Hacker News Reader

Comment by kumarski | Hacker News Reader

My father is an ag soil chemist of 50+ years.

I'm an industrial systems eng. w/ a specialty in polymer-textile-fiber engineering. (Mostly useless skillsets in the US now)

Gonna share a few lessons here about agriculture that I try to convey to EECS, econ, Neuroscience, and the web developer crowd.

- You can only grow non-calorically dense foods in vertical farms

- It takes 10-14 kwh/1000 gallons of water to desalinate. More if it gets periodically polluted at an increasing rate.

- Large majority Agrarian populations exist because the countries are stuck in a purgatory of <1 MWh/capita annum whereby the country doesn't have scaleable nitrogen and steel manufacturing.

- Sweet potatoes and sweet potatoes are some of the highest satiety lowest input to output ratio produce. High efficiency.

- In civilizations where you are at < 1MWh/capita annum - there is not enough electricity to produce tools for farming, steel for roads, and concrete for building things. The end result is that the optimal decision is to have more children to harvest more calories per an acre.

- Property, bankruptcy, and inheritance law have an immense influence on the farmer population of a country.

I remember telling some "ag tech" VCs my insights and offering to introduce my father who has an immense amount of insight on the topic from having grown things for as long as he has....My thoughts were tossed aside.

View on HN

> 1MWh/capita annum

Oh this is fascinating! I never thought of this but of course energy consumption per capita is going to be an indicator of how industrialized a country is. I briefly checked the two countries I am a citizen of (Canada, Hungary) and counterchecked with one of the poorest countries I know of (Chad) and the numbers are as expected: 14.6, 4.1, 0.013 (oof).

Yes, and for this reason (IIRC) GDP/capita scales linearly with MWH/capita.

On the topic of energy, I find this correlation fascinating as well: https://jancovici.com/wp-content/uploads/2016/04/petrole_gra.... Oil consumption predicts GDP. Complete article: https://jancovici.com/en/energy-transition/oil/is-the-price-...

Now think about what it means to do with less energy.

Energy availability is wealth.

I highly recommend Energy and Civilization[1], the most engrossing dry recitation of facts I've ever read.

[1] https://mitpress.mit.edu/books/energy-and-civilization

> You can only grow non-calorically dense foods in vertical farms

Purslane (Portulaca oleracea) grows on the sidewalk already, and often next to some wild amaranth (Amarathus Hybridus). What is the point of more efficiently producing specific crops, when there are all these underutilized nutritious plants growing without any human input (or should I say growing despite human input)? This is another problem that I see with the technification of food production in general (including the Green Revolution). Some food wants to be free, but people keep looking for whatever makes the land produce more money in the short term, not what makes it produce more nutrition, etc., because the latter does not adapt so well to the market.

I think this is the type of magical thinking that pervades a lot of these ag startups.

Just because you noticed "Hey, this unpopular grain grows so easily it's popping up from the sidewalk!" does not mean you can actually then go and scale that to providing food for hundreds of millions of people.

We developed modern agriculture the way we did for a reason. Some of those reasons are no longer valid, true, but a whole lot of them are still very very pertinent.

I think vertical farming / rooftop farms / etc are at this point largely an exercise in virtue signaling more than they are actually improving the food system. They might be improving it for high-income Whole Foods types who are upset Amazon owns their favorite store now (ie: us here on HN) - but those people aren't exactly the ones who need to see improvement.

Yes I agree with you, but maybe I did not express my point with enough clarity. I mentioned this two species because they count as food of negligible caloric value (sidewalk amaranth is mostly useful because of the edible leaves). As far as I understand, human caloric needs at present population levels require the large scale conventional methods of production (including the Haber-Bosch process). But caloric needs aside, there are populations, which are not consuming enough greens or fruit, even when it grows in front of them. I have seen it in the (warm and tropical dry) countryside here, they think they have to eat lettuce, when they have lots of hibiscus plants growing in their garden. It turns out the leaves of hibiscus aren't just edible, but their nutritional value is similar lettuce (I suspect higher in Mg since it's a lot greener). The answer of some technologists to this problem is to grow more lettuce. Another plant that grows plentifully here certain months of the year is Ipomoea Triloba. I don't think anyone should try to produce them at scale following the conventional methods. In the case of cities I imagine there could be gardens with a synergistic diversity of plants, which try to keep a high total yield, but not a high yield of any specific plant.

