Food miles are misleading: what actually drives your import carbon footprint

A retailer’s sustainability questionnaire has landed on your desk and it wants a carbon number for your programme. Or a buyer has said your Peruvian fruit “travels too far”, and you suspect that is wrong but cannot show it.

Distance is a poor guide to the carbon in a box of fruit. What decides it is the mode that carried the box. Moving one tonne of cargo one kilometre in a deep-sea container costs roughly 10 to 25 grams of CO₂e. The same tonne-kilometre by articulated lorry costs somewhere between 60 and 200 grams, and by long-haul air freight it costs 500 to 1,100. A road leg is therefore about five to ten times worse than a sea leg over the same ground, and an air leg is thirty to sixty times worse. Once you have those three numbers, most food-miles arguments reverse.

Distance is the wrong number #

Here are the published emission factors, well-to-wake — that is, including the energy used to make the fuel as well as to burn it.

Three things about that table.

The sea number is small and it is stable. A big container ship is the most carbon-efficient way anyone has found to move a heavy thing a long way, and it stays efficient whether the voyage is 3,000 km or 15,000. Distance on water is close to free in carbon terms compared with distance on a road.

The road band is enormous, and that is the number you can actually change. A lorry’s emission factor depends on the size of the vehicle, how full it is, and how often it runs back empty. The same road kilometre can cost 60 grams or 200 depending on those three things alone.

And the sets disagree with each other. DEFRA, the GLEC Framework and the academic literature all publish different figures for the same lorry, because they average over different fleets. None of that changes the ranking. The ranking is the only part of the table your buyer’s question depends on.

Three thousand kilometres by road, eleven thousand by sea #

Take two ways of putting winter fruit into a Rotterdam warehouse.

Morocco is the short one. Most Moroccan produce never touches a deep-sea container. Volume goes by refrigerated lorry from Souss-Massa, the growing region around Agadir, up through Tanger Med and across to Algeciras on a ferry, then overland from there. Agadir to Rotterdam that way is roughly 3,300 km of road. At 100 grams per tonne-kilometre — a well-loaded articulated reefer trailer, at the low end of the road band — that is about 330 kg of CO₂e per tonne of fruit.

Peru is the long one. Callao to Rotterdam through the Panama Canal is roughly 11,000 km by sea. At 20 grams per tonne-kilometre for a refrigerated container, that is about 220 kg of CO₂e per tonne, plus whatever road leg follows the discharge.

So the Moroccan fruit travelled a quarter of the distance and arrived with at least as much carbon attached. Use 150 grams instead — entirely normal for a trailer that runs back empty — and the Moroccan tonne rises to roughly 500 kg against the Peruvian 220.

This is not an argument against Moroccan fruit, which is grown where it grows for good reasons and reaches a European buyer in days rather than weeks. It is an argument against reading a map. The lorry legs at both ends of a chain, not the ocean in the middle, are where a produce import’s transport carbon usually sits. So a lane with a long sea leg and short road legs at each end can beat a nearby origin that arrives entirely by road.

For most produce, the farm beats the freight #

Transport is a smaller share of a food’s total footprint than almost anyone expects. Across the global food system it accounts for around 5% of emissions, and air freight — the part that dominates every conversation — carries about 0.16% of all food tonne-kilometres. Roughly 60% of them go by sea.

What dominates instead is the farm. How the crop was grown, what land it displaced, how much heat and light went into it — those are far larger than the shipping on most commodities.

Tomatoes are the clean example, because Europe grows them two completely different ways. A winter tomato from a heated northern glasshouse carries the gas that heated the glass. The same tomato from an unheated plastic tunnel in southern Spain or in Souss-Massa carries a long lorry ride instead. The lorry ride is much the cheaper of the two. Published comparisons put a Dutch January tomato at several times the footprint of a Spanish one delivered to the same shelf. The distance went up and the carbon went down.

You will also meet the New Zealand lamb study in this argument, because it is the one everyone quotes. Treat it carefully. Lincoln University published it in 2006, it measured energy rather than full greenhouse gases, and it was commissioned while the country was defending its exports against food-miles labelling in Britain. The direction it pointed has held up. The study itself is twenty years old and has been overtaken by the food-system work above, so cite that instead.

The same arithmetic makes the longest lane we cover look strange. New Zealand to Rotterdam is roughly 20,000 km and about six weeks at sea — around 400 kg of CO₂e per tonne, or the same as 4,000 km in a lorry. The fruit came from half a world away, and four days of European road haulage produces as much.

Air freight is the exception, and the exception matters #

Everything so far argues that distance is a weak signal. Air freight is where it becomes a strong one, because the emission factor is so large that the kilometres stop being cancelled out by anything else. Thirty to sixty times a sea leg is not a rounding difference. It is the whole footprint.

That means the single largest decarbonisation lever available to a fresh produce importer is not sourcing closer. It is moving a commodity off aircraft and onto water. One crossing achieves more than ten years of shortening road legs, and it saves money at the same time — which is not how sustainability levers usually behave. We keep the map of which commodities have crossed and which are next in the air-to-sea frontier.

The uncomfortable corollary is that the technology that keeps fruit alive for longer is doing more for produce emissions than local sourcing is. Controlled atmosphere, better packaging and faster services are what let cherries, blueberries and asparagus leave the aircraft. A reefer programme that buys a commodity four more days of quality life is a climate intervention, even though nobody sells it as one. New technology opens new lanes covers how those days are actually bought, and what buying them costs in spare days at the far end.

What the reefer adds on top #

A refrigerated box is not a dry box with a plug. Two things sit on top of the propulsion figure, and a serious estimate has to carry both.

