One question I have had on my mind is about solar powered LED lighting.
Solar cells capture a wider spectrum than plants (plants are green ergo not green eaters).
At what levels of efficiency do solar cells and LEDs have to be before you receive a net gain by having the solar cells in the sun instead of the plants. How far away from that are we now?
Underground farming in arid regions seems like the best scope for something like this.
Not an expert so I can't vouch for its veracity, but I'll inline it:
"This can theoretically be more efficient than growing plants under sunlight. Chlorophyll mostly absorbs light in the 400nm to 700nm band which is about half of all the photons that hit lights. Even in that band ~24% (https://en.wikipedia.org/wiki/Photosynthetic_efficiency) of the energy in them in lost due to the higher energy photons (closer to the 400nm side) being converted to lower energy photons.
"So right there plants are only 25% efficient at collecting energy from photons. This is why you see indoor plants grown under red LEDs. Modern commercially available solar panels are about 21% efficient so that is actually getting close to plant can do (I'm not going to include inefficiencies in power conversion right now). Solar panels in the high 40% efficiency range exist, but they are crazy expensive and only used on things like satellites.
"So using current (very advanced) technology we could actually get about double the plant output per area land by using the most advanced solar panels we have to collect light energy to power LEDs."
Assuming zero production costs, 100% efficient LED's, 100% efficient energy transfer, 100% land coverage with Panels that perfectly tile the surface, zero panel degradation, zero dust, etc etc.
In the real world wind energy probably works out much better.
This math is incorrect. Look at the wikipedia article percentages - each one is a percentage of the preceding one, not of the total light hitting (if you add up the percentages, they sum to way more than 100%). The 24% figure is 24% of 30% of 47%, or about 3% in actuality. 40% efficient solar panels also do not exist as far as I know (and are impossible) - I think the user is confusing other forms of solar energy collection with photovoltaics. See my other answer below for a better breakdown.
Yeah, but imagine this on deserts. You havae too much light and almost no agricultural land. So, use panels to gather light, use hydroponics/aeroponics to grow plants and now you can distribute your sunlight to many levels of racks (which can be underground under panels which means even more efficiency than on normal land.
Not sure if you are responding to the correct parent?
Quote:
Sure it's great for optimizing land use. Agricultural land is cheap. The real question is cost.
Yes, deserts are even cheaper than agricultural land. But that isn't the problem.
The problem is the upfront capital cost of building an underground greenhouse system, the solar panels and getting the water.
Yes, once you have all that, you can produce more efficiently than normal production. And in a couple of hundred years[1] you will actually be ahead.
[1] No, I haven't done the maths. Maybe it is 10 years, but the point is still the same. It's never going to produce the profits needed to make an investment of that magnitide[^1]
I think Las Vegas would be a great place for vertical farming. Heavy local demand, constrained water usage, cheap power, high ambient temperature. What's not to like?
The cost of building large 5+ floor concrete structures that can take significant floor loading in kilograms per square meter? Building vertical anything isn't cheap, this is why you see massive cloud scale datacenters built into "shell" warehouse buildings that are tilt-up concrete walls, the whole thing 1 story high sitting on a concrete slab. Usually in some places where land is cheap in $/square meter. Example: Datacenters in The Dalles, OR.
Actually, they do use green - they're more reflective in green than red/blue, but it's still mostly absorbed.
Apparently at high illuminations you want to be using green too - the issue is that red and blue are absorbed so well that they don't make it past the surface of the leaf and thus the plant quickly maxes out on them, while green light penetrates further and thus reach more tissue.
So when you see pictures of leaf greens being grown under purple light - that's because leaf greens actually need less light than other vegetables, you probably wouldn't want to do that with other crops.
This is why we still use high yellow/green CFLs per so many red/blue LEDs. It's not all about getting those high peak spectrum binned LEDs for each type and concentration of chlorophyll.
I don't have a definitive answer for you, but you can kind of back into a rough approximation (someone please correct my math if I went astray).
