Shipping Container
Roof Types

Everyone looks at a steel box and assumes the strength is uniform. It is not, it is not close, and every decision about container roofing follows from that one fact. A container floor is built to take a loaded forklift. The roof is built so two people can walk on it while lashing cargo.

So the real question is never what roof does my container need. The container has a roof. The question is what job a second roof is being asked to do — shed water, carry snow, stop condensation, or span between two boxes — because those four jobs point at four different answers, and most people buy a roof shape before they have worked out which one they are solving.

The Lid Is Not a Roof

A shipping container is engineered with extraordinary precision for one job: being stacked nine high on a rolling ship and lifted by its corners. Every structural decision follows from that, and almost none of those decisions produce a good building roof.

Where the strength actually is

Vertical load travels through the four corner castings into the four corner posts. That path is enormously strong — published structural analysis of stacked containers puts roughly 116 kips per corner, about 116,000 pounds. The floor is likewise overbuilt, designed for a 250 psf live load. The roof is the exception, and it is a dramatic one.

The test result everybody misreads

Containers are tested under ISO 1496, and the roof strength test — commonly cited as TEST 07 — applies 660 pounds over an area of 2 feet by 1 foot. That figure gets quoted constantly as evidence that container roofs are strong, and the arithmetic seems to support it: 660 pounds over 2 square feet is 330 psf.

Where the 20 psf figure comes from, and how much to trust it

In an engineering discussion of exactly this question, one structural engineer calculated roughly 20 psf capacity for the corrugated steel roof panel under simple-support assumptions. Another argued that actual capacity exceeds the simple calculation, because the panel distributes load in more than one direction and because passing TEST 07 implies some reserve.

Both are likely right, and the disagreement is instructive rather than disqualifying. The engineers disagree about the number and agree about the direction: the container roof panel is at the low end of building-roof capacity, it was never intended as a load-carrying deck, and anything beyond incidental load needs a structure above it.

It also ponds

Container roofs are not flat, but they are not usefully sloped either. The corrugations run across the width and there is a slight crown, and none of that is enough to reliably move water off a roof that has settled even slightly out of level. Water stands.

Standing water is the dominant long-term failure mode in container construction, and it compounds in a specific way. Water weighs 5.2 psf per inch of depth. Four inches of ponded water is 20.8 psf — by itself at the estimated capacity of the panel. And ponding is progressive: load causes deflection, deflection deepens the pond, the deeper pond adds load. It is the one roof condition that manufactures more of itself.

What a Second Roof Is Actually For

Before choosing a shape, work out which of four jobs you are buying. They are separable, they have different answers, and buying the wrong one is how people end up with a handsome gable over a container that is still rusting from the inside.

Where a Roof May Touch the Container

This is the part with real engineering in it, and it is settled: load goes to the corners, attachment goes to the top side rails, and nothing structural ever bears on the roof panel.

The three surfaces

Engineering guidance is consistent on the remedy: rather than relying on roof capacity, strengthen the roof structure, slope it for drainage, and over-frame to redirect loads to the corner fittings — the container's primary load path. That is the same principle that governs what happens under the container, running in the opposite direction.

The three attachment methods

On the CSC argument, and who is making it

Clamp manufacturers argue that welding and drilling invalidate the container's CSC plate — the safety approval certifying it for international shipping — and that clamps preserve it. That is factually correct and often irrelevant.

The Eight Roof Types Compared

In rough order of structural ambition.

Exposed container roof — the default, and the expensive one

Doing nothing is a choice, and in a dry climate on a container used for storage it is often the right one. As a building roof it is the configuration that generates the most long-term cost, because it accepts ponding, offers no snow capacity, and gives condensation nowhere to go.

If you are leaving it exposed, the mitigations are drainage and coating: shim the container so it actually drains in one direction, keep the roof clear of debris that dams water, and maintain an elastomeric or similar coating on a schedule. That is maintenance forever rather than a one-time build, which is the honest trade. Our container painting and coating guide covers the surface prep that makes a roof coating last.