Refrigeration energy. The unit runs the whole voyage, and on a container ship it is drawing from the vessel’s generators. The GLEC Framework handles this as a surcharge over the base container factor, in the region of 20% to 40% depending on how cold the cargo is held — a banana at 13 °C and frozen seafood at −20 °C are not the same load. A lower setpoint is an emissions decision as well as a quality one. It is not a decision worth taking on emissions grounds: the fruit’s specification wins that argument every time.

Refrigerant leakage. The gas the unit pumps around is itself a potent greenhouse gas, and units leak slowly over their lives. Much of the fleet still runs R134a or R404A, both with global warming potentials in the thousands. New builds use blends such as R452A, which is better but still over a thousand. The destination most of the industry expects is carbon dioxide itself, which scores one. We cover the phase-out and why it may have to happen twice under what is changing under the lid.

What that means for a questionnaire: a per-shipment figure built from a generic container factor is understating a reefer. And if the person who produced the figure cannot tell you which reefer surcharge they used, they did not use one.

What your buyer can actually require #

Here is the part most importers have not been told, and it changed recently.

The CSRD — the EU rule that made large companies report their sustainability data, including the Scope 3 emissions in their supply chains — was substantially rewritten by the Omnibus I Directive, which entered into force in March 2026. Two things in it matter to you.

First, almost no importer has to report. The obligation now falls only on companies with more than 1,000 employees and more than €450 million of net turnover, for financial years starting from January 2027. Your retailer is over that line. You are almost certainly under it.

Second, and this is the part worth knowing: a reporting company may not require sustainability data from a supplier of 1,000 employees or fewer beyond a defined, limited set. That set is aligned to a voluntary standard written for smaller businesses. You may declare that you are under the threshold, and the declaration is taken at face value. You have a right to refuse anything past the cap, and your retailer is treated as compliant for having respected it.

So the legal pressure on you is close to zero. The commercial pressure is not, and confusing the two is the mistake. A retailer that cannot compel a number will still prefer the supplier who has one, and listings are decided on preference rather than on directives. The importers who will find this easy in three years are the ones who started keeping per-shipment records now, while it was optional.

You are already paying for some of this, incidentally. Shipping sits inside the EU Emissions Trading System, and carriers pass the cost of their allowances on as a surcharge on European trade lanes — so next to the bunker fuel line on your carrier’s invoice there is now a carbon one. Wind-assisted ships are one of the ways it gets smaller.

Questions worth asking about your own lanes #

Six things decide the answer, and you can get all six from people you already talk to.

• What is the mode mix, door to door? Not just the sea leg: the whole chain, including the leg from the packhouse to the load port.
• How long are the road legs at each end, and do the lorries run back empty? This is the largest lever most importers have, and the only one they control directly.
• What share of the programme still flies? Including the early-season volume that quietly goes by air when the price justifies it.
• What setpoint is each commodity carried at, and is controlled atmosphere on the booking? Both change the reefer surcharge, and CA may be what keeps the cargo off aircraft at all.
• What is the carrier’s fuel programme, and are you buying into it? Biofuel and green methanol are sold as separate products with their own certificates. You do not get them by default.
• Which standard did the number come from? A figure with no method behind it is worth nothing in a questionnaire, and ISO 14083 or the GLEC Framework is what a retailer’s own consultants will be using.

Every one of those questions needs the same underlying thing: what each shipment actually did, rather than what the booking said it would do. Which vessel, which route, how many days, which legs by road. That record is easy to keep while a container is moving and close to impossible to reconstruct a year later, when the questionnaire arrives. It is the job Trackberry does on the containers you already have on the water.

FAQ #

Are food miles a good measure of a food’s carbon footprint? #

No. Distance is a weak signal because emissions per tonne-kilometre can be fifty times higher on one transport mode than on another. A tonne of fruit carried 11,000 km by ship can arrive with less carbon attached than the same tonne carried 3,000 km by lorry. Transport also accounts for only around 5% of global food-system emissions, so how the crop was grown usually matters more than how far it travelled.

Is sea freight better than road freight for emissions? #

Yes, by a wide margin per tonne-kilometre. A deep-sea container ship runs at roughly 10 to 25 grams of CO₂e per tonne-kilometre and an articulated lorry at 60 to 200, so sea is about five to ten times more efficient over the same distance. The road figure varies far more than the sea one, because it depends on the size of the vehicle, how full it is, and how often it runs back empty.

How much lower are sea freight emissions than air freight for produce? #

Around thirty to sixty times lower per tonne-kilometre. Long-haul air freight sits at roughly 500 to 1,100 grams of CO₂e per tonne-kilometre against 10 to 25 for a container ship. That is why moving a commodity from air to sea is the largest single emissions reduction available to a fresh produce importer, and why it also cuts the freight bill by roughly ten times.

Does CSRD apply to my import business? #

Almost certainly not directly. After the Omnibus I Directive of March 2026, the obligation falls only on companies with more than 1,000 employees and more than €450 million of net turnover, from financial years starting in January 2027. You will still meet CSRD through customers who do have to report. But such a customer may not require data from a supplier of 1,000 employees or fewer beyond a limited set aligned to a voluntary standard, and you may decline anything past it.

How do I work out the emissions of one shipment? #

Multiply the weight of the cargo by the distance travelled on each leg, then by the emission factor for that mode, and add the legs together. Use ISO 14083 or the GLEC Framework as the method, because that is what your buyer’s own reporting is built on. Two mistakes are common. The first is using the port-to-port distance rather than the door-to-door one. The second is applying a dry-container emission factor to a reefer, which understates the shipment by 20% to 40%.