To simplify the question as much as possible, let's assume an Earth-like planet, part of a galaxy-spanning civilization. 30% is above water level, the rest is ocean of salt-water.

This world is an agri-world, purely used to produce as much food as it can and ship the production to other systems that can't sustain themselves when it come to crop, livestock and such. It keeps just enough food to sustain the population on site. All the landmass is used for farming. For the sake of simplification, it is only producing crops and not meat of any kind.

On Earth, farming use large quantities of water, but since it's all on the same planet, it's (I assume) a neatly closed system and we are only concerned about the ratio of salt to freshwater.

My question is, if you are shipping farm products of world, aren't you removing water from the equation? Would you witness a significant drop in the quantity of water of this world over time - or is the water contained in crops negligible in the grand scheme of things?

  • 5
    $\begingroup$ They would be very much more sustainable when it comes to water than when it comes to phosphorus... A world like Earth has water aplenty, phosphorus not so much. $\endgroup$
    – AlexP
    Aug 15 at 18:17
  • 4
    $\begingroup$ If you wanna make this believable, you’re gonna need very good reasons why a galaxy-spanning civilization would transport something as simple as food between systems instead of just building greenhouses and just importing the required molecules once and recycling them. $\endgroup$ Aug 16 at 12:27
  • 2
    $\begingroup$ @AlexP I read one article (or maybe it was a YouTube broadcast), that had a point that once 'post-scarcity' is achieved, it actually only moves on to other scarcities, and phosphorus was pointed out as being an extremely hard limit on how much life a system can support. $\endgroup$ Aug 16 at 15:56
  • 3
    $\begingroup$ If you are concerned about running out of water, it may also be interesting to be concerned about destinations gaining too much water $\endgroup$
    – user78090
    Aug 16 at 16:34
  • 3
    $\begingroup$ I suggest you read "The Moon is a Harsh Mistress" by Robert Heinlein -- He deals with exactly this scenario. (albeit Hydroponic farming on the moon) $\endgroup$ Aug 16 at 20:42

10 Answers 10


Technically it's Not Sustainable

But you're probably ok.

Current worldwide food production is 4,000,000,000 tonnes of food per year. Let's assume that food is 100% water, just to make things easy.

All water is eventually sourced from the ocean, and it evaporates and is transported to rivers, lakes, etc by the water cycle. We then use this freshwater to irrigate crops.

Earth's oceans have a mass of approximately 1,400,000,000,000,000,000 tonnes.

So we could theoretically export all the food that Earth produces for 350,000,000 years before we drained the oceans dry.

350 million years is probably long enough that you could nudge some long period comets into the appropriate orbits to replenish your water stocks.

You can easily adjust this value by scaling the current population of the Earth. There are currently about 8 billion people on Earth. So if you need your agri-world to feed 20 billion people, then that's about 3 times the current world population, and you would drain your oceans in about a third of the time calculated above - 116 million years.

  • 2
    $\begingroup$ I'm curious about the downvote? $\endgroup$
    – codeMonkey
    Aug 16 at 15:33
  • 4
    $\begingroup$ Wouldn't there be significant negative effects before the oceans were completely drained, though? $\endgroup$ Aug 16 at 20:23
  • 7
    $\begingroup$ @Feathercrown - Ecological collapse will probably occur significantly before the 350 million year mark as ocean salinity relentlessly increases. But ecological collapse would probably also occur when you clear cut the surface of the planet to grow wheat. I'm assuming that the kind of culture that creates "agri-worlds" to feed their hungry hyper-urban-worlds is also the kind of culture that doesn't care about obliterating local ecosystems. (provided that the timeline to obliteration is long enough to turn a profit) $\endgroup$
    – codeMonkey
    Aug 16 at 20:57
  • 8
    $\begingroup$ Eh, presumably said hyper-urban worlds that consume megatonnes of food will also produce megatonnes' worth of compostable waste in the process, so as long as that's shipped back to the agri-worlds to serve as more fertilizer... $\endgroup$
    – Shadur
    Aug 16 at 21:10
  • 5
    $\begingroup$ Can you do this math for the co2? I suspect you will run out of carbon long before you run out of water. $\endgroup$
    – Ariel
    Aug 16 at 23:12

Cargo ships hate to run empty.