Not the original person, but I agree with you. A decentralized food system does not require as much of the distribution network we have, nor does it require scale.

There are food sources that can be had if they were grown local, and grown in a way that makes sense for that locality. By "local", I mean at the neighborhood, and home level. They can be grown synergistically. Even the old, but simple idea of (thoughtful) companion planting can create better yields than monoculture crops.

Our current food system is optimized towards the kind of scale that allows an easier way to control and meter food resources. VCs chase after multiples on investment, and are not incentivized to truly decentralized food systems.

For others following along, there _are_ already people who have guardens with a synergistic diversity of plants. And not only that, have practices that, over time, create yields that require less time to maintain. The permaculture design community have been exploring and implementing these ideas for over 50 years, and people have gotten them to work. They may not have achieved 100% self-sufficiency, but you would be surprised how much they are able to achieve by synergizing not only plants, but also animals, and human social inclinations into the whole system.

Some people here in Australia do pick dandelion leaves and other wild greens, but the general advice (unsure if backed by evidence) is you probably don't want a lot of it in a larger city because of roadside pollution.

I was happy when parsley sprung up like a weed around my house, but it all died out in the drought. Hopefully it comes back soon.

The human input is providing the moisture trap (the concrete lid on the ground preventing water vaporising into the air), so of course plants will thrive there once the level of lime leeching out is low enough to be tolerable.

But in the end you have several tons of concrete covering tens of square metres of ground, to allow one or two square metres of cereal grains to grow. That is not particularly efficient use of time or resources.

I'm not sure why you think the plant will only grow with a concrete lid on the ground. That plant thrives in the marginal spaces in the sidewalk with poor soil quality, and can probably do better if allowed grow in the lawns where we normally cultivate inedible, ornamental grass.

Furthermore, there are several techniques to help trap moisture that does not require a concrete lid. I'm in the Phnoenix desert, and I use the trimmings from my zucchinni plant as green mulch. Along with the palm fronds that my neighbors love to trim, they create very effective moisture traps. I need to trim my zucchinnis anyways for their optimal health.

I could also plant things in a way to take advantage of the vertical spaces as well. Okra, for example, thrives in the low desert, and provides dense shade that already traps moisture. If I add something that likes shade, and stays closer to the ground, that forms a kind of living mulch that does not require a lot of ongoing maintenance. And if I plant pole beans to take advantage of the sturdy okra stalks, they fix nitrogen and help keep the soil fertile.

I didn't come up with these ideas. They may not be widely known, but people have come up with these design patterns and used them well.

That's right. I've seen purselane thriving along with other weeds away from any concrete, growing taller and with bigger leaves during the first month of rains. A week later someone comes with a machete, and that's the end of it. I do think however that most edible weeds require some sort of disturbance that human presence tends to bring.

same exact boat.

the VC that use to approach us for insights would just never listen. my father literally knew the researchers that tried it in the past and failed.

it did not stop this VC from investing his LP's money in a vertical farm. although i suspect his willingness to allocate other people's money in this manner, for this particular company, had more to do with the social side of things re the founders and other investors.

> Sweet potatoes and sweet potatoes

Is this a typo? Did you mean to write a second vegetable?

Probably regular potatoes; they are incredibly calorie-dense and resilient.

Great comment, thanks.

HN just keeps delivering. It is almost impossible to believe how much embedded technical knowledge is lurking here. You could colonise Mars with it.

> HN just keeps delivering. It is almost impossible to believe how much embedded technical knowledge is lurking here. You could colonise Mars with it.

Of course, in this context you should be prepared for the possibility that if you could tap all that expertise what you'd actually get is "reasons that colonizing Mars actually can't work". ...hopefully not, of course, but beware mixing hopes and dreams and reality.

