Vertical Farm Basics and Overview

A vertical farm is a facility that buys back the sun with electricity

When you’re weighing whether to step in or invest, the thing you’re usually left holding first is a number. “What share runs in the red,” “investment in the tens of billions” - you take one figure you saw in the news and make it the yardstick for your own project. But the average of what, exactly? A vertical farm is a facility that buys back the outside sun and climate with electricity to raise living things. Depending on how you combine “how far you replace it with electricity” and “what you grow,” the profitability and the risk become different animals while the name stays the same.

The misunderstanding the word “factory” invites

Shelves lined with vegetables, the glow of LEDs, people in white work coats. Lately you see the word “vertical farm” everywhere. When people hear “factory,” most picture that production line. Press a button and standardized products keep coming out, cheaper and cheaper. Stop it, and you can start it up again. Put in the equipment and you can make anything. That’s the intuition a factory of machines has trained into us.

But look at a vertical farm with that intuition intact and you’ll probably read it wrong. Even though they go by the single phrase “vertical farm,” what’s inside differs as much as one living creature does from another. A room fully closed off with LEDs, and one that controls the inside while using sunlight the way a plastic greenhouse does. Two things this different, sitting under one name - enough to make you wonder whether they deserve it.

What I want to do here is set the word “factory” aside for a moment and hand you one way of seeing what is actually going on inside. Let me give the conclusion first. A vertical farm is not “a factory where standardized products come out cheap” but a facility that buys back the outside sun and climate with electricity to raise living things, and it becomes a completely different animal depending on the type (where you take the light from) and the crop. Make this view your entry point, and three things click into place: why the crops you can grow are skewed, why the economics are so tight, and why you can’t treat it all as one lump.

Two types, split by light source

I said they’re different animals, but the dividing line that matters most is a single one. Do you use the sun, or do you make the light, all of it, with electricity?

The type that uses sunlight - the so-called greenhouse type - think of it as an extension of a greenhouse. Light comes in through the roof for free. So what’s the factory-like part? It measures and controls temperature and humidity, CO2, watering, that range of things. It’s a facility that manages everything but the light.

The other one, the fully closed PFAL, the all-LED type, throws the sun away entirely, in a windowless room. Then it buys back the light that was originally free with electricity, and remakes it itself. So it’s less “going out of your way to use expensive electricity” and more “the electricity bill for reproducing the whole outside weather inside a room.” Not just the light - HVAC, humidity, all of it electric. That’s the price of cutting yourself off from the outside air.

This “buying back with electricity” view bites hardest on the fully closed PFAL. Because the greenhouse type gets light for free, think of it as buying back everything but the light - the control of temperature, humidity, and CO2.

And this “electricity bill for reproducing the outside weather inside a room” shows up clearly when you look at orders of magnitude. An analysis pulling together the energy consumption of several facilities puts PFAL energy consumption at a median of around 78 MJ per kg of harvest, 27 MJ in a greenhouse, and the open field far below both. (See 1, 2) Even looking at 1 kg of lettuce, there are estimates of around 160 MJ of primary energy for indoor hydroponics versus just under 11 MJ in the open field. (See 1, 2) “Replacing light from the free sun with electricity” isn’t a story that stops at “feels like it’d be expensive.” It’s a story of becoming one or two orders of magnitude heavier.

That said, how much this weight bites changes with the type and the scale. In estimates for large facilities, electricity takes a big share; but when you’re small-scale, selling at a high price, with a buyer nearby, labor and packaging can come to dominate instead, and electricity ends up being just one part of the cost. What’s common is the property that “the cost of artificially providing a substitute for outside nature always falls somewhere,” and it doesn’t always show up in the form of an electricity bill.

Why the crops you can grow skew toward leafy greens

Lettuce, spinach, baby leaf. What you tend to hear about in vertical farm talk is leafy greens. With that electricity bill in mind, the reason comes into view. Leafy greens don’t need strong light, they’re short and you can stack many tiers on a shelf, and the growing period is short too. So they just barely justify the cost of making light with electricity. This direction gets pointed out the same way in several overviews. (See 4)

A closed-type artificial-light facility has high running costs for lighting and HVAC, and to make it pay you need high-value-added crops - that’s the structure. (See 4) What can be grown commercially is, for now, mostly limited to short leafy greens. (See 5)

What about crops that bear fruit, then? Both the amount of light needed and the electricity bill come out an order of magnitude different. Tomatoes need an enormous amount of light to set fruit. (See 5) And in crops that grow tall, as the stem stretches the light reaching the lower leaves keeps weakening, and light-use efficiency falls off. That’s why it’s pointed out that what PFAL can grow commercially is limited to small plants. (See 5) There’s also the point that fruiting vegetables are, by the very shape of their bodies, poorly suited to vertical stacked cultivation in the first place. (See 3) But this is a story about today’s varieties and prices. If breeding for smaller size, low-light tolerance, and shorter growth cycles advances, the range of what you can grow could shift. It isn’t a fate fixed by today’s economics.

