Silos and the mill settle differently, so the link must move
A grain plant looks like one complex on a drawing. Structurally it is at least two, and often three: the silo battery, the mill or the dryer, and the receiving shed. They share a site, a layout and an owner. They do not share a foundation, a load pattern or a settlement history.
That matters because something has to connect them. Grain moves between the silos and the mill by conveyor, and the conveyor needs a bridge, and the bridge has two ends. Those two ends are where a plant that looks fine on the general arrangement turns into a maintenance problem three years later.
Two structures, not one
A silo battery is a set of vertical cylinders carrying stored product. Its load is dominated by the product, it is applied over a compact footprint, and the resulting pressure on the ground is high and concentrated.
A mill is a framed building carrying machines on floors. Its load is spread over the footprint, and a large part of it arrives in the first months of operation as equipment is installed and filled.
The two therefore load the ground differently from the first day, and the ground responds differently under each. This is not a defect and not something a better foundation necessarily removes — it is a normal consequence of putting a heavy point load next to a lighter spread load.
Settlement is a design input, not a defect
If a geotechnical report is available it will give an estimate, and that estimate should reach the steel drawings. If it is not available, the honest position is that the differential is unknown, and the connection should be detailed to tolerate movement rather than to resist it.
The failure mode when nobody decides this is predictable: the bridge is drawn as a rigid box fixed to both structures, because that is the simplest thing to draw and the cheapest thing to fabricate. Then the ground moves, and the bridge becomes the member that transfers the silo's settlement into the mill frame. The symptom appears far from the cause — cracked blockwork at the far end of the mill, a door that stops closing, bolts that come loose on the opposite side of the building.
What the link has to do
A conveyor bridge between two structures that can move relative to each other needs one end that carries the vertical load and the horizontal restraint, and one end that carries the vertical load and slides. That is the same logic as a gallery between towers, and it is worth reading as one problem rather than two.
Practically: a fixed bearing at one end with the bracing arranged around it, and a sliding or slotted bearing at the other, with a travel allowance that covers the estimated movement plus a margin for the estimate being wrong. The sliding end also needs something to stop it walking off the bearing — a keeper, not just a pin.
Silos are their own wind case
A tall battery of cylinders is a different aerodynamic object from a low shed, and the two do not experience the same pressure distribution. On a plant where the mill is wide and low and the silos are tall and narrow, a single assumed wind value applied to both is a simplification that hides a real difference.
The number to use is the one that applies at the site, expressed in km/h, taken from the local code that governs there. What matters for the steel is not only the magnitude but the shape: the silo battery attracts load on a small plan area and delivers it into a small foundation, while the shed spreads a similar total over a much larger one.
The receiving side is a third structure
Where a receiving pit and shed sit in front of the silos, they are a third set of loads — a pit is a retaining structure, the shed over it is light, and the truck traffic is a live load nobody models well.
That pit also fixes the level of everything downstream, which is why it is worth settling before the silo layout is frozen rather than after.
What I want before drawings
- The geotechnical report, or an explicit statement that there is none.
- Silo diameter, height, capacity and whether they are on a common raft or individual pads.
- The conveyor route, with the height of each end and the travel direction.
- Whether the plant intends to add silos later, and on which side.
- The wind speed for the site in km/h from the applicable local code, not a round number.
- Who owns the interface — because on a plant with three structures, the connection nobody owns is the one that fails.
None of this is exotic. It is simply the difference between a plant that was designed as a system and one that was designed as three purchases.
Written by Johnny Joe, who advises owners and project managers in Africa and South America on steel buildings. If this decision is live on a project, send the layout and the constraint that worries you and I will tell you what it changes.
Related: expansion joints in a long building · a gallery restrained at one tower · the receiving pit decides the building · eave height