Low-slope overlay — the underrated answer

Tapered sleepers — ripped timber or light steel — run along the container roof from one top rail to the other, building a slope of an inch or two over the eight-foot width. A deck goes on the sleepers, and a membrane or standing seam panel goes on the deck.

It solves ponding completely, creates the vented cavity that solves condensation, adds very little weight, and barely changes the building's wind profile. It is the highest ratio of problem solved to money spent in container roofing, and it gets skipped almost universally because the finished result does not read as a roof.

Shed — the workhorse

A single sloping plane. Rafters bear on a tall ledger at one top rail and a short one at the other, or on a knee wall at the high side. Described in the container-building literature as extremely cheap, very simple, and buildable in a couple of days.

Why it wins: the simplest framing of any true roof, easiest to get watertight, no ridge, no valleys, no hips to flash.

Solar. A shed roof gives one large uninterrupted plane at a consistent angle, which is the best possible surface for panels. If solar is in the plan, that is a strong argument for shed over gable on its own.

Overhangs. Extend 12 to 18 inches past the container to keep water off the walls and doors. Container side walls are corrugated and shed water straight down into the base rail, so eave overhang genuinely matters here.

The limitation: on a single 8-foot-wide container a shed roof reads as an afterthought, and the high side needs a knee wall that has to be framed and clad.

Gable — the one people want

Two planes to a ridge. It is the shape a house is supposed to be, and that is most of why it gets chosen. It sheds water and snow well, it creates attic volume that is excellent for insulation and for hiding mechanicals, and it makes a container building read as a building rather than as a container. Our A-frame roof case study walks a real 40ft gable build from framing to materials list.

Hip

Four sloping planes. More framing, more flashing, no gable ends. The reason to choose it over a gable is wind: a hip roof has no flat vertical end wall for wind to push on, and performs measurably better in high-wind and hurricane-exposed regions. The reason not to is that it is the most complex framing on this list and it eliminates the usable attic volume a gable gives you.

Butterfly

Two planes sloping inward to a central valley, usually for rainwater collection or for a particular architectural look. It works, and the entire roof depends on one drainage path staying clear. A blocked valley on a butterfly roof does not drip — it fills. If you build one, oversize the valley, oversize the outlet, fit overflow scuppers, and accept that it is a roof with a maintenance schedule.

Living / green roof

The load numbers end this conversation quickly for an unmodified container. Saturated growing medium is heavy, and the load is permanent rather than seasonal. Against an estimated panel capacity near 20 psf, even a shallow sedum tray assembly is far outside what the container roof can carry.

It is entirely achievable over a purpose-built structural deck spanning to the corner posts — which is to say, a green roof on a container is a structural engineering project that happens to have plants on it, not a roofing choice. Budget and plan it as the former.

The Breezeway, in Detail

Two containers set parallel with a gap between them, and one roof spanning the whole assembly including the open space. It is the configuration that makes container building feel like architecture rather than like stacked boxes, and it is structurally the most demanding thing on the list.

Why the form works

The breezeway is a container-era rediscovery of the dogtrot house — a Southern vernacular form of two cabins under one roof with an open central passage. The passage is shaded, it is covered, and because it is open at both ends it accelerates airflow between the two volumes. In a hot, humid climate that central space is the most comfortable part of the building for much of the year.

It maps onto containers almost too neatly. Two 8-foot-wide boxes are exactly the wrong width to live in comfortably and exactly the right width to flank something. The gap does the work the containers cannot.

Span

Practical guidance puts the useful range at 10 to 20 feet between containers, with about 15 feet as the balance point for most projects. The material changes across that range:

The failure mode: differential settlement

This is the one that ends breezeway projects, and it has nothing to do with the roof.

You now have two independent foundations carrying one rigid structure. If one container settles relative to the other, the roof spanning between them is twisted. That twisting opens seams, shears fasteners, and breaks the weather seal along the length of the building. Guidance on the configuration is blunt about the tolerance: both containers on a unified foundation, set with laser-level precision, because even an inch of differential height introduces structural stress.

