How to Maximize Container Space: Orientation & Box Size

How many boxes you get into a container comes down to three things: which way each box is allowed to face, how the box dimensions line up with the inside cross-section of the container, and whether the cargo goes on pallets. In my experience, the gap between a sloppy plan and a careful one is somewhere around 10–25% of the internal volume, which on a 40’ high cube (76.3 m³) is roughly 7.6–19.1 m³. Those decisions belong in the load plan, before the container shows up. If you’re deciding orientation and packing method at the dock, you’ve already lost the space.

Below are the ways to maximize container space, roughly in order of impact, followed by a worked example and the mistakes that cost the most.

1. Let the boxes rotate#

Many plans lock every box upright by default. Allowing all six orientations usually adds 3–10 percentage points of utilization, as a rule of thumb. Take a 50 × 40 × 30 cm box in a 40’ high cube. Kept upright, 1,080 boxes fit (84.9%). With rotation allowed, 1,128 fit (88.7%). That’s 48 more boxes, or 2.88 m³ (+3.8 points), in the same container at the same freight cost.

Some cargo really must stay upright (this side up): liquids, glass, machinery. But “this side up” arrows get printed on plenty of boxes out of packaging habit. The CTU Code tells packers to observe handling marks such as “this side up”, so if the box carries the mark, follow it. If the product doesn’t care which way it faces, take the mark off the packaging spec and state in the load plan that the boxes may be rotated.

2. Size the box to the container width#

This is the biggest lever you have, because the inside cross-section barely changes from one container to the next. ISO 668 fixes the external dimensions (every size shares the same 2.438 m width) and sets minimum internal dimensions, so inside width and height don’t vary much. Our calculator uses an inside width of 235 cm and a height of 239 cm for a standard dry container, 270 cm for a high cube, and a length of 1,203 cm for 40’ units.

Now look at the width. At 62 cm wide, only 3 boxes fit across (235 ÷ 62, rounded down), using 186 cm and leaving 49 cm, or 20.9% of the width, empty. Trim it to 58 cm and 4 fit, using 232 cm with 3 cm to spare. One more box per row is a third more boxes (4 ÷ 3 = +33%), and that gain repeats down the full 1,203 cm of length.

If you control the outer box, the case that holds the inner packs, test candidate sizes in the Container Loading Calculator before you order molds or packaging.

3. Floor load when the math works#

Floor loaded means the boxes are stacked directly on the container floor, with no pallets. Pallets cost space: the decks and the gaps around them typically take 10–20 percentage points of utilization, in my experience (the numbers are in how many pallets fit in a container, and the layout on each pallet in how many boxes fit on a pallet).

The catch is unloading. A floor loaded 40’ takes a crew hours to empty by hand; on pallets, the same cargo comes off in minutes with a forklift operator. So the real question is whether the freight you save per container is worth more than the extra handling cost at destination. As a rough guide, floor loading often wins for high-volume, low-value goods on stable lanes. If the cargo is going into a modern distribution center (DC) network, you often won’t get a choice: many DCs require pallets.

4. Check weight limits and weight distribution#

Before you chase space, check whether weight will stop you first. Our calculator uses a payload of 28,800 kg for a 40’ standard container and 28,620 kg for a 40’ high cube. Payload is the maximum gross weight minus the container’s tare weight, so carrier figures vary with the rating: one carrier lists 28,750 kg for a 40’ standard and another 26,840 kg; for a high cube, 28,600 kg and 28,800 kg. On the road, the limit usually comes sooner. For a 40’ on US roads, my working estimate is roughly 19–20 t of cargo once you allow for the truck and chassis. That’s a rule of thumb, not a legal limit. The detail is in our container weight limits and VGM guide.

Where the weight sits matters as much as how much there is. The CTU Code (the IMO/ILO/UNECE Code of Practice for Packing of Cargo Transport Units) sets these guidelines for a container that is lifted level:

• Length: the combined center of gravity of the cargo should sit close to the middle of the length. Normally it shouldn’t be more than ±5% off-center; the rule of thumb is that 60% of the total cargo weight sits within half the length. In specific circumstances, such as with special spreaders, up to ±10% may be acceptable.
• Width and height: keep the lateral center of gravity near the middle of the width and the vertical center of gravity below half the height of the cargo space.
• Stacking: heavy boxes on the bottom, light ones on top. The code says not to stow heavy goods on top of light goods. That keeps the lower boxes from being crushed by the weight above them (crush damage) and keeps the stack stable when the ship rolls.

