Setting a Fence That Stays Up

The fence is a sail, and the post is its mast

Stand a six-foot solid board fence in an open paddock and you have built roughly six square feet of sail per running foot of run. Wind pushes on that face, the pressure resolves into a horizontal force at about mid-height of the panel, and every bit of it travels down the post into the ground as a bending moment. Nothing else in the assembly carries it. Rails carry the infill, infill carries the pressure to the rails, and then the whole story arrives at the post — a cantilever with one fixed end buried in soil that may or may not deserve the word fixed.

Once you see the fence that way, the arguments people have on site resolve themselves. Why does a 4x4 that was fine at four feet start leaning at six? Because sail area went up by half and the lever arm got longer at the same time, so the moment at grade climbed far faster than the height did. Why does the neighbour's identical fence stand while yours racks? Theirs sits in the lee of a garage and yours faces a fetch across two open lots. Why does the gate post always go first? It carries wind load plus a swinging dead weight on one side, permanently trying to pull it out of plumb.

Wind pressure is not a number you get to choose. It comes from the design wind speed for the site, the exposure category, and the height of the fence — and every one of those is set by the local building authority working from a wind-load standard, most commonly ASCE/SEI 7 Minimum Design Loads and Associated Criteria for Buildings and Other Structures. Coastal and hurricane-prone jurisdictions run substantially higher design speeds than sheltered inland ones, and some publish fence-specific prescriptive requirements that override the general method entirely. Ask the building department what governs before you dig, because the same fence detail that passes inspection in one county gets rejected two hours away.

Reading the site's wind before you read the plan

Walk the run before you price it. What you are looking for is fetch — the uninterrupted distance wind travels before it hits the fence line. A back garden ringed by mature trees and two-storey neighbours is a different structural problem from the same fence on the exposed side of a corner lot with a ploughed field opposite, and pretending otherwise is how you end up warranty-replacing posts.

Note the topography too. A run along the top of a bank or the crest of a slope sees accelerated flow; wind speeds up as it climbs and the fence at the top catches the fastest of it. A run at the bottom of the same slope sits in comparatively dead air. On a long property both conditions can occur within one contract, and the honest answer is to detail the exposed stretch heavier rather than average the whole job to a middle spec that is over-built at one end and under-built at the other.

Funnels deserve a separate look. Gaps between buildings, the throat between a house and a garage, alleys aligned with the prevailing wind — these accelerate local flow well above the open-field figure. Fences set across a funnel take a beating out of all proportion to their height. If the client insists on a solid panel across a gap like that, either shorten the bays, deepen the posts, or tell them plainly that a semi-open infill will survive and a solid one probably will not.

Then find the prevailing direction and check what the fence does about it. A run broadside to prevailing wind is the worst case. A run parallel to it barely loads at all until the wind veers, which it will. Corners and returns are quietly the strongest part of any fence because each leg braces the other, which explains why the middle of a long straight run, with no return within thirty metres, deserves your extra concrete.

Answer one: how deep the post goes

Embedment resists the overturning moment by mobilising lateral bearing in the soil, and that soil is the weakest link in the chain far more often than the post is. Dense gravel and stiff clay push back hard. Soft silt, loose fill, recently landscaped topsoil and anything within the influence of a nearby excavation push back very little. The hole depth that works in one is negligent in the other, and no table on the internet knows which one you are standing on.

Your first hole is the site test, and it costs nothing. Look at the spoil. If the auger walks through it, if the sides slump before you get the post in, if you hit made ground or old builder's rubble, you are in weak soil and the design assumption has changed. Deeper, wider, or both. Where you meet groundwater partway down, note the level — saturated soil loses a great deal of its lateral capacity, and a post founded in it is on a schedule.

Frost changes the question from strength to movement. In freezing climates the base of the footing wants to sit below the local frost depth so that the freeze-thaw cycle cannot lift it; that depth is set by the local jurisdiction's adopted code and varies enormously across a single country. Getting depth right for wind but wrong for frost produces a fence that stays plumb through a gale and then jacks itself crooked over one winter, which is a harder conversation with the client because there was no storm to blame.

Shape the concrete so it sheds water. A collar that finishes flat or dished at the top holds a puddle against the post for years — the specific mechanism that rots timber posts and rusts steel ones at the exact section carrying the most bending stress. Crown it, slope it clear, and keep concrete off untreated wood. Backfilling a gravel base under the post so water drains rather than pools takes a few minutes and buys years.

