Founding and Anchoring a Shed or Pergola Kit

Runners Land Under Joists, Not Under Panel Edges

Turn the floor panel over before you touch the ground. It arrives already framed — a perimeter, a set of joists at whatever centres the manufacturer chose, a sheet deck nailed or stapled over them — and that framing is the entire brief for the base. The runners are not there to hold up a floor. They are there to put a continuous, dry, non-settling bearing under joists that were designed on the assumption that something would be there. A runner that lands halfway between two joists is not supporting the frame at all — it presses on the underside of the deck sheet while the joists either side of it carry on spanning to the next runner that does catch one, and that is how a kit shed floor acquires the permanent dip under the mower.

Orientation comes out of the same inspection. Runners lie along the length of the shed and the joists cross them, so the width in the skid calculator is the dimension the runners are spaced across and the length is the dimension each runner runs end to end. Put those the wrong way round and the arithmetic is still correct — correct about a different building — and the error arrives as a runner order that is too short and too numerous. A 2.4 by 3.6 metre floor returns five runners of 3.6 metres each, which across 2.4 metres places them on 600 millimetre centres: marginally tighter than the two-foot convention the count came from, and that is the direction to err in.

Two placements are worth deciding rather than inheriting. Put a runner directly beneath the door opening: it is the only part of the floor taking a concentrated wheel load, it is where the threshold gets cut for a ramp, and it is the first area to rot because it is the only one rain reaches. And let each runner project 50 to 100 millimetres past the end of the floor frame, so the end grain is clear of the bearing and a bar can get under it when the shed has to be shifted in ten years. Each of those projections is a saw cut through pressure-treated timber, exposing a face the treatment never reached — flood it with a field-applied preservative to the manufacturer's instructions, which is the ground AWPA M4, the Standard for the Care of Preservative-Treated Wood Products, covers.

The pad beneath them runs past the runner grid on every side rather than stopping at it, for the same reason a slab edge is never built at the edge of its sub-base; the tonnage and the compaction that implies are worked through next door in the garden-room guide. The runners sit in ground contact whatever is under them, so they want the ground-contact use category — Use Class 4 under BS EN 335, or its equivalent under AWPA U1 — rather than the above-ground decking timber stacked beside it in the rack, which looks identical and is not.

A Pergola Kit Lands on a Surface Somebody Else Built

A pergola kit is drawn for open, level, unobstructed ground and it is almost never delivered to any. The realistic site is a twelve-year-old patio with a fall across it for drainage, an inspection chamber somewhere inconvenient, a gully against the house, a soakaway nobody can locate and a paving edge that stops 150 millimetres from where post three would like to stand. None of that is a reason to abandon the kit. All of it is a reason to set the post positions out on the ground with chalk before a single bracket comes out of the box.

Four things veto a post position, and only one is structural. An inspection chamber cannot be built over — access has to be maintained, and on a public sewer that is a formal matter with the water authority rather than a preference. A post base bolted close to a free slab edge has no concrete around it to develop anything, which is the next section. A base landing across a paving joint bears on two independently bedded units that will not stay coplanar. And a post in the low corner of a patio's drainage fall is standing in the puddle, whatever the timber is treated to.

Move a post and the beam over it gets longer, which is where a kit stops being a kit. The section supplied was sized for the span on the drawing, and a rated span for that beam is the only thing that says whether it can carry more — from the kit's own literature, or failing that from a published span table for the species, grade and section, such as the AWC Span Tables for Joists and Rafters or supplier tables derived from the strength classes in BS EN 338. If the answer is no, the resolution is another post rather than a longer beam. A 6.0 by 3.0 metre run at an 8 ft spacing ceiling returns ten posts: four down each long side, plus one intermediate on each short end.

Two limits on that number are worth stating plainly, because both cut in the reader's favour. The spacing choice tops out at 8 ft, so a kit whose beam is genuinely rated to span 3.6 metres clear sits outside the menu entirely — feed a 3.6 by 3.6 frame in and it reports eight posts where the kit correctly ships four, which is an artefact of the ceiling rather than a verdict on the kit. And the model assumes a beam runs along all four sides, so on a pergola beamed on two sides only, the intermediate posts on the short ends have nothing to support and come out of the count by inspection.

