Sizing a steel warehouse: the five numbers that decide the frame
Most warehouse enquiries I receive arrive in one of two shapes. Either a single line — “we need a warehouse, 40 by 80” — or a full architectural floor plan with toilets, offices and door swings drawn to the millimetre, and no eave height anywhere on it.
Both are hard to price, and for opposite reasons. The first leaves a supplier guessing at everything. The second buries the five numbers that matter inside a hundred that do not.
The point of this note is to separate them, so that whatever you send next is answerable.
Why only five
A steel warehouse has more variables than five, obviously. Sheeting profile, purlin spacing, gutter type, bracing arrangement, insulation, skylight panels, roller doors, personnel doors, office partitions, mezzanine, sprinkler, roof-mounted plant — the list is long, and each one changes something.
But almost all of them change one thing: a surface, a fitting, or a lining. They are decisions taken on top of an already-fixed structure, and they can be revisited without touching the primary steel.
Five variables behave differently. They move the primary frame, the foundation, or the erection method at the same time. Change one of those after the drawing is signed and you are not amending a building, you are buying a second one.
That is the whole test: does this number change the steel, the ground it stands on, or the way it gets lifted into place? If yes, it belongs in the first conversation. If no, it can wait.
The five numbers
1. Clear width
The distance between column lines, measured to the inside face, at floor level. Not the same as the overall external width, and frequently confused with it — a couple of metres of discrepancy that silently eats a racking aisle.
Width is the single largest driver of member depth, because bending demand over a span rises faster than the span itself. The rafter does not merely get longer; it gets heavier per metre, and so does the column that carries it.
The practical consequence is that clear width is a cost curve, not a target. There is no maximum — you can build very wide — but there is a width beyond which the steel stops being the cheap part of the building. I have written about where that curve turns separately: how wide a warehouse can go without internal columns. The short version: ask what a column would actually block, not whether you can avoid one.
2. Bay spacing
The distance between consecutive portal frames, measured along the length. This is the number most often left blank, and it is the one with the most misleading intuition attached to it.
Fewer, wider bays means fewer frames, fewer foundations, and fewer crane picks — which sounds like free money. It also means each frame carries more roof area, so rafters and columns grow, and the purlins spanning between frames must be deeper or closer together. Somewhere in the middle there is a spacing where the total is lowest, and it is not the same spacing for a 15 m wide shed as for a 35 m wide one.
I have worked through that trade-off, including how length interacts with it, in bay spacing is a trade-off, not a default.
3. Eave height
The height from finished floor level to the underside of the rafter at the eave line. State which of those two references you mean, because “6 m high” has at least three meanings on a warehouse drawing and they differ by more than a metre.
Eave height does two things at once. It lengthens the column, which increases its effective buckling length, and it increases the wall area that the wind pushes against. Both effects grow faster than the height itself, which is why the difference between 6 m and 9 m is not half again as much steel.
It also fixes something that has nothing to do with structure: whether a truck can tip inside, whether racking can go to four levels, whether a mezzanine can be added later. The eave height decision is worth reading before you commit, because it is the most common late change on warehouse projects.
4. Design wind speed at the site, in km/h
This is the number supplied correctly least often, and the one where a wrong assumption is invisible until something goes wrong.
Wind load on a low-rise building scales with the square of the wind speed, so a modest difference in the assumed speed is not a modest difference in the frame. Two suppliers quoting the same shed at different assumed wind speeds are not quoting the same shed — and the lower of the two will look like the better price while being a different structure.
Ask for it in km/h, in writing, on the datasheet. If a supplier sends a wind figure quoted in some other unit, convert it yourself and confirm the conversion back in writing before comparing anything — unit mix-ups on datasheets are common enough that it is worth the one extra email.
“We do not have a local code” is not an answer, and it comes up constantly across the markets I work in. Wind still blows. What it means is that the design basis has to be agreed and written down rather than looked up — I have set out how that conversation should go in design loads when there is no local code.
5. What the frame has to carry
The fifth number is not one number but a short list, and it is the one owners most often assume is the supplier's problem. It is not, because only you know what the building will do.