Something as complex as colonizing Mars would really need to be tried before anyone should claim to have expertise on if it's possible. It looks like our current technology and knowledge is up to the task, so I'd say it's worth seeing if humanity can do it.

Re: <1 MWh/capita annum

Which brings up another issue, which is energy density/m2 of land. To support industrialization/high density urbanism the only fuel sources that do this are currently fossil fuels, or nuclear, but none of the renewable fuels have the energy density.

So if these countries want to increase the amount of MWh/capita, the most efficient (only?) pathway is through high-energy density fuel sources, which right now is being achieved through the use of fossil fuels. To me, this is (one of) the main reasons nuclear energy needs to be prioritized as a climate change solution.

ETA: And, now that I think about it, another way to squeeze more effectiveness from your grid is to build super energy-efficient buildings that reduces the overall and peak grid energy consumption.

Why is energy density important in that context? If all other variables were identical between a high density an a low density solution, the high density one would of course be preferable. But if the low density solution is cheaper and relies less on pre-existing long-distance grid infrastructure, why would high density still be considered the most efficient, or possibly the only viable pathway?

IANAE but the general idea from the discipline of urban geography is that industrialization, and also the knowledge economy relies on the economies of scales that come from high density, mixed-used urbanization. These regions rely on extremely high power that must be supplied from scarce land resources.

I'm not going to provide a source, but my understanding is that one square meter can provide roughly 1 KW of electricity at noon on the equator. Looks like that's about 4MW per acre. Five acres for the "traditional" family farm. I bet where I live that irradiation equals 1MW per acre, and of course it varies seasonally. Sounds ridiculous, until you've tried to water any sizable amount of square footage, and contemplated the heat energy required to evaporate all that water which doesn't go into the plants and isn't runoff.

1 KW sounds high to me. Granted I work in higher latitudes (42). Rule of thumb is 1KW for peak solar radiation, which translates 100W from a solar panel (which are 15% - 20% efficiency).

And the sun in eight light minutes away. The sheer power is mind boggling.

Hmm. If the target is 1MWh in a year, that seems trivial to deliver in a place like Chad. Insolation in Chad should be right around 6 or 7 kWh per sq metre. [0] Chop off 85% of that, because we can't harvest all of the solar energy that falls on a square metre of land. We can only reliably capture maybe 15 to 20% of it. But even if you harvest that 1.5 kWh for only 4 hours a day, (which is pretty easy in the middle of the Sahara where you should get about 8-12 hours a day), it comes in around 120 kWh per month. Well above the 1MWh per annum target.

So on the distribution side, they have a population of only 13 million, with a geographic size of 1.2 billion sq metres, which lands us at roughly a thousand times the amount of energy they need to meet the 1MWh per annum per Chad citizen. If they were harvesting it that is. (And that's using 20% efficient solar panels alone. No wind, etc.)

Chad's problem is a lack of anything to trade in exchange for the equipment to harness the sun and build the storage and distribution network. (Grid distribution infrastructure, maybe a liquid air storage facility or 3, solar panels etc) But the sun itself provides them more than enough energy to meet that 1MWh per annum per capita target.

[0] https://en.wikipedia.org/wiki/Solar_irradiance

Solar panels themselves are only 10% - 15% efficient at converting solar to electricity. So I think your estimate here is off by a factor of 10.

Grandparent estimated capturing 15%-20%, so they’re only off by a factor of 0x-2x from your estimate range.

If you google “solar panel efficiency”, you will instantly find several sources that all indicate that most (cheap!) solar panels are 15% to 20% efficient, which is 50% higher than the number you’re providing. Top end residential solar panels are over 20% efficient, somewhere in the neighborhood of 22%. Specialized/research solar panels can reach efficiencies of over 40%.

Also, energy density is not nearly as important as you seem to fundamentally believe. It’s pretty much all about cost and availability.

In countries that use very little energy per capita, even just a single solar panel (with a small inverter and a lead acid battery) per person (or even per family) is life changing, giving them enough energy to run a small refrigerator, a light bulb, and a place to recharge their smartphone without having to go into town and pay someone.