With staple crops it gets more extreme. Growing soybeans under fully artificial light, for instance, is projected to need a thousand to two thousand times the energy of the open field. (See 1) Rice and wheat too: with today’s varieties and electricity prices, all-LED can’t be expected to pay on the electricity bill alone, several analyses point out. (See 6, 7)

This is where the word factory betrays you. An ordinary factory: the more equipment you put in, the more things you can make, broadening both volume and variety. A vertical farm’s fully artificial light is the reverse. Because of the constraint of buying light back with electricity, the set of things you can make actually gets squeezed tight. And it’s not like a line of machines you can stop and start up again, either. Going by the feel of the PFAL and leafy-greens floors I’ve seen, what’s growing inside is alive, so you can’t pause it like a line because demand dropped and then start it again from the same spot. Once the seed is sown, sold or unsold, it grows at the speed of its own growth. It’s not a device for mass-producing anything cheaply, but a device for putting out a limited set of crops steadily, cut off from the weather. So even though it’s the same “factory,” it’s not a factory that scales up - it’s a factory that irons out the swings.

Checking first whether the crop you want to grow pencils out makes the talk of types and costs that follows hit much closer to home.

The other dividing line is who decides the environment

Put a thermometer in a plastic greenhouse and open the windows by hand when it gets hot. From there, all the way to sensors measuring and skylights opening automatically, CO2 being added, the nutrient solution concentration adjusting itself. In fact, there’s a dividing line here that’s different from the first one.

For the greenhouse type, it’s less that the boundary with ordinary farming is blurry and more that it’s all one unbroken stretch. From opening and closing windows by hand to fully automatic, it’s one continuous slope, and there’s no line drawn somewhere saying “factory starts here.”

If you had to draw a line, you’d look at “whether the environment is measured in numbers and machines are deciding and acting toward those numbers.” As long as a person looks at the sky and opens the windows on a hunch, it’s a greenhouse; if the environment keeps being controlled toward set values with no one present, it leans toward factory - it’s a gradient.

So this is a different axis from the line we split on first, “use the sun or throw it away.” The first line was about where you take the light from; this line is about who decides the environment. When you look at a vertical farm, it’s worth keeping in mind that these two axes cut separately.

Don’t apply the industry’s single number to your own project

In vertical farm news, haven’t you seen talk like “about half of operators run in the red”? Or, the other way around, felt the urge to copy a success story’s initial investment figure straight into your own estimate? Is a single figure like that something you can use directly for your own situation? Once you hold the view that “it’s a different animal by type and by crop,” how you handle this changes.

To put the conclusion first: it’s a number you must not carry over to your own case. In fact, it’s the most dangerous way to use it. “Half in the red” and “the success story’s initial investment was X” are, by everything said so far, nothing more than an average taken by lining up things that differ in both type and crop. It’s a loss rate that counts fully-artificial-light lettuce, greenhouse-type tomatoes, and small near-hobby facilities all together as one “vertical farm.” There’s no guarantee anywhere that that average applies to the combination of type and crop you’re about to attempt.

In fact, numbers produced by mixing types and crops swing quite a bit. In surveys of large-scale protected cultivation and vertical farms in Japan, one report puts operators in the red at 49%, while another reports that even after subsidies on the scale of 50 billion yen were poured in, 75% were still in the red as of 2017. (See 8, 9) I cite these numbers not to apply them to your own project, but to show how much the population swings by itself.

Yet when you divide the population by type, a clear difference - separate from the swing - emerges. In the fiscal-2025 survey of large-scale protected cultivation and vertical farms, the share that was profitable or breaking even in the latest financial results was over 70% for both greenhouse and combined types, while the fully closed PFAL stayed at about 50%. The earlier “49%” and “75%” are figures that blend type and crop, so they swing widely from report to report; but re-divided by type, the difference - “PFAL is roughly fifty-fifty, greenhouse and combined types are around 70%” - holds up even as the year and the population change. Mix it and it swings; divide it and the difference looks robust. “A different animal by type” is not a matter of feel - it’s a difference that shows up even in domestic primary surveys.