- Roof imposed load. Access for maintenance, and anything mounted on the roof — solar panels, air handling units, dust collectors. A roof designed for sheeting and nothing else is not a roof designed for a PV array.
- Suspended services. Sprinkler mains, cable trays, ductwork, conveyor supports. These hang from the frame at points the frame was or was not designed for.
- Crane or monorail. If anything lifts inside, the bay it lifts in is a different structure. Even a future crane should be declared now, because the column and foundation for a crane bay are not something you add afterwards.
- Mezzanine. A floor inside a warehouse loads the frame at points chosen by the floor, not by the frame. Declaring it early is cheap; inserting it later usually means a separate structure standing inside the building — which is the subject of a mezzanine inside a warehouse.
Floor slab loading is the fifth number's close relative and deserves its own line. What the ground slab must carry — racking legs, forklift wheels, bulk stack, or a hit from a loaded truck — decides slab thickness and reinforcement, and it decides whether the racking can be fixed down. That is a separate note in this series; the point here is simply that the slab is not a constant.
The order to settle them in
They are not independent, so the sequence matters. In practice:
- What happens inside, and what moves. This is not one of the five, but it generates four of them. Traffic paths give you width and door positions; storage height gives you eave height; lifting gives you crane.
- Clear width, from the layout rather than from a preference.
- Eave height, from the tallest thing that has to move or stack, plus clearance.
- Bay spacing, which you can leave to the supplier's optimisation provided it does not clash with your door and partition grid.
- Wind speed, from the site, agreed in writing.
- The load list, as a written schedule rather than a sentence in an email.
Note that bay spacing is the only one of the five you can safely delegate. That is worth remembering when you are assembling a brief: three numbers you must own, one you can hand over, one you must obtain.
What a late change actually costs
The reason to spend an extra week on these five is not tidiness. It is that a late change propagates outward from the frame in every direction at once.
Push the eave height up by two metres after the frame is designed and you have changed the column length, the wind area, the cladding quantity, the gutter height, the downpipe runs, the door heights, and the crane capacity needed to erect it. Nothing in that list is a drawing edit. Each is a purchase order.
Push the clear width out by four metres and you have changed the rafter section, the column section, the base plate, the anchor bolts, and — in poor ground, which is most of the ground I am asked about — the foundation. The foundation is usually already poured by the time someone notices.
By contrast, deciding in month two that you want a different cladding colour, an extra personnel door, or a mezzanine office in the corner costs a revision and a small quantity change. That is the category the other twenty variables live in, and it is why they do not need to hold up the first conversation.
Two things that are not on the list but decide everything
Where it is going. Country and site, not because of paperwork, but because of wind, ground and transport. A design that is efficient for an inland site with a wide access road can be the wrong design for a coastal site with a narrow one, before you have changed a single dimension.
What it must not do. The exclusions. If your brief does not say whether foundations, erection, insulation, or electrical are inside the supplier's scope, two quotes for the same building will differ by a wide margin and neither will be wrong. This is the quietest cause of a bad comparison and it has nothing to do with steel at all.
I have written separately about how to write a building brief that can actually be priced, and about what really drives the cost of a steel building once the five numbers are fixed.
Where to go next in this series
The five numbers each have their own note, because each one deserves more than a paragraph:
- How wide a warehouse can go without internal columns — where the cost curve turns.
- Bay spacing is a trade-off, not a default — the only one of the five you can delegate.
- The eave height decision — the most common late change on warehouse projects.
- Design loads when there is no local code — how to agree a basis instead of looking one up.
- Why a steel warehouse rains indoors in a humid climate — the roof build-up decision, and the one rule that fixes it.
- The warehouse floor: what the slab carries — the other half of the load list, decided before the pour.
For the surrounding picture — how warehouses sit next to cold-chain, cross-dock and port transit buildings in this business — the logistics and warehousing notes collect those together.
Written by Johnny, who advises owners and project managers in Africa and South America on steel buildings. If you have a span in mind and no comparable quotes yet, send me the five numbers and I will tell you what I would change first.
Related: warehousing notes · all notes · what drives the cost