In deep urban environments, space is definitely at a premium, but I struggle to imagine that many people are excited to set up a dirty diesel plant in the town square when air quality is already bad enough due to heavily polluting vehicles running in the city center, let alone set up a small nuclear plant, even if wealthier countries would allow such a proliferation of nuclear technology and fuels. Once the nuclear fuel is spent, then you have this extremely toxic, dangerous waste that has to be put somewhere, and those people probably have a lot of other things on their minds, so you could just be rad poisoning their town by giving them nuclear if they don’t dispose of the waste properly.

On the other hand, the rooftops are prime locations for solar panels and solar water heaters, and because 175W/m^2 of solar converted electricity is actually a ton of energy, it’s still plenty of power.

Unfortunately, cost and availability are king. Wealthy countries are happy to sell old assets at a steep discount, and so shiny new solar panels have to compete with second hand fossil fuel plants / generators that don’t lend themselves to great air quality.

Solar is an extremely sensible solution. Nuclear is not a clear answer at all for developing countries, unless you happen to have the design for a clean, portable cold fusion reactor in your pocket. Fission has a lot of problems that are manageable, but managing those problems takes significant money.

Your comment about a solar panel and battery reminds me of my van with solar panel and 100ahr lead acid battery and a 2000W inverter.

Things I've done with it. Run a Engel Cooler. Run a microwave. Boil water. Probably 1-2 gallons a day. Charge batteries for power tools. Run a skill saw. Power lights indefinitely.

With just that you're getting close to a middle class lifestyle.

So you can't grow potatoes vertically? Can you elaborate? Is it a function of physiology, i.e. calorie dense vegetables need far more leaves and supporting stems than can be practically stacked vertically?

I imagine space is a factor, but energy will be a big one as well. Calorie dense foods will likely need more space and energy (light) inputs. Vertical farms are very water efficient, so I don't think that matters much.

Vertical farms make a lot more sense with fresh vegetables like leafy greens that grow quickly, command high prices if grown organically, and benefit from being closer to market.

Potatoes are the exact opposite. If it ever becomes more cost effective to grow corn, wheat, and potatoes in virtual farms then outdoor agriculture is dead. While I don't agree with the article that it will never happen, it might require energy advances like fusion power or drastically higher _rural_ land values and water prices.

Greenhouses make sense long before vertical farming, just look at agriculture in the Netherlands, it's mind boggling how much they produce for such a tiny country.

Can you expand on this?

I get that to store a calorie in a potato I need to supply a calorie of energy from somewhere else.

But why is fusion power required instead of better UV lamps in my vertical farm? (Assuming I had enough electricity to run them)

I think the suggestion is that fusion power here is a stand-in for electricity that is too cheap to meter -- basically free power.

Yes, that's exactly what I mean, sorry for skipping a step in my logic.

The total amount of electricity to power those UV lamps should be on par with what the Sun sends to the potatoes fields. Maybe that's the reason for fusion. It didn't do the math.

Actually no not really. Plants only absorb two wavelengths of light. It's currently more efficient to convert sun into solar power via panels and then to light LEDs supplying only the wavelengths that plants use. Despite the seeming inefficiency here, the fact is that plants are even more inefficient at absorbing light not at the right wavelengths than solar panels.

Could one imagine a material that would absorb solar spectrum and emit the preferred frequencies? Something like a polymer one could stretch over fields to get more from the suns rays.

>> "Actually no not really. Plants only absorb two wavelengths of light. It's currently more efficient to convert sun into solar power via panels and then to light LEDs supplying only the wavelengths that plants use. Despite the seeming inefficiency here, the fact is that plants are even more inefficient at absorbing light not at the right wavelengths than solar panels."

> Could one imagine a material that would absorb solar spectrum and emit the preferred frequencies? Something like a polymer one could stretch over fields to get more from the suns rays.

Would you call that a "solar transmitter"?

https://en.wikipedia.org/wiki/Transmitter :

> Generators of radio waves for heating or industrial purposes, such as microwave ovens or diathermy equipment, are not usually called transmitters, even though they often have similar circuits.