Answer two: how far apart the posts stand

Bay width is the lever you have when depth is constrained. Every extra foot between posts adds directly to the tributary sail area one post must carry, so the moment at each footing scales with it. Tightening from eight-foot bays to six on an exposed run cuts the load on each post by a quarter without touching the hole depth — often cheaper and always faster than deepening every footing on the job.

Set the spacing to the infill module, not to a round number. Pickets, boards and panels each have a natural repeat, and forcing a stock panel into a bay it does not fit produces either a ripped-down filler board at every post or a gap that whistles. Lay the run out end to end first, decide where the odd bay lands, and put it somewhere it reads as deliberate — beside a gate, at a return, against the house — rather than letting it fall in the middle of the most visible elevation. Run the length and bay count through the calculator below before you order, because material quantities and post count move together and a change to one silently changes the other.

Rails matter more than people expect. The rail is what delivers the panel's wind load into the post, and on a tall fence a two-rail arrangement leaves the boards spanning far enough to cup and rack under pressure. Three rails on anything six feet and up separates an infill that stays flat from one that starts twisting fasteners out of the top rail after a couple of seasons. Fix the rail to the post with a connection that resists withdrawal, not just shear — end-nailing into end grain is not a connection under cyclic load.

Answer three: what the infill lets through

Solid infill maximises the sail. That is the whole trade-off, and it is the one clients understand least because they are buying privacy and hearing about aerodynamics. Say it plainly: a board-on-board or fully closed panel catches everything the wind brings, and standing it up needs deeper posts, tighter bays, or both. That cost is real and it belongs in the quote at the start, not in a variation after the first storm.

Spaced pickets bleed pressure through the gaps and load the structure far less for the same height. Shadowbox — alternating boards on opposite faces of the rails — gives near-privacy from an angle while leaving a genuine air path, and it is the detail worth proposing when someone wants a solid six-footer on an exposed corner lot. Lattice tops do the same thing at the level where the lever arm is longest, precisely where relief helps most. Fix the picket count and gap once the bay layout is settled, so the spacing that survives the wind also divides evenly across the bay.

Watch what the client adds afterwards. A reed screen, a windbreak mesh, a run of climbing plants trained across a spaced picket fence — each converts a permeable fence into a solid one without a single post being changed. The structure was designed for the fence you built, and it is now carrying a different one. Tell them, in writing if the job warrants it, that infilling the gaps voids the assumptions the footings were sized on.

Gates deserve their own paragraph. A gate leaf hangs its whole weight off one post as a permanent eccentric load, then adds wind load on top, then gets slammed. Oversize the hinge post relative to the line posts, deepen it, and brace or tie it to its neighbour where the layout allows. A diagonal brace within the leaf running from the bottom hinge corner up toward the latch keeps the gate square; the reverse diagonal in timber does nothing useful in tension and lets the leaf drop.

Where wind gets in through the details

Fasteners fail before members do under cyclic load. Wind is not a single push — it is thousands of reversals, and every one works a nail a fraction looser. Screws, ring-shank nails or structural connectors in the rail-to-post joint hold where smooth shank nails walk out. On any exposed run, spend the extra on connection hardware before you spend it on heavier timber; the timber was not the thing that was going to fail.

Corrosion picks the worst spot to work. Water tracking down a post face and sitting at the concrete collar attacks the section with the highest bending stress, and by the time you can see it the capacity has already gone. Match fastener metallurgy to the treated timber you are using — the wrong combination corrodes the fastener from inside the joint where nobody looks. Coastal salt air accelerates all of it.

Panel fixings are the quiet failure. A prefabricated panel clipped into channels or toe-nailed at the corners will lift bodily out in a strong gust, which is how whole panels end up in someone's conservatory. Positive fixings top and bottom of each panel, into the rail or the post, cost a handful of screws per bay.

Finally, check the finished run in a wind, not in still air. Push the middle of each bay hard with a flat hand at the top rail. Movement at the post base means the footing is not doing its job; movement in the panel with a solid base means the rails or fixings are. Both are cheap to fix on the day you built it and expensive to fix from a ladder in November after a call from an unhappy client.