Copying a success story’s initial investment figure is more dangerous still. That’s a single number that only held together with that person’s crop, that location’s electricity rate, that scale, that buyer, all as a set. Copy just the price tag of the equipment and there are things that don’t come with it. Remember the swing that gets ironed out - the value of putting product out steadily, cut off from the weather. At what price was that person’s buyer paying for it? Leave that out, and you can build the same equipment and it still won’t pencil out.

The weight of energy is the same. It varies by more than an order of magnitude across facility type, crop, and region, and no clear industry-wide downward trend has emerged yet, the analysis says. (See 1)

One thing I want to flag here is that “the weight of energy” mixes two separate things. One is environmental load (CO2 emissions and so on), and because this moves by more than an order of magnitude depending on the power source and how waste heat is used, you can’t lump it together as “vertical farms are good for the environment” or “bad.” (See 1, 10) The other is the sheer amount of energy consumed per kg of harvest. This amount of consumed energy doesn’t go down even if you switch the power source to renewables. You can lower the environmental load, but that doesn’t make the weight of the electricity bill itself disappear.

So when you see a single number, there are four things to check first. Is this a greenhouse or a fully closed PFAL? What’s the crop? Where’s that electricity cost from - what region’s rates? Who was it being sold to, and for how much? A number that doesn’t have these lined up can serve as a reference, but you can’t use it as the root of your own estimate. Put the other way, other people’s numbers aren’t useless either. For getting a feel for orders of magnitude, or for framing the next question you should check, they’re plenty useful. The average loss rate and the successful operator’s investment figure, seen not as answers but as material for framing questions, don’t go to waste.

Move on to the breakdown of what the “half” in that loss rate is concretely made of, and you’ll start to see where your own type and crop sits.

What to check first so you don’t miss

When you start looking into “what’s a vertical farm like, anyway,” trying to decide from the type first usually gets you lost. My recommendation is the reverse. Set the type aside for now and start from “what swing do I want to iron out.”

First, decide from the exit. To whom, what, and why at that price will it get bought? Is there a buyer within your reach who pays for the very fact that you can deliver without ever running out, all year round? Without this, whatever you build won’t pay.

Second, place the crop that buyer wants. Once the exit is set, the crop is almost linked and decided. And once the crop is decided, the strength of light needed, the height, the growing period all follow.

Third, the type finally appears. Or rather, by this point the type isn’t something you choose yourself - it gets narrowed down from the crop’s side. If it’s leafy greens that don’t want much light, fully artificial light comes into range too; if it’s a fruiting crop that needs strong light, the greenhouse type is the only option - that sort of thing. The type decided here contains both of the two axes from earlier: where you take the light from, and how far you leave the environment to machines. The type is not the starting point.

Fourth, only now do you go read other people’s numbers. Choose and read only the cases where that four-piece set from earlier - type, crop, the location of the electricity bill, buyer - lines up with yours. Reverse the order and start from an image like an “LED lettuce shelf,” and you end up holding just the picture of the equipment with no exit. Exit, crop, type, then numbers. Just keep this order intact and you won’t miss by much. Note that this is about priorities - don’t start from the picture of the equipment; in reality, constraints like the funds, land, and location you have on hand bite at the same time. It does not mean you can put off thinking about funding.

By now you should be able to see that under the single phrase “vertical farm,” facilities that are completely different animals stand side by side depending on where you take the light from, who decides the environment, and what you grow. Not the device that comes to mind when you hear “factory” - the one that mass-produces anything cheaply - but a facility that buys back the outside weather with electricity and equipment to iron out the swings in a limited set of crops. Make that your entry point, and three things click into place: the reason the crops skew, the reason the economics are tight, and the reason a single number can’t be relied on.

From here, you can move on to what’s next depending on whether your interest is which type pencils out, which crop pays off, the substance of why so many are said to run in the red, or how revenue is decided on the floor.

One last thing. The hands-on feel covered here centers on the range of growing leafy greens in a PFAL. For the greenhouse type, and for fruiting vegetables and large overseas cases, the substance isn’t uniform enough to talk about as one lump here. So this article isn’t the place that hands you an answer - it’s an entry point for separating out “which type, which crop, which question (is it economics, crops, or the floor) am I actually curious about” and moving on from there.