Would "absorption spectroscopy" specialists have insight into whether this is possible without solar cells, energy storage, and UV LEDs? https://en.wikipedia.org/wiki/Absorption_spectroscopy

(edit) The thermal energy from sunlight (from the FREE radiation from the nuclear reaction at the center of our solar system) is also useful to and necessary for plants. There's probably a passive heat pipe / solar panel cooling solution that could harvest such heat for colder seasons and climates.

Also, UV-C is useful for sanitizing (UVGI) but not really for plant growth. https://en.wikipedia.org/wiki/Ultraviolet_germicidal_irradia... :

> UVGI can be coupled with a filtration system to sanitize air and water.

Is that necessary or desirable for plants?

https://www.lumigrow.com/learning-center/blogs/the-definitiv... :

> The light that plants predominately use for photosynthesis ranges from 400–700 nm. This range is referred to as Photosynthetically Active Radiation (PAR) and includes red, blue and green wavebands. Photomorphogenesis occurs in a wider range from approximately 260–780 nm and includes UV and far-red radiation.

Photomorphogenesis: https://en.wikipedia.org/wiki/Photomorphogenesis

PAR: Photosynthetically active radiation: https://en.wikipedia.org/wiki/Photosynthetically_active_radi...

Grow light: https://en.wikipedia.org/wiki/Grow_light

Are there bioluminescent e.g. algae which emit PAR and/or UV? Algae can feed off of waste industrial gases.

Bioluminescence > Light production: https://en.wikipedia.org/wiki/Bioluminescence#Light_producti...

Biophoton: https://en.wikipedia.org/wiki/Biophoton

Chemiluminescence: https://en.wikipedia.org/wiki/Chemiluminescence

Electrochemiluminescence: https://en.wikipedia.org/wiki/Electrochemiluminescence

Quantum dot display / "QLED": https://en.wikipedia.org/wiki/Quantum_dot_display

Could be possible? Analyzing the inputs and outputs is useful in natural systems, as well.

Even if artificial lighting and natural lighting were equally efficient you would still have to cover the same number of acres in either solar panels or plants. In other words, the denser your vertical farm the more land it consumes. Clearly vertical farming is meant for some really exotic situations in which you might have access to electricity but are in an environment in which you can't grow the plants you want. That situation would probably be a mars colony or a fallout shelter.

Well, it's more efficient (I forget by which factor). So, if the factor were 2, you would need 1 acre of solar panel for every 2 acre-equivalents of planting space. Since you're okay with going vertical, the acre-equivalent could be much less than an acre. So yeah... electricity being much easier to transport than produce, it would actually make more sense to cover a large part of unpopulated area with solar panels, and then farm in very very small portions of the earth.

From an environmental perspective, this is certainly a 'win'. Reducing the amount of land needed for agriculture is a win to both consumers and the environmentalists and the farmers.

What powers the UV lamps?

Sun + water is cheap and plentiful. Small scale farms can sell potatoes at $0.50/lb or less. Amish farms with oxen can go a little less.

Capital and operational costs for vertical farms don’t seem to make sense, unless there’s some disaster in the Colorado watershed or a trade war that makes hothouse winter produce a viable business again.

This is totally unrelated, but I saw people plowing fields with oxen in Cuba. A scathing indictment of socialism if ever there was one[1]. It's easy to forget at times how large parts of the rest of the world live.

[1] Don't get me wrong, it's a spectrum and some socialism is a very good thing. But not like in Cuba or the old USSR.

>But why is fusion power required instead of better UV lamps in my vertical farm?

Because of the second law of thermodynamics. Your UV lamp is not going to produce light that contains more energy than the electricity you used as input. That energy needs to be produced via solar panels if you want maximum efficiency. If we ignore nuclear or fusion all energy on earth is derived from sunlight.

And geothermal.

You probably can grow them (you can "grow" a potato in a cup of water on your counter), but probably not profitably. Potatoes have a fairly low commodity price relative to their light and space demands. Additionally, they store and transport really well.