Your farm world exports crops. Cargo ships leave full of crops. Then they come back. What do they bring on the return trip? Maybe the worlds buying crops have ample water but lack other things that crops need. They can send their water back in the cargo ships, replacing the water for the farm world and partly offsetting the cost of the crops they are importing.

It would be a fine visual - the cargo tug towing a giant gently sublimating iceberg. It releases it to free fall into the atmosphere of the farm planet, then hits it with a missile on the way down to break it up. Cool alien rain!

Of course somewhere in the story your protagonists will find some other use for an iceberg in space...

  • 7
    $\begingroup$ If the entire planet is solely for farming, the ships are likely to return with either waste or fertilizer (and a few luxury products for those living on the planet). This is needed to offset the soil exhaustion, as that will be your biggest limiter on the production of crops. Moreover, a civilization that is capable of spaceflight, a multi-planet empire, and setting up farm-worlds will know about the exhaustion problems and are likely to have rules/solutions in place for the import/export ratios of resources. $\endgroup$
    – vinzzz001
    Aug 16 at 9:15
  • 1
    $\begingroup$ One of Sherri Tepper's novels talks about the dispersal of trace mineral soil supplements by exploding rockets. I could see such being done by incoming cargo ships - they could simply disperse the supplements when using aerobraking to land as part payment for the crops. Simple and efficient assuming they are cheap enough to spread widely. $\endgroup$ Aug 16 at 11:32
  • 2
    $\begingroup$ FTL so cheap/easy you ship water back on one leg of the trip... yikes. $\endgroup$
    – John O
    Aug 16 at 13:49
  • $\begingroup$ It also keeps the filthy heretics in check. They'll never know if the missles fired from the Holy Armada are just to vaporize the iceberg or to carry an Exterminatus in the Emperor's name. $\endgroup$
    – Mindwin
    Aug 16 at 15:10
  • 1
    $\begingroup$ In this vein, if they're using a space elevator to lift the crops out of the planet's gravity well, they need a counterweight to send back down to the planet. You could have some water-bearing comets that they put in orbit around the planet, and they rip ice off of those and send them down to counter-weight the crops being lifted by the elevator, thus replenishing the water lost $\endgroup$
    – TheSneak
    Aug 18 at 19:58

You will be shipping back manure.

You are shipping mass off-world.

At these scales, you don't care that some of the mass is water, you care what proportion is hydrogen, oxygen, carbon, nitrogen... etc. An amount of energy went into chemical processes to fix that into digestible forms.

The end result of digestion is mostly the same amount of mass, with a (lower energy) chemical makeup. Wherever you have sent this food to doesn't particularly want piles of manure around, and you have all these freighters that would otherwise be empty for the return trip.

  • $\begingroup$ Perhaps not "manure" but biosolid residuals from sewage treatment. It could be returned with a moisture content comparable to the food that's exported. $\endgroup$
    – Theodore
    Aug 16 at 15:13
  • 4
    $\begingroup$ @Theodore To a lay person "biosolid residuals from sewage treatment" could be summarised as "manure" $\endgroup$
    – Caleth
    Aug 16 at 15:14
  • $\begingroup$ This is a good point to raise. It might take millions of years for the water drain to matter, but the drain in minerals will probably happen in a generation. (Although I think what we're really looking for is "fertilizer".) $\endgroup$
    – JamieB
    Aug 16 at 16:20
  • $\begingroup$ @Caleth This stack isn't about reductions to the lowest common layperson denominator. The activated sludge process is over 100 years old. At least read the wikipedia article. Biosolids are mostly made up of dead microbes that have already digested the waste. $\endgroup$
    – Theodore
    Aug 16 at 16:54

Technically yes. Every unit mass shipped offworld will remove 1 unit mass from the planet. If a percentage of that mass is water then you'll remove that percentage of total mass as water.

