This idea will not work, but it is delightful to imagine and I had fun thinking about it.
Farming is only really profitable in the USA when done on good farmland at large scale, on the order of several square miles. A square mile is 640 acres. The average price of an acre of good farmland is about $4,200, so buying a 5 square mile farm might cost about $13 million.
Irrigating an acre requires 5,000-8,000 gallons of water per day, so one desalination plant could handle at most 10 acres. To irrigate a farm of 5 square miles would require at least 320 of these desalination plants. Assuming each costs $13,000 to build, that’s a cost of about $4 million for the desalination plants alone. The land would cost another $2 million or so. My guess is that the total costs of terraforming 5 square miles of land add up to around $10-$20 million and probably require extensive permitting, since you’re extracting water from the common aquifer. I’m uncertain if your listed price for the desalination plant includes the cost of labor to build/install it. Either way, though, my guess is that your costs in terraforming this farmland will exceed your revenues, and that the buyer will perceive the land as inferior to better farmland with more established farming infrastructure that they could buy elsewhere.
Selling the minerals extracted from desalinated water might make some money, but in addition to evaporating off the water, you’d also need to separate the minerals and transport them to a buyer. The more minerals per gallon, the less viable the irrigation, and vice versa.
Yeah, and I think aquifer depletion would be a pretty big problem this would run into. Irrigating farmland requires ~twice the annual rainfall you get in texas. So terraforming projects might have to import water from the ocean.
Based on that Fermi estimate, the plan actually sounds surprisingly viable. It’s already on the same order of magnitude as traditional farmland even without optimizing very hard, which means just a bit of added cleverness should put it in the realm of tractability.
That’s a fair point. Let’s look at some other sources and include fixed costs, and do some sanity checks on water requirements.
An inch of rainfall is about 27,000 gallons per acre, so to supply a 5 square mile farm with an inch of rainfall equivalent is about 87 million gallons of water. High yield corn needs 22-30 inches of rainfall per year. Peak season for corn is about May-September, so let’s assume we therefore need about 22 inches of rainfall equivalent (1.9 billion gallons) spread out over 5 months or about 150 days. That averages to about 13 million gallons of water per day. That tracks very well with the original estimate of 5,000 gallons/acre/day, which would come out to 16 million gallons per day for this farm.
This company says a 27 MGD plant cost $87 million. If cost scales ~linearly with capacity, that suggests a $50 million price tag for the desalination capacity—now perhaps 4 times the normal cost of a farm of equivalent size. That puts the cost of desal much higher than the lower bound I used from the OP.
They also claim a unit cost (including capital cost, debt service, and operating cost) of at least $1.25 per 1,000 gallons of desalinated brackish water, which would therefore cost about $20,000 per day for this farm and create an added $3 million cost per year.
I don’t have a great source here, but it looks like groundwater use for farm irrigation is typically free. Corn and soy yield optimistic profits of $100-$200/acre, or perhaps $320,000 for a 5 square mile farm. A $3 million cost per year is again an order of magnitude higher than those expected profits.
This idea will not work, but it is delightful to imagine and I had fun thinking about it.
Farming is only really profitable in the USA when done on good farmland at large scale, on the order of several square miles. A square mile is 640 acres. The average price of an acre of good farmland is about $4,200, so buying a 5 square mile farm might cost about $13 million.
Irrigating an acre requires 5,000-8,000 gallons of water per day, so one desalination plant could handle at most 10 acres. To irrigate a farm of 5 square miles would require at least 320 of these desalination plants. Assuming each costs $13,000 to build, that’s a cost of about $4 million for the desalination plants alone. The land would cost another $2 million or so. My guess is that the total costs of terraforming 5 square miles of land add up to around $10-$20 million and probably require extensive permitting, since you’re extracting water from the common aquifer. I’m uncertain if your listed price for the desalination plant includes the cost of labor to build/install it. Either way, though, my guess is that your costs in terraforming this farmland will exceed your revenues, and that the buyer will perceive the land as inferior to better farmland with more established farming infrastructure that they could buy elsewhere.
Selling the minerals extracted from desalinated water might make some money, but in addition to evaporating off the water, you’d also need to separate the minerals and transport them to a buyer. The more minerals per gallon, the less viable the irrigation, and vice versa.
Yeah, and I think aquifer depletion would be a pretty big problem this would run into. Irrigating farmland requires ~twice the annual rainfall you get in texas. So terraforming projects might have to import water from the ocean.
I’m also concerned that there might be little-no topsoil.
I considered that. I wonder if you could plant some hardy, dry-climate crop the first few years and till it under to improve the soil composition.
Based on that Fermi estimate, the plan actually sounds surprisingly viable. It’s already on the same order of magnitude as traditional farmland even without optimizing very hard, which means just a bit of added cleverness should put it in the realm of tractability.
That’s a fair point. Let’s look at some other sources and include fixed costs, and do some sanity checks on water requirements.
An inch of rainfall is about 27,000 gallons per acre, so to supply a 5 square mile farm with an inch of rainfall equivalent is about 87 million gallons of water. High yield corn needs 22-30 inches of rainfall per year. Peak season for corn is about May-September, so let’s assume we therefore need about 22 inches of rainfall equivalent (1.9 billion gallons) spread out over 5 months or about 150 days. That averages to about 13 million gallons of water per day. That tracks very well with the original estimate of 5,000 gallons/acre/day, which would come out to 16 million gallons per day for this farm.
This company says a 27 MGD plant cost $87 million. If cost scales ~linearly with capacity, that suggests a $50 million price tag for the desalination capacity—now perhaps 4 times the normal cost of a farm of equivalent size. That puts the cost of desal much higher than the lower bound I used from the OP.
They also claim a unit cost (including capital cost, debt service, and operating cost) of at least $1.25 per 1,000 gallons of desalinated brackish water, which would therefore cost about $20,000 per day for this farm and create an added $3 million cost per year.
I don’t have a great source here, but it looks like groundwater use for farm irrigation is typically free. Corn and soy yield optimistic profits of $100-$200/acre, or perhaps $320,000 for a 5 square mile farm. A $3 million cost per year is again an order of magnitude higher than those expected profits.
That’s more like the sort of picture I was expecting.