Bracing

A breezeway roof has a large covered opening in the middle of it and two supports that are free to move independently. Diagonal cross-bracing in the plane of the roof is what prevents racking under wind load, and it is the detail most likely to be omitted on a DIY build — because a braced roof looks over-engineered right up until the first serious storm.

Condensation and the Vented Air Gap

This is the job a second roof does that nobody buys it for, and it is the one that determines how long the container lasts.

The steel is the condensing surface

A container roof is a large, thin, uninsulated sheet of steel with conditioned air below it and outdoor temperature above it. When the steel drops below the dew point of the air touching it, water condenses on it. In a container this is so routine it has a name in the shipping industry — container rain — and in a building it means water forming on the underside of the roof, running down, and collecting where you cannot see it.

What the vented version looks like

• An air gap between the container roof and the underside of the new roof deck. The sleepers or rafters that create the slope also create the gap.
• Intake at the low edge and exhaust at the high edge, so air actually moves through the cavity rather than sitting in it. A gap that is sealed at both ends is not ventilation.
• Soffit venting with insect screen. Guidance for container roofs specifies at least an inch of air gap in the soffit, covered with wire mesh — airflow in and out, while keeping wasps and rodents out of a cavity you will never inspect again.
• Insulation below the container roof, not above it, in most retrofit cases, so the steel stays on the outside of the thermal envelope and dries to the ventilated cavity above. Our insulation guide covers the assemblies and the dew-point arithmetic in detail.

Get this right and the container roof becomes a protected interior surface with dry air moving over it, which is the single best thing you can do for the lifespan of the box. Get it wrong and the new roof becomes an umbrella over a corrosion chamber.

Snow, Water and the Numbers

Where the structural case for a framed roof is real, these are the figures that make it real.

Snow, by depth and density

Snow load is depth times density. Density varies enormously — fresh powder runs about 7 pounds per cubic foot, settled snow around 15, and wet or packed snow 25 or more. That range matters more than the depth does.

Shaded figures are at or beyond the estimated capacity of an unsupported container roof panel. Read the table and the conclusion is uncomfortable but clear: roughly a foot of wet snow, or a foot and a half of ordinary settled snow, reaches the number. That is a normal winter over a large part of the country, not an extreme event.

Standing water

And unlike snow, water does not wait for weather. A roof that ponds four inches ponds it every time it rains, and the deflection that allows the pond makes the next pond deeper.

Choosing Yours

The order to decide in

1. Work out which of the four jobs you are solvingDrainage, snow, condensation, spanning. Be honest, because the answer sets the budget.
2. Check your actual snow load with the building departmentRather than assuming. If it is negligible, the structural argument for a framed roof disappears and the drainage argument remains.
3. Decide the roofing material and the pitch togetherThe material sets the minimum pitch, not your preference.
4. Plan the attachment to the top rails and corner castingsNothing bears on the roof panel. Nothing.
5. Design the vented cavity before you design the roof shapeIt is far easier to add ventilation to a plan than to a finished assembly.
6. If it is a gable or a breezeway, get it engineeredUplift and differential settlement are both outside what rules of thumb cover.

Numbers Worth Carrying

A Note on Sources

The source landscape here is worse than usual. Container roofing content is dominated by two kinds of publisher: companies selling containers or container products, and DIY blogs working from each other. Neither is a good source for structural numbers, and the structural numbers are the part that matters.

The load figures above therefore come from structural engineering discussion and from a structural engineering trade publication rather than from container vendors. Those sources are more reliable and also less tidy — the engineers in the cited thread disagree with each other, and that disagreement is reported here rather than resolved, because resolving it would mean inventing a precision nobody has.

Snow, ponding and roof-area figures on this page are computed arithmetic: snow as depth × density, ponding at 62.4 lb per cubic foot giving 5.2 psf per inch, areas from nominal container dimensions. Reproduce them rather than cite them.

FAQ

This guide covers published figures and general practice. Snow load, wind exposure and uplift requirements are all site-specific and code-governed — your building department has the governing numbers. Any framed roof on a habitable container structure, and any breezeway span, should be designed by a licensed engineer for your location.