A fruit packing house is a wet building with a dry end

By Johnny Joe, export engineer at Hebei Hongji Shunda Steel Structure Engineering Co., Ltd. · 25 October 2026 · 6 min read

A packing house is usually described as a shed with a line in it. On the drawing it is one building with a span and a height. On the floor it is two buildings: a wet end where fruit arrives, gets tipped, washed, graded and packed, and a dry end where packed cartons wait for a truck.

Almost every packing house I have seen go wrong went wrong on the seam between those two, or at a detail that treated the whole envelope as if it had one climate.

One envelope cannot serve two climates

The wet end gets hosed. There is chlorinated water on the floor, brushes running, wax, fruit acid, and staff with a hose who are paid to use it. The dry end wants to stay dry: cartons lose strength fast when the air is humid, and a stack that has absorbed moisture fails at the bottom layer during transport, days later and a long way from the building that caused it.

So the boundary between the two has to be a wall with a door, not a change in floor finish. It also has to be a wall that a pallet truck can get through, which is where the drawing usually gives up and a plastic strip curtain appears instead. A strip curtain is a reasonable compromise; a change of floor colour is not a boundary at all.

Water attacks the wall base, not the roof

Insulated or single-skin, a wall panel ends somewhere near the floor, and that end is the most attacked part of the building.

If the panel sits on a kerb that is flush with the floor, washdown water runs straight under the panel and sits in the bottom rail. The fix is a raised plinth or a concrete upstand, so that the panel base is above the wet floor and water runs off the wall rather than behind it. It costs almost nothing at design stage and is expensive afterwards, because by then the floor is poured.

The second mechanism is chemical rather than physical. Chlorine and fruit acids attack cut edges, exposed fixings and scratch lines far faster than they attack a coated face. In a hall washed daily, the specification that matters is not the coating on the flat sheet, it is what happens at every cut and every screw.

And the old rule holds: nothing upward-facing below about 2 m (6.5 ft) that can hold water. Every horizontal ledge in a washed building is a reservoir.

The ceiling drips at the worst moment

A single-skin metal roof over a wet, humid interior is a condensation machine. Water forms on the underside of the sheet, collects along the low point of each profile, and releases in drops — usually onto the product, and usually on the shift when the outside air is warm and humid and the interior has been cooled by washdown.

There are three honest answers and one dishonest one. A liner panel with a drained cavity keeps the drips off the product but adds weight to the purlins, so it has to be in the frame design from the start rather than added at handover. An anti-condensation membrane absorbs and re-releases moisture and reduces dripping but does not stop the water being there. Ventilation that keeps the underside above the dew point removes the cause but is a decision about the roof and the walls together.

The dishonest answer is to specify more insulation and hope. Insulation slows the rate; it does not stop a cold sheet in a humid room from reaching dew point.

The dispatch dock is where the building meets the truck

The dock is a height, a slope, a canopy and a door, and each of them is a decision with a cost that is hard to revisit.

It is also the hygiene boundary. A dock that is open to the yard all day undoes the separation between wet and dry ends in one shift, no matter how good the internal wall is. Whether the answer is a dock leveller, a ramp, a level entry or a canopy over the queue depends on the truck fleet and the site fall — the same calculation that decides a raised dock or a level entry on any warehouse.

Canopy length is a queue calculation and canopy width is a cycle-time calculation, not a habit — the reasoning is the same as for a truck loadout canopy.

Floor flatness is a machine specification

"Flat enough for a forklift" and "flat enough for a sizing line" are two different numbers, and the second one is much tighter.

The practical issue is not the flatness itself but the joints. A floor has to be poured in bays, and every bay edge is a step the machine will find. The joint layout should follow the line layout: put the joints where the line can tolerate them, and let the pour sequence absorb the inconvenience rather than the machine.

This is one of the few decisions that has to be made before the frame is ordered, because the joints and the building grid want to agree.

Bins and cartons are not a uniform load

Bins of fruit are heavy, they stack high, and they stand in one place for hours. Cartons are lighter but they go higher and they lean against each other.

Both land as point loads on the slab, and near the wall they land on the same strip that carries the panel base and the washdown water. That is why the slab edge detail and the wall base detail are really one detail, and why a packing house is a poor candidate for a slab designed purely by area loading.

If a mezzanine is planned for storage or a future line, decide early whether it touches the frame — the same question as for a mezzanine inside a warehouse, and the same answer applies: a mezzanine that hangs off the columns changes the column design, and one that stands on its own does not.

What I want before drawings

• The line layout with machine footprints, and the direction fruit travels through the building.
• Where water is used: pressure, temperature, chemicals, and how often.
• Which end is wet, which is dry, and what separates them.
• Dock count, truck types, and whether trucks queue on site.
• Whether a liner panel or any ceiling is expected under the roof, decided now rather than at handover.
• Stack heights and bin types, so the slab can be checked where the load actually lands.
• The wind speed for the site in km/h from the applicable local code, not a round number — a wide, low packing shed with large dock openings is the shape that generates local suction at the eaves, and that is where storms find the weak fixing.

None of these are steel questions either. All of them change the steel drawing.