Bolting to a Slab You Did Not Pour

The hardware bag says the fixings are for concrete, and that is the sum of what the kit knows about the thing it is bolted to. Three facts decide whether they hold, and all three are unknown when the box is opened: how thick the slab is, what is underneath it, and how far the fixing sits from a free edge. A base rail bolted to a 100 millimetre unreinforced garden slab is a reasonable detail. The same rail bolted to a 50 millimetre paving flag on five mortar dabs is not a fixing at all — it is a flag with a bolt through it, and the flag comes up with the shed.

Finding out is a ten-minute job and worth doing before the first real hole. Drill one exploratory hole with a masonry bit in a corner nobody will see, and feel and measure where the resistance stops; the change from concrete to bedding sand or sub-base is unmistakable through the machine, and the depth at which it happens is the number you needed. Where the surface is flags rather than a slab, lift one. On the BS 7533 series of pavements built from precast concrete flags or natural stone slabs, the bedding under them is a construction layer and not a structural one, and no post-installed anchor changes that.

The arithmetic that follows catches people out because a catalogue length reads generously and then loses most of itself before it reaches concrete. A fixed-length wedge or screw anchor gives up the base plate, any shim pack or levelling grout under it, and the washer, nut and thread projection above; only what is left is embedment. Mechanical anchors also want a hole deeper than the embedment so drilling dust falls clear of the tip, tabulated as a minimum hole depth rather than assumed. Between the two, the anchor in the bag frequently cannot reach the embedment its capacity was published at — and the slab is frequently not deep enough for it to try.

The documents that govern this are worth naming rather than paraphrasing. Anchoring to concrete is designed under ACI 318 Chapter 17 in the United States and under BS EN 1992-4, Eurocode 2 Part 4, Design of fastenings for use in concrete, in Europe. Any one product's capacity belongs to one qualified embedment and one installation procedure: mechanical anchors are qualified under ACI 355.2 and adhesive anchors under ACI 355.4, with the values in the product's ICC-ES evaluation report or its European Technical Assessment and the method in the manufacturer's printed installation instructions. Edge distance, spacing and whether the concrete is assumed cracked or uncracked all move those numbers, and the mechanism behind the edge-distance limit — the breakout cone — is worked through in the anchoring-into-concrete guide rather than here.

One structural point about pergola post bases, because it is routinely misread. A flat plate bolted to a slab is not a moment connection: it locates the foot of the post and resists uplift, and that is all it is honestly good for. What stops a freestanding pergola racking is the frame above — the beam-to-post joints, the knee braces, the rafters and purlins acting as a plane — so a kit with no braces and four flat plates is relying on those joints entirely. Worth knowing before somebody decides the braces are ugly and leaves them in the box.

The Count Is a Budget; the Positions Come From the Frame

Run the convention on the 2.4 by 3.6 metre shed — a 12.0 metre perimeter at roughly six-foot spacing — and it returns seven anchors. Seven does not map onto a rectangle. Four are corners, leaving three intermediates to share between four sides, and every way of doing that leaves one side longer than the rule allows. Build eight: a corner at each corner and one intermediate at the midpoint of every side, which puts the longest unsupported run at 1.8 metres. The count is a budget to buy against and a check on the kit's own hole pattern; the positions come from the geometry, and where they disagree the geometry wins.

The positions also have to land on something structural, which quietly removes most of the perimeter from consideration. On a sheet-metal or resin kit the anchor points are the pre-punched rail holes and there is nowhere else to put them. On a timber shed the strap has to reach a runner, a floor bearer or a wall stud — never the cladding and never the sheet deck, both of which tear out around the fixing long before the anchor moves in the soil. An anchor sitting 100 millimetres off the even spacing for that reason is the correct outcome, not a compromise.

As for how much any of it holds: an auger or screw-in ground anchor's capacity is a strongly soil-dependent figure published by its own manufacturer, and no general number could honestly be printed here for it. What can be said is where the chain usually breaks, and it is almost never the soil. It is the strap, the buckle, or the fixing through the strap into the timber — a single screw through a galvanised strap into a runner has a withdrawal capacity that is a published property of the fastener and the timber, and it is routinely the smallest number in the assembly. Design wind actions come from ASCE/SEI 7 or BS EN 1991-1-4 with its National Annex, and neither produces an anchor count for a shed; the manufacturer's instruction does, and a local requirement for accessory structures overrides both.