A packing list, a container list and a shipment schedule
The list is built from the drawing, not from the box
Every project starts with a component number, and that number is not a packing decision. It is a position: a specific member at a specific grid line, at a specific elevation, with a specific job to do when the frame goes up.
That one fact fixes the order of the paperwork. The packing list is generated from the erection drawing, and only then allocated to containers. It is never generated from the container and matched back to the drawing afterwards, because the moment you work that way the list stops describing a building and starts describing a shipment. Those are not the same object, and the difference only becomes visible at the far end of the journey, when someone is standing in front of a pile looking for one member.
What the packing list carries is short: the container number, the component names and numbers inside that container, the piece count, and the weight. Nothing on it can be estimated. A number that is approximately right is worse than no number, because it will be believed.
"The number points at the drawing" is also the reason a supplier cannot invent a marking scheme at the end of production. If the marks are assigned by the packing team, they describe the shipment; if they are taken from the drawing, they describe the building. Only the second one lets a site team put steel up without calling you.
Two documents, two questions
Alongside the per-container list there is one shipment schedule covering the whole project: every component in the job, its net weight, its gross weight, and its piece count.
The two documents are not a long version and a short version of each other. They answer different questions. The schedule answers is the building complete. The packing list answers what is inside this box.
Keeping them apart is deliberate, and the reason is worth stating. A complete schedule and an incomplete shipment can coexist perfectly happily, because the schedule counts what was made and the list counts what was loaded. Merge them into one sheet and that distinction disappears: a filled row looks identical whether the member is on the water or still standing in the yard. The moment you cannot see the gap between "made" and "loaded", you have lost the only place where the error was visible.
Where the three stop agreeing
Mismatches do not usually arrive as a single dramatic error. They arrive as small, individually reasonable decisions, each made by someone who could not see the other three documents. The recurring ones:
- A mark changed in the workshop. A member is remade or re-cut and picks up a new mark, while the schedule still carries the old one. Both are correct on their own day.
- A partial shipment. Part of a bay goes in the first sailing and the rest waits. The schedule is still complete; two packing lists now describe it.
- Consolidation at the port. Boxes are re-stuffed to use space, so the contents of container A end up split across A and B while the paperwork stays with A.
- Packages counted instead of pieces. Forty bundles is not forty members. Once those two numbers are used interchangeably in conversation, the list has stopped being a record.
- Gross weight drifting. Reinforcement added at loading is real weight that appears in no earlier document, and the first honest number arrives at the weighbridge.
None of these requires anyone to be careless. They only require that the paperwork is assembled after the fact instead of generated with the drawing.
Three counts, three different sets of eyes
The way we handle it is to count the same shipment three times, and the point is not redundancy. Each count is looking for a different kind of error.
The workshop counts what it made. This is a production check: is everything on the list physically present, painted, and marked. It catches the member that was made twice and the one that was never started.
The design department counts against the drawing. This is a completeness check, and it is the one that matters most, because it is the only count made by people who know what the building needs rather than what the order said. Completeness against the drawing, condition, and whether the sorting is clear enough for a site team to work from. A shipment can be complete and still be useless if the sorting is wrong, and only this count is placed to notice.
The trade department counts with a camera. This is an evidence check. The loading is photographed and filmed through the whole operation, and the material is sent to the customer before the truck leaves — not after arrival, and not only if something goes wrong. It answers a different question from the other two: not is it complete, but can we both see the same thing.
Three counts, three questions: was it made, is it right, and can we prove it. A single count, however careful, answers only the first.
Small parts: the five per cent that protects a programme
Bolts, connection plates and the other small items are where a full container and a stopped site can be the same shipment. They take up almost no volume, they are the first thing to be short, and a site cannot improvise them.
What they get before they travel: protection against rust, protection against water, and a unit-weight check in the workshop before dispatch, because the fastest way to find a short package is to weigh it. Each small package is opened and checked individually rather than trusted as a bundle. Large packages are reinforced last, after everything inside them is confirmed, so that nothing works loose in transit.
Then the part that is easy to skip: the consumable and easily damaged small items travel at about five per cent over the counted quantity.
That margin is not generosity and it is not a rounding habit. On a site, small items get dropped, cut wrong, over-torqued, or lost between two trucks. The cost of carrying five per cent extra is trivial next to the cost of a crew standing still waiting for a box of bolts to clear customs. It is a schedule decision that happens to look like a quantity decision.
The weighbridge is the last honest number
Before the truck leaves, the load goes over a weighbridge and the figure is recorded. This is the only number in the whole chain that nobody can talk its way out of.
Compared against the shipment schedule, it does two jobs at once. It confirms the loading is broadly what the documents claim — a container that is several tonnes light usually means something did not get on. And it produces the figure that the forwarder, the port and the customer will all eventually use, which stops three parties from working to three different totals.
At that point the paperwork is closed, the imagery is already with the customer, and the container is sealed. There is no later moment at which anyone can improve the record.
What to ask a supplier before the first container
- Are the component marks taken from the erection drawing, or assigned during packing? Ask for one mark and trace it back to a grid line.
- Is there a project-level schedule separate from the per-container lists, showing net weight, gross weight and piece count?
- Who counts, and how many times, and does anyone count against the drawing rather than the order?
- Will you send photographs and video of the loading before the truck leaves, as a matter of course?
- Is the load weighed before departure, and will you send the figure?
Five questions, and none of them are about steel. They are about whether the building on the drawing and the shipment in the water are still the same object by the time it matters.
Written by Johnny Joe, who advises owners and project managers in Africa and South America on steel buildings. If a packing list is what worries you on a live project, send one container list and the project schedule and I will tell you where the two would disagree.
Related: what a 40-foot open top can carry · erection sequence as a design output · modular or stick-built · temporary bracing at erection · what is in a shop drawing set