Earth has 1.3 billion cubic kilometers of water. Of that 2.5% is fresh water. That's 32.5 million cubic kilometers of fresh water. Given that most commercial grain has a water percentage (by weight) of 10%-15%, you could remove 32.5*10^18 kilograms of grain before having a 1% impact on the amount of fresh water on the planet.

  • 3
    $\begingroup$ About that fresh water... The water in the rivers (and thus used for agriculture) comes mostly from rain, and most rain water comes from the sea. Pure natural solar-powered desalination, as it were. $\endgroup$
    – AlexP
    Aug 15 at 18:16
  • $\begingroup$ @AlexP If you wanted to do a more detailed calculation sure. But you'd need to factor in that 68% of fresh water on Earth is in glaciers and ice caps, and 30% of fresh water is in underground aquifers. A very very small fraction of total water is replenishble fresh water. $\endgroup$
    – sphennings
    Aug 15 at 18:24
  • $\begingroup$ Yes, true, but most of the water used in agriculture comes from rain or from rivers. There are of course some very notable exceptions where misguided countries practice agriculture using fossil water. $\endgroup$
    – AlexP
    Aug 15 at 18:43

Whether or not it's sustainable isn't the question

Sustainability is a great context for a story, so whether or not such a planet is factually sustainable or not is really irrelevant, right? Besides, you're correct, unless something is done, the world is losing water.

But, what can be done?

Freeze Drying

Nothing leaves the planet without being freeze-dried. This minimizes the water loss. Frankly, it likely minimizes it to a degree that it's impractical to consider the implications for anything but a very long term.

But what might make the most sense is...

Mine the ice giants!

Granted, the term "ice giant" when referring to planets like Neptune and Uranus is a bit of a misnomer. Oh, they likely have boat loads of ice... but you'd need to work your way through miles of other nasty stuff (like methane) to get to it.

But that doesn't mean they're not solutions! Ice giant planets, ice asteroids, ice-covered moons. NASA claims the solar system and beyond is awash in water.

Which is a long way of saying you really don't have a problem. If your economy allows you to ship wheat off the planet, it will allow you to ship water back to it at no populated planet's expense.

  • 2
    $\begingroup$ Drying was also my first guess. Interstellar transport is very very expensive, so any bit of water in the food is weight (volume does not matter, assuming space elevators) that is NOT the desired stuff (food) $\endgroup$
    – Hobbamok
    Aug 16 at 10:07

As codeMonkey calculated, water is not what will be used up first.

On the other hand, there are other limiting factors in agriculture:

  1. Nitrogen. Most of the nitrogen is in the atmosphere and it is way less dense than the liquid water. One will carry away comparable amounts of water and nitrogen by exporting food.

  2. Phosphorus. Most of it is in the litosphere, but is at least as much limited.

  3. Carbon. Where the real problems is. Just like with the nitrogen, one will carry away comparable amounts of carbon. On the other hand, it is ~0.04% of our atmosphere today. Make it 0.02% and create an eternal ice age. A terraformed planet will likely not have much fossil carbon deposits (coal, oil, gas). Carbide ores may be an option, but only to delay the inevitable problem.

One will need a big industry mining these important elements on a celestial bodies beyond the frost line (whatever equivalents you have for Jupiter and Saturn satellites and the icy planets beyond). Or somewhere else.


I can't see how an "agri planet" can stack up economically.

Even if you have an orbital tether, you still burn up a significant fraction of the food's calorific value just to reach synchronous orbit, and a boat-load more to reach escape velocity.