Amateur hydroponicist here. It depends on what you mean by 'vertically'. You can certainly grow potatoes in containers that are stacked vertically. You can also grow them hydroponically. However, the issue I've noticed in the hydroponics community is that no one is interested in growing potatoes. That is really the problem with these vertical ag startups and such. They focus on ridiculous foods like greens, which -- while nutritious and easy to grow -- cannot form the bulk of a human diet.

As a community, vertical agriculture need to focus on high calorie crops like potatoes or sweet potatoes or at least something useful like beans.

But circling back to the beginning. You can't really grow a potato with less industrial input vertically than you can with regular land, so unless you are really out of land (and the United States at least is not running out of land anytime soon), it doesn't really make sense to do so. Potatoes are really easy to grow -- you stick them in the ground and dig them up a few months later. Anyone can just buy a few acres of land, fertilize it, stick in some seed potatoes, and get a pretty decent crop that more than covers their costs. This is currently way easier than the amount of setup it would take to use containers. If you were to use conventional growing containers, you would need to import large amounts of soil / substrate. If you were using hydroponics, you'd also have to buy large amounts of hydroponic substrate or expensive nozzles for aeroponics. Either way, it's more expensive.

Anyway... wish me luck, I'm starting some potato growing experiments this summer to see if I can develop new container, vertical, and hydroponic techniques. I'm particularly interested in growing potatos without a substrate and without expensive aeroponics. Currently investigating 'aeroponic' drip systems.

Do you have a blog or a youtube channel where we could follow your endeavor?

No not really unfortunately. I'm not great at that kind of thing.

> The end result is that the optimal decision is to have more children to harvest more calories per an acre.

For a photo-essay about this there's the New Humanitarian article here: https://www.thenewhumanitarian.org/report/94947/lesotho-weat...

It describes the interaction between climate change, HIV/AIDS, and poverty.

(It used to be called "Too poor to farm").

I guess I was always under the impression that vertical and urban farming would be done for "specialty" crops like herbs or kale or something, never for high volume cash crops like potatoes or corn. I can see a benefit for these "specialty" crops because they aren't done to the same scale (maybe I'm wrong about that)

Yeah that was my take as well. I think there's some sort of trade-off point here, though I don't know where it is. Yes modern outdoor agriculture is hyper efficient, although I think the author's comments about cash flow self-sufficiency gloss over a lot of government subsidies and bank bridge loans. In any case, indoor ag should be able to exploit the lower weather/pest risks, lack of need for damaging pesticides, consistent conditions, 365 day growing season, proximity to markets, etc. at least for some combination of products.

Even for specialty products, you’re competing with cheap travel and labor.

Guatemala or Peru are short hops.

The parents comment certainly is accurate for most staples, cereals, and even meat (raised on corn) but entirely misses the need for more fresh veggies and leafy greens in the US. In the states it generally costs more to buy food made with fresh vegetables. That’s where indoor and vertical ag shines. Many developed countries have plenty of empty calories. Too many. Indoor Ag provides an opportunity to provide scale and freshness to a number of non-grain foods, with less pesticides and preservatives. I hope the field is successful at that goal.

So you plan is to replace calories with mostly empty (calorie and nutrient-wise) greens...

What is it about vertical farms that prevents calorically dense foods? Also, I know something like lettuce would not be dense, potatoes probably dense, but is there a cutoff/metric, e.g., calories-per-gram for this determination?

Thanks for the insights!

I'd imagine it's sunlight / artificial light + growth to harvest time. It takes energy input to make food calories, there's really no getting around it with technology. The only way to increase the calories in the harvested food is intense direct light in a short time, or less intense (shaded) light over a longer time. In vertical farms, that's a higher cost of production.

Hi You write: Sweet potatoes and sweet potatoes ?Could this be a typo what's the other one please?

> - You can only grow non-calorically dense foods in vertical farms

How about seeds and sprouts?

I grow sprouts on my kitchen counter. Mostly they require lavage. For the last day or two they require light to turn green and possibly develop flavor; they don't need much.

> How about seeds

You need adult plants to produce seeds, defeating the purpose.

> and sprouts?

Alfalfa sprouts? That's very much a non-calorically dense food.

Just one thought, everybody can buy an imported tractor these days. And steel too.

Reply on news.ycombinator.com