Ground Screws vs Concrete Foundations

The part of a veranda that decides whether it is still standing straight in twenty years is the part you will never look at. Everything above ground gets discussed at length — colour, roof glazing, projection, whether the posts land somewhere sensible relative to the back door. What sits underneath those posts is usually settled in a single sentence, if it comes up at all. That is the wrong way round: the ground is the one part of the job that cannot be adjusted afterwards without taking down everything above it.

At Roma Verandas the question of what lies beneath a patio is raised early, and the honest answer at enquiry stage is often that nobody knows yet. This guide explains what a veranda foundation is resisting, why a paving slab on its own is not a foundation, how a concrete pad is formed, how ground screws differ in practice, and which ground conditions push the decision one way or the other.

What the Posts Are Actually Carrying

Start with weight. An aluminium frame is light, but glass is not: laminated safety glass in the thicknesses used overhead weighs somewhere in the region of twenty-five to thirty kilograms per square metre, so a roof of four metres by three carries three to four hundred kilograms of glazing before the beams and gutter are counted. Polycarbonate is a fraction of that. Snow adds more: design loads across lowland Britain are modest, but even a modest figure over twelve square metres runs to several hundred kilograms, and the sharper version is snow sliding off the house roof above, concentrating a season's accumulation on the strip nearest the wall in seconds.

The load that governs the foundation, though, is wind, and it acts upward. A veranda roof behaves rather like a free-standing canopy: air accelerating over the top generates suction, and any wind that gets underneath pushes up. Both pull the same way, and a structure heavy enough to sit still under its own weight is not automatically anchored against uplift. It is uplift rather than weight that the fixing at the foot of each post has to resist. Add wind pressing on side screens and the post becomes a lever: a base plate perhaps two hundred millimetres across resisting a two-and-a-half-metre post, which turns a modest horizontal push into a much larger tension in the windward bolts.

Why a Paving Slab Is Not a Foundation

A typical modern patio is a slab of thirty-five to fifty millimetres, bedded on mortar or sand over a hundred or so millimetres of compacted stone, if it was laid properly. Many older patios are worse: slabs spot-bedded on five dabs of mortar with voids beneath. That is entirely adequate for what it was built for, which is people and furniture spread over a wide area.

It is not adequate for a post. An expansion anchor set into forty millimetres of concrete has almost no embedment, and the failure mode is not the bolt shearing but a cone of slab pulling out around the fixing, or the slab lifting off its bed and taking the post with it. The problem is not that paving is weak. It is that it is not connected to anything. A foundation has to reach into ground that will not move, and a slab by definition sits on ground that already has.

The Concrete Pad and What It Involves

The traditional method is a mass concrete foundation cast under each post. The paving is lifted, which matters in itself: matching a slab discontinued a decade ago is often impossible, and the reinstated area rarely disappears completely. A hole is dug out — commonly four to six hundred millimetres square and a similar depth, deeper where the ground is poor — past topsoil and made-up material into firm undisturbed subsoil, then filled with a standard structural mix.

The post is then fixed onto bolts cast into the wet concrete using a template or, far more commonly, onto resin anchors drilled once the pad has cured, which is more forgiving of small setting-out errors. Curing is a genuine constraint: resin anchor manufacturers generally specify a minimum concrete age before an anchor may be loaded. There is also spoil, roughly a quarter of a tonne per post, and it has to leave the garden. None of that is a criticism. Concrete is thoroughly understood, works in almost any ground and is easy to inspect. What it costs is time, mess and disturbance to a patio the customer may well have paid a lot for.

Ground Screws and Where They Change the Calculation

A ground screw is a galvanised steel shaft with a helical flight or coarse thread at the lead, driven in with a hydraulic drive head, commonly in lengths from around eight hundred millimetres to a metre and a half. Capacity comes from friction along the shaft and bearing on the helix, and it works in tension as well as compression, which suits the uplift problem particularly well. There is no cure, so the frame can go up the same day, and no spoil. Where the patio is staying, a diamond core is taken through the slab and the screw goes down the hole, leaving a neat disc to patch rather than a lifted and relaid area. It is also reversible, which matters for a rented property or a customer who suspects they will move.

The honest limitation is obstruction. Rubble, an old shed base, a buried raft or the compacted stone under a former driveway will stop a screw, and when one refuses the options are to shift the position, drill through, or fall back to a pad — a decision made on the day, not in an office beforehand. A machine also has to reach the garden, which rules the method out where the only access is through the house or down a narrow side passage. A screw is also only as good as the material it engages: very soft or made-up ground needs a longer screw, or more of them.

Reading the Ground Before You Choose

Clay is the most common complication in British gardens. It shrinks and swells with moisture, and close to mature trees the zone of seasonal movement reaches surprisingly deep; guidance for house foundations near trees routinely calls for depths approaching a metre. A veranda is far lighter than a house, but a post that rises and falls a few millimetres each year shows up as a twist in the roof line or a sliding panel that binds in August and not in February. Clean sands and gravels are the easier end of the scale: they drain freely, move little with the seasons, and give a screw good friction provided it is driven deep enough. Made ground deserves similar caution: on estates built from the 1960s onwards, rear gardens were frequently levelled with rubble and site arisings half a metre deep. Nothing about that material can be assumed, and a trial hole is worth more than confident guessing.

The issue that causes the most trouble in practice, though, is buried services. Soil pipe runs, rainwater drains to a gully, land drains, an old soakaway, the supply to an outside tap and, increasingly, the pipework to an air source heat pump all live in exactly the strip of ground where veranda posts want to go. A five-hundred-millimetre excavation and a metre-long screw are equally capable of finding a drain. Inspection covers, gully positions and the line you would draw between them tell you most of what you need before anything is dug. A design with a little flexibility in post spacing is therefore more robust than one where a post must land on an exact spot.

The Question Worth Asking Early

The most useful thing a homeowner can do before a survey is establish — or freely admit they do not know — what is underneath the patio. When was it laid, and was it laid over an older one. Was the garden ever levelled or terraced. Does a drain cross it, and where does the downpipe discharge. Photographs of the paving, the chamber covers and the whole rear elevation are worth more at that stage than any set of measurements.

Neither method is universally right. Ground screws are quicker, cleaner and reversible and they suit uplift well, but they can be defeated by ground nobody knew was there. Pads are slower and messier, but they can be formed anywhere and adapted as the hole is dug. In most gardens either would work perfectly well, and the choice comes down to access, the state of the paving and how much disruption you will accept. What is not optional is that something proper goes underneath the posts. If you would rather have the method settled with a contingency understood than quoted on an assumption, that is worth raising at survey stage — and occasionally the truthful answer is that we will know when the first slab comes up.