On a planetary scale it would be more efficient to develop zero-g agriculture and get the water from comets and asteroids. Which I suppose answers the question "no, you don't run out of water", at least on the planet surface.

The other problem is at the receiving end. Let's assume that the main market is planets who have exceeded their carrying capacity, even after entirely displacing their native ecosystems, so we can assume a global populations of 100bn+, and that they each eat around a tonne of solid food per annum.

De-orbiting 100 bn tonnes of food per annum means absorbing almost half a petawatt on an ongoing basis, equivalent to 1.5% increase in planetary insolation. That's going to make our current climate change look like a picnic.

  • $\begingroup$ Agricultural subsidies. That's how. That planet is one giant EU budget line item. $\endgroup$ Aug 17 at 11:29

This answer addresses an issue which is peripheral to the main question BUT utterly crucial.

For interplanetary (let alone interstellar) food transportto be viable you are going to need space transportation technologies not only far beyond what we have now but also far beyond what we imagine will ever be possible.

I could dig up numerical figures for the following but it makes the point adequately without them. I may add some if there is enough interest shown.

At present we use chemical based propulsion systems. A Saturn V delivered about 4% of its launch mass to orbit. An utterly inconceivably good chemical rocket systm is unlikely to be deliver better than 10% launch mass to LEO.

To deliver payload to the surface of Mars from LEO requires as much energy per kg from LEO as it takes to get TO LEO. So using the Inconceivable V launcher technology ("IV")we can deliver 10% x 10% = 1% of launch mass to the surface of Mars.
Using Saturn V technology it's 4% x 4% = 0.16% of launch mass or 1/625th of launch mass.

Delivery to near earth asteroids is about 0.5% using the IV and Lunar surface is somewhere in between asteroids and moon.

A Space Elevator reduces the earth's surface to LEO energy cost to about zero. (Energy per kg to orbit on an elevator is a few MJ at 100% efficiency . Allowing for substantial inefficiencies that's say roughly $1 in electricity costs.

But, even if we get off the planet's surface for free, we still have to get to the destination and to re-enter the payload. Antimatter systems will bring the cost down and the payload fraction up - once we manage to get antimatter down to commodity pricing. As antimatter is currently priced at about 2500 trillion dollars per gram (2.5 x 10^12) we, or your posited civilisation, have some way to go from our present situation. That 1 gram costs about 100,000 times as much per kWh as current electricity prices, which is actually quite encouraging. But unlikely to be commodified any time soon.

SO - unless we have a new Non-Existium V drive technology it is likely that almost any practical alternative is going to be cheaper. Orbital growing systems, underground systems, float in the air, ... .

The only way to make this sensibly viable is if you can produce a magic matter transport system that requires minimal energy and delivers payload between planets, or stars, rapidly or instantaneously. In a story this is possible. In our reality, start planning local farm systems.


No and yes

As you already understood shipping out food means shipping out a lot of water. A country like Earth can have a lot of water, but most of the water will eventually become salted or polluted. The fresh and clean water on Earth come from a continuous cycle of evaporation, rain and filtering by sand or clay layers when it seeps underground. So fresh water will never be abundant like sea water and shipping it out will eventually deplete it.


Shipping water through space is expensive and pointless. Water is heavy, but it can be available everywhere, why ship it? The energy cost of freeze drying the food for interplanetary commerce would make sense. So all the food will be exported freeze dried.

Still some finite resources like phosphates would risk depletion, but they can be imported back.


How are you managing fuel? When you burn fuel you produce water and carbon dioxide - both things that will help your agriculture.

Right now you need 8 times as much fuel (by mass), as payload. So if you burn your fuel in the atmosphere of the planet, you'll actually add 8 times as much water+co2 as you export.

If fuel comes from outside (since you said the plant is agriculture only) you'll actually have the reverse problem if you do this long enough.

Also, going back to the original question, you'll actually have a much bigger problem with co2 - you'll run out of that way before you run out of water.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .