deeper, refilled with fresh Type 1, and whacker-plated in two passes. The remaining sub-base stayed tight under compaction, so the excavation stopped at the failed edge zone.

Cutting and Pouring the Haunch

The concrete haunch had to work as both bearing and buttress. It needed to sit under the edge course and also rise behind the outside face of the block. A trench was cut through the prepared sub-base along the full run, tight to the string line, wide and deep enough to seat the blocks and still leave a substantial body of concrete rising up the back of the outer course.

The concrete mix used a coarse all-in ballast blended with cement. It was kept dry enough to hold a shaped face after placement. That consistency allowed the bed to support the blocks immediately and allowed the rear face to be drawn up into a stable haunch.

Concrete went into the trench as a continuous pour. Edge blocks were bedded straight onto the green concrete and tapped down to the string with a rubber mallet. As each block seated, concrete was pulled up behind the outer face and shaped back at roughly forty-five degrees. The finished haunch stopped short of the top of the block, leaving the upper part clear so soil or gravel could return against the edge without the concrete showing.

Continuity matters along this type of border. Any gap in the haunch leaves unsupported blocks either side of the void, and thrust then concentrates at those points. The pour was handled in short working sections, with each fresh section tied into the previous one while the concrete was still wet. The finished restraint acted as one unbroken beam along the full exposed run.

At the low end, an ACO drainage channel already formed part of the layout and carried surface water to a gully. The new haunch was brought up to the channel’s own concrete surround and stopped there. The ACO unit already provided lateral restraint through that stretch, and its fall had to remain undisturbed.

The green concrete was left overnight before the second course and adjacent field blocks were returned. Early loading can crack a new haunch at the base, especially where the concrete has been shaped high behind the edge course.

Returning the Courses

After the haunch had cured, the second course was laid back on a fresh bedding layer of sharp sand. The sand was screeded to match the level of the retained field so the lifted section sat flush with the paving that had stayed in place.

The stacked blocks went back in their original order. That kept the Argent colour blend consistent across the repair and avoided a patched appearance where the rebuilt section met the retained paving.

Jointing With Resin Compound

The joints were filled with a brush-in resin compound. On an edge that had already opened and allowed bedding fines to wash out, resin gives a firmer joint and sheds water more effectively than loose kiln-dried sand.

The blocks were bone dry before application. The compound was swept diagonally across the joints in two directions to fill the gaps fully, and surplus material was brushed clear from the face before moisture reached it. Once the surface was clean, a light misting activated the cure.

Sandstone, Porcelain and Efflorescence Beside Block Edges

Where sandstone paving slabs meet a reset block border, riven sandstone needs a full wet mortar bed. Voids under natural stone hold water, and that water path can drive efflorescence up through the slab surface. The slab bed described here used a mortar course over the same MOT Type 1 base, with each stone buttered on the back with slurry primer before bedding.

That primer bonds the mortar to the underside of the stone. Sandstone is porous, yet the underside can still release unevenly if the bond is poor. Skipping the slurry can leave hollow patches that later ring when tapped and lift at the corners.

Porcelain behaves differently. A porcelain priming slurry has to be painted onto the back of every unit before the slab touches the bed. Porcelain absorbs almost nothing, so an ordinary mortar bed will not grip it properly. The slurry provides the chemical key. Without it, a porcelain slab can look sound until frost lifts it from the bed intact.

Efflorescence had already bloomed on part of the older sandstone as a white salty haze. The salts migrate out of mortar and stone as they dry. A dilute acidic efflorescence removal treatment cleared the bloom after the surface had been pre-wetted and then rinsed thoroughly.

Pre-wetting matters because a dry slab pulls acid deep into the pores and can etch the finish. On newly laid sandstone, the better defence is the full mortar bed and the slurry primer, which reduce the water path that carries salts up to the surface.

What the Excavation Still Could Not Show

A thin mortar strip had been asked to hold a long exposed border on its own, and once it was broken out the missing haunch was obvious from the failure pattern. What the spoil could not tell anyone was why the original edge had been built without concrete in the first place, and whether the rest of the perimeter had been formed the same way beneath sections that had not yet moved.