Load too much weight toward the front and the axle weight (the weight carried on each axle of the chassis) can go over the limit even when the total weight is legal. On US Interstates, for example, a vehicle can weigh 80,000 lb in total, but a single axle is limited to 20,000 lb and a tandem axle to 34,000 lb. The code also warns that weight limits along the route can be well below the permitted gross weight.

The CTU Code is non-mandatory. It was approved in 2014 by the IMO Maritime Safety Committee, the UNECE Inland Transport Committee and the ILO Governing Body as a revision of the 1997 guidelines, and it covers the whole intermodal chain, by sea and by land.

5. Fill the gaps, then brace what’s left#

Two gaps come up in almost every load.

• The top layer. The top 20–30 cm is often too low for another full layer. If you ship more than one product (SKU), that’s where the small, light boxes go.
• The void at the door. Any space you can’t fill with cargo has to be filled with dunnage (air bags, timber) and braced. Cargo that shifts in transit damages itself and can fail inspection. The CTU Code tells packers to fill void spaces when necessary and to use blocking or lashing.

Don’t chase the last few percent. In my experience, once you push past about 90% of the theoretical pattern, the extra labor usually costs more than the freight you save. Boxes forced into odd spots slow down loading and arrive crushed.

6. Simulate the load instead of dividing volumes#

Container volume divided by box volume is a ceiling, not a count. The real count can match it but never beat it, and it usually falls short, because the division ignores orientation, how the boxes line up against the walls, and the door opening. As a rule of thumb it overstates the count by 10–25%; in the example below, it’s 32% over the grid count and 22% over the calculator’s.

The Container Loading Calculator runs in your browser and places each box in an actual position, mixing orientations where rotation is allowed. It shows the result in 3D, with the box count, utilization and the limiting factor (volume, weight or quantity), so you can see whether weight or space is the real bottleneck. It also checks whether each item fits through the door opening, and it rates the weight distribution by how far the cargo’s center of gravity sits from the middle, front to back, side to side and in height.

Worked example: one SKU, three orientations#

Take one SKU packed in a 60 × 40 × 40 cm outer box (0.096 m³) weighing 8 kg, going into a 40’ high cube with internal dimensions of 1,203 × 235 × 270 cm. Here’s what a single-orientation grid gives you for each way the box can sit:

The worst orientation costs 60 boxes, or 5.76 m³. The cause is the 60 cm side running across the width: only 3 fit (235 ÷ 60, rounded down), leaving a 55 cm strip empty along the whole container.

With every box facing the same way, 600 is the best you can do, but it isn’t the best possible. With rotation allowed, the calculator mixes orientations and fits 650 (81.8%), 50 more than the best grid. If the boxes have to stay upright, the calculator also lands on 600 (75.5%).

Weight isn’t the constraint here. 600 boxes at 8 kg come to 4.8 t (650 boxes, 5.2 t) against a 28,620 kg payload, so this load runs out of space long before it reaches the weight limit. Every slot you win back is freight you don’t pay for.

Now check the actual order: 500 boxes. The CBM Calculator gives 500 × 0.096 = 48 m³, which is one 40’ high cube at 62.9% of nominal capacity, rated as a fit. The Container Loading Calculator loads all 500 at 62.9%, with quantity as the limiting factor. It doesn’t show how many more boxes would fit, but you can work it out: room for 100 more if you go by the 600 grid, or 150 by the calculator’s maximum of 650. That’s room for another product, and it can save you a second booking.

Mistakes that waste the most space#

• Trusting the volume division. 76.3 m³ ÷ 0.096 m³ = 794.8, so someone books 794 boxes as “fitting”. That’s 194 boxes (about 18.6 m³) more than the 600 grid, or 144 (about 13.8 m³) more than the calculator’s 650. That cargo has nowhere to go. It either ships LCL, which costs more per CBM than FCL (see FCL vs. LCL), or it waits for the next vessel. For realistic loaded CBM by container type, see how many CBM fit in a container.
• A load plan that never reaches the dock. You plan 600, the crew stacks the boxes the way that feels natural, and 540 go in. Now 60 boxes go back to the warehouse, get handled again, and the paperwork has to be redone. Give the loading supervisor a one-page drawing that shows which way the boxes face.
• Forcing the last few boxes in. Boxes wedged sideways into leftover gaps arrive crushed. If the bottom layer of a six-high stack gets crushed, it can turn into a damage claim for that whole layer.
• Optimizing orientation for dense cargo. Dense cargo can reach the practical road limit for a 40’ in the US, roughly 19–20 t by my estimate, before the floor is even covered. Once weight is the limit, a better orientation gains you nothing. Work out which limit you hit first, then look at the pattern.