Siting a Field Warehouse on Flood-Prone Ground

Tony Miller

6 days ago

7 min read

A warehouse in flood-prone ground needs its floor set above the highest recorded flood line, plus a safety margin, on a plot with a real access road, not on whichever patch happens to look driest the day you visit. Get the site wrong and the structure, the stock and the road out all fail together.

What is the first check before you commit to a site?

The highest recorded flood line, not this year's dry-season appearance. A plot that looks fine in September can sit under half a metre of water in August, and the people who know that are the ones who have lived through more than one wet season on that ground: county officials, chiefs, longstanding NGO staff, the boda-boda riders who stop using a road every year at roughly the same time. Ask them where the water reached last time and the time before, and treat "it never floods here" from anyone who has not seen a full rainy season on the site as unverified.

Where that local knowledge is thin or contested, UNHCR's own flood-risk guidance for humanitarian settlements pushes the same principle further: site selection should rest on a proper flood-risk assessment against local topography and hydrology, not a single visual inspection, and where risk cannot be brought down to an acceptable level the answer is to relocate the site rather than engineer around a bad one. For a warehouse that is exactly the trade-off: an elevated structure on a low-lying plot still floods at the access road even if the floor stays dry.

Walk the boundary before you sign anything. A site that slopes gently away from a river or seasonal watercourse, sits visibly above the surrounding land, and has no old flood debris, waterlines on trees, or dead vegetation at the base of its slope is a far safer starting point than a flat plot near a road junction that is merely convenient.

How high should the hardstanding and plinth sit above the flood line?

Above the highest recorded flood line, plus freeboard, the margin engineers add because no flood record is ever a guarantee against the next one. It is a well-established idea in floodplain engineering generally: in the United States, FEMA sets one foot (0.3 m) above the base flood elevation as its minimum encouraged freeboard for structures in a mapped flood zone, and notes that communities regularly build in more once wave action, debris and an unreliable flood record are accounted for.

South Sudan's rivers are not gauged and mapped to that standard, so the flood line you are working from is usually an estimate built from local testimony and the last two or three seasons, not decades of recorded data. On that basis, treat the highest reported flood mark as a floor, not a ceiling, and build the hardstanding and any plinth a working half-metre above it as a minimum, more where the record itself is uncertain or where the site sits inside a known river system rather than beside a minor seasonal channel. A raised, compacted hardstanding pad, laterite or murram over a drained sub-base, does more for a warehouse's flood resilience over its working life than almost any other single decision, because it protects the floor, the stock and the foundations of the structure together.

How do drainage and bunding keep water off the stored stock?

By moving water away from the stock faster than it can accumulate under it. A hardstanding pad needs a real fall, not a flat compacted surface, so that rain sheds to a perimeter drain rather than pooling against the base of the structure. That perimeter drain then needs somewhere to go: a natural watercourse or a constructed channel that does not simply return the water to the site during a flood peak, which is a common design failure on flat ground.

Bunding earns its place around anything you cannot afford to lose to standing water: fuel stores, chemical stock such as water-treatment tablets or flocculant sachets held ahead of distribution, and any area where forklifts or pumps sit when not in use. A low perimeter bund, even a modest one, keeps surface water pooling outside the working area rather than inside it, and buys hours during a flash flood that matter more than the bund's height suggests.

What access road and turning circle does the site actually need?

One that survives the wet season, which is the constraint that rules out more otherwise-good sites than flood height does. South Sudan has roughly 20,000 km of roads, of which only 400 km are paved, and most rural roads become inaccessible once the rains set in. A warehouse on perfectly dry, elevated ground is still useless if the last two kilometres of feeder track turn to black-cotton mud the first week it holds this year's stock.

Score the access road on the same visit as the plot: is it built up above the surrounding land, does it have working culverts rather than simple dips through seasonal watercourses, and has it actually carried a loaded truck in the last wet season, not just a motorbike. For the yard itself, plan the hardstanding and apron around the vehicle you will actually receive, not an average one; a standard humanitarian Mobile Storage Unit measures 24 x 10 x 5 m, a 240 m² footprint holding around 840 m³, and an articulated truck needs a genuine turning circle to reach it and leave again without reversing blind across soft ground, which is exactly where trucks get stuck.

How do you anchor a structure on soft or saturated ground?

Not with simple stakes driven into wet clay, which is the single most common cause of a modular structure failing in the first storm after it goes up. Soft or seasonally saturated ground needs a proper compacted sub-base under any anchor point, driven or screw-type ground anchors set well below the depth that saturates first, and, where the soil genuinely will not hold an anchor, ballasted or water-filled counterweights instead of relying on the ground at all.

Test the ground before you build, not after: dig a test pit at a few points across the footprint at the wettest part of the year you can manage, or ask locally where the water table sits at the peak of the rains. A structure anchored into ground that turns to soup for three months of the year needs a different foundation approach entirely, typically a raised, compacted and drained pad carrying the anchor points rather than the natural soil carrying them directly. This is where a warehouse partner who designs, ships and erects the structure as one job earns its keep: the foundation design has to match the actual ground, not a generic spec sheet written for a different climate.

Is it worth trading distance to the population served for higher, safer ground?

Usually, yes, because a warehouse that floods, washes its access road out, or sits unreachable for six weeks of the rainy season serves nobody at any distance. South Sudan's 2025 flood season is the scale to weigh this against: by late October, around 1,024,500 people were flood-affected across 29 counties, with Jonglei and Unity accounting for roughly 87% of that total. Those are the same low-lying counties where the shortest route to a population in need also tends to be the wettest ground on offer.

The honest trade-off is a few extra kilometres of access road, sited on higher ground with a reliable route, against a warehouse that sits closer to the population but cannot be reached, resupplied or evacuated for a third of the year. In most of South Sudan's flood-prone counties that is not a close call: a stock point that works for nine months of hard access beats one that looks convenient on a map and is unreachable for three. Where the population genuinely cannot be served from higher ground at all, the answer is usually a smaller forward stock point resupplied from a properly sited main warehouse, not a compromise on where the main site sits.

Frequently asked questions

How much higher than the flood line should a warehouse floor sit?

Above the highest recorded flood line plus a working margin, since FEMA's general floodplain-engineering minimum is one foot (0.3 m) of freeboard above the base flood elevation. Where the flood record itself is an estimate built from local testimony rather than gauged data, which is normal in South Sudan, build in more than that minimum, not less.

What ground conditions make a site unsuitable for a modular warehouse?

Ground that stays saturated for weeks at a time, has no natural fall for drainage, or shows old flood debris and waterlines above the level you plan to build on. A test pit dug at the wettest point of the year, or reliable local testimony on where the water table peaks, tells you more than a dry-season site visit ever will.

Does the access road matter as much as the flood height of the plot itself?

Yes. South Sudan has roughly 20,000 km of roads and only 400 km paved, and most rural roads become impassable once the rains start, so a warehouse on safely elevated ground still fails if the feeder road to it does not survive the wet season.

Should we build closer to the population we serve or on the higher, safer ground further away?

Generally the higher ground, because a warehouse that is unreachable, flooded or washed off its access road for part of the year serves nobody at any distance. Where the gap is too wide to close, run the higher main site with a smaller forward stock point rather than compromising the main warehouse's location.

How big does the hardstanding need to be for a standard relief warehouse?

Size it to the structure and its vehicle apron together, not the structure alone. A standard humanitarian Mobile Storage Unit is 24 x 10 x 5 m, a 240 m² footprint, and the apron around it needs enough compacted, drained hardstanding for an articulated truck to turn without leaving the prepared surface.