The claim is a good slogan and a bad unit of comparison. "One box equals ten mu" — one container matching about 0.67 hectares of farmland — does not survive arithmetic on output, but it survives, roughly, on land. Ten mu of lettuce requires something like twenty containers, which together occupy a single parking lot. The land saving is real and on the order of twentyfold. The output saving is a category error. Separating the two is the only way to think clearly about whether a container farm pays.
First, the capacity arithmetic
One mu is 666.7 square metres, so ten mu is some 6,667 square metres. Field lettuce in a temperate climate yields roughly 4 kilograms per square metre per crop, over two crops a year — about 8 kilograms per square metre annually, or some 50 tonnes across ten mu.
Now the box. At the 2024 Bright Food–Duoduo Vertical Farming Challenge on Chongming Island in Shanghai, the Shanghai Academy of Agricultural Sciences team harvested 1,680 heads of lettuce weighing 299.55 kilograms from about 42 cubic metres in roughly fifty days, with over ninety percent reaching Grade A commercial standard. Seven such cycles a year gives about 2.1 tonnes.
So one container produces roughly 4 percent of what ten mu produces. You need closer to twenty-five containers to match the field. The honest restatement of the slogan is this: ten mu of lettuce output fits into about 300 square metres of footprint — a twentyfold saving in land, not a tenfold gain in output.
That reframing matters commercially, because land is the input a city cannot manufacture and electricity is the input a container cannot avoid. The claim is a land-use claim, and only a land-use claim.
What the capital actually buys
The purchase price of a container farm is mostly not the container. A used forty-foot high-cube is a commodity item; the cost sits in what goes inside it: multi-tier racks, LED arrays with tuned spectra, HVAC and dehumidification sized for a sealed box, recirculating nutrient delivery, sensors, control software, and whatever automation the operator chooses. Add site works — a level pad, a power drop, drainage, sometimes a structural assessment if the unit goes on a roof.
Because the specification varies enormously, published price ranges are close to meaningless as planning inputs. What is stable is the structure of the cost: lighting and climate equipment dominate capital, and both scale with the number of tiers, because every added layer needs its own photons and its own share of cooling. Unlike a field, where the second hectare costs less per unit than the first, a container farm's costs scale almost linearly with capacity. There is very little economy of scale inside a single box; scale comes from replicating boxes, which is precisely why standardisation is the sector's central engineering problem.
The line item that decides everything
Energy is the variable that determines viability, and here the competition data is unusually useful. The champion team, from China Agricultural University, reached 0.33 kilograms of lettuce per square metre per day while holding consumption to 12.25 kilowatt-hours per kilogram of lettuce.
Run that number. At an industrial tariff of roughly USD 0.08–0.10 per kilowatt-hour, electricity alone costs about USD 1.00–1.25 per kilogram of lettuce — before a single seed, tray, nutrient dose, hour of labour, or dollar of depreciation. Field-grown lettuce wholesales in most markets for well under a dollar a kilogram.
That single comparison explains the shape of the industry better than any business plan. A container farm cannot compete on commodity lettuce, and no amount of operational discipline changes the physics: a sealed box has no sun, so every photon is purchased. The 12.25 figure was celebrated as industry-leading precisely because it is close to the floor of what the technology can do.
An illustrative model
To see where the money goes, take a single unit producing about 2.1 tonnes of lettuce a year and assume the following — these are planning assumptions, not quoted prices, and they should be replaced with real ones before anyone signs anything:
Item
Annual (USD)
Revenue — 2.1 t at USD 8/kg premium wholesale
16,800
Electricity — 2.1 t at 12.25 kWh/kg, USD 0.09/kWh
(2,300)
Labour — 0.5 FTE at modest urban wage
(15,000)
Nutrients, seeds, packaging
(1,500)
Maintenance and consumables
(2,000)
Depreciation — capital over 10 years
(10,000)
Result
(14,000)
The model loses money, and it loses it for a specific reason: labour, not electricity, is the largest operating line item once a unit is small. A container farm produces a few tonnes a year, which is not enough volume to absorb a full-time wage, yet the daily work — seeding, transplanting, harvesting, cleaning, monitoring — does not shrink with the box. This is the trap the sector keeps stepping into, and it is why a wave of well-funded vertical farming companies in North America and Europe collapsed or contracted sharply between 2023 and 2025. The failures were rarely agronomic. They were unit-economics failures at small scale and over-ambitious scale-ups alike.
Where the numbers do work
Three configurations change the arithmetic honestly.
High-value crops. Microgreens, herbs, edible flowers and specialty salad leaves sell at multiples of lettuce prices — often USD 30–60 per kilogram wholesale — while growing on shorter cycles. Because electricity cost per kilogram is roughly constant across leafy crops, raising revenue per kilogram is the only lever that works without new physics.
Owner-operators. When the labour line is the proprietor's own time rather than a wage, the same box can look profitable while generating a modest income rather than a return on capital. Many functioning container farms are exactly this: small businesses, not investments.
Structural advantage at the site. A rooftop in a dense, expensive city; a northern winter where field supply stops; a restaurant or retailer willing to pay for same-day harvest and a specific cultivar; a market where imported greens carry days of transit loss. In each case the container is being paid for something other than kilograms — freshness, continuity, food safety, or avoided logistics — and those are legitimate revenues.
He Dongxian, a professor at China Agricultural University and a judge at the Chongming competition, assessed several teams as reaching internationally leading levels on yield and energy consumption with the capability for industrialisation, and argued that with stable technology the model can reach profitability at scale. That is a defensible claim about replicated, automated, standardised production — not about a single box dropped in a car park.
The return that is not on the spreadsheet
Strip the slogan away and the container farm's case is narrower and more durable than "one box equals ten mu." It converts expensive horizontal land into cheap vertical space, removes weather and season from the production schedule, and puts harvest within hours of the plate. Those are real advantages, and cities with scarce land and high-value local demand will pay for them.
What it does not do is beat a field on cost per kilogram of a commodity crop. Any business plan claiming otherwise is comparing a box to ten mu on the wrong axis. The right question is not how much a container can grow, but whether the specific site, crop and customer can carry a cost structure in which every photon is invoiced.