High-efficiency solar cells are about 20% efficient across the daylight spectrum.
For some reason, no one gives hard stats on how much energy LEDs waste as heat, but the claims are that incandescents waste about 90% of energy input as heat, and LEDs are about 80% more efficient than that. So let's say LEDs convert about 80% of input power into light at their operating wavelengths as a guess.
So if we put these together, we'll get about 16% of the input daylight back out as light at the wavelengths we want.
According to wikipedia [1], about 47% of light that falls on chlorophyll is wasted because it falls outside the ideal range for plants (400-700nm). Another 3% or so is lost due to wavelength conversion. The rest we can't really do anything about because it has to do with biological limitations. Put another way, our total system has to be about 50% efficient to be better than just putting the plants in the sun. As you can see, we're not even close. In fact, it's not even possible since the theoretical maximum efficiency of solar panels is only about 34% [2].
Update on this for future reference: using multi-junction solar cells, the theoretical ceiling for efficiency is about 68% [1], and we can probably assume LEDs will get up to about 90% efficiency in producing desired wavelengths. Thus in an ideal future scenario it might be possible to do better with solar collection + LEDs than plants themselves do, but we're pretty far off in actuality.
The Shockley-Queisser limit only applies to single-layer cells under unconcentrated light. Different cell designs have surpassed it by significant amounts in practice already.
Do you have a link where I can read more? I'm aware of light concentrators being more efficient, but that defeats the whole point since the surface area required is far more than size of the panel itself.
Fair point, but that's not really helpful in this particular situation since plants can't use heat as an energy source. We're only concerned with the % of input power that is converted into the specific wavelengths we actually want (400-700nm).
> Underground farming in arid regions seems like the best scope for something like this.
This sort of mediated growing environment seems quite interesting as an approach to managing agriculture as environmental changes from global warming progress. In some areas, perhaps conventional greenhouses would suffice alone. But in others, underground agriculture might provide an additional buffer against climate changes. This would allow normalization of growing conditions in regions where the climate has drifted and/or become unpredictable in ways that would severely damage agricultural output.
Actually if you're doing underground farming and such like that, it could be a net positive pretty quickly because you could be swamped by other concerns such as available water for growing and the energy costs to get it.
Aquaponics uses much less water than hydroponics plus you don't use energy to make fertiliser. Your inputs are light & fish food and some top up water.
I'm a big fan of aquaponics too, but I've never seen numbers on the energy required to source, process, and transport the fish food. It gets even more complicated if you include the energy required to raise the fish that get caught and turned into fish food.
In my old system, I used bottle fly larvae to feed the fish so the energy input was less (waste refuse that would have been composted anyway) . It was the least fun part of the system because of the "gross factor" of maggots.
I live in Cleveland currently and in the winter time there is no sun. Thus, LEDs are an excellent alternative, and can be used elsewhere during nighttime. The benefits are not "bringing the sun via proxy" ... the benefits are "we have focused on the wavelengths that most rapidly and efficiently excite chlorophylls alpha and beta (harvest sun color and the noon-time color of the summer, if I may approximate the "feeling" of the two active bumps... carotenoids are the main chlorophyll in carrots).
Plants do still absorb green and are arguably better off if they do (although the latter may be less certain). The bits that absorb green are more sparse, but light will bounce around inside the leaves until it hits one.
The next step is optimizing the plants to grow better in these greenhouse conditions - currently the plants waste lots of energy. For example my company, TAXA Biotechnologies[1], has identified a gene that doubles the rate of growth of lettuce per unit light energy. It wouldn't work in the wild as the cost is reduced pest-resistance but growing indoors there are other ways to control that.
[1]www.taxa.com
Yes, this is a must. I wonder if industrial greenhouses could benefit from some sort of induced day-night cycles and thermal gradients to cuddle the little plants. Maybe it will allow them to grow better and to develop a better taste. Once you have LEDs possibilities are many.
Last time I checked (it's been a while since I had a salad), lettuce already doesn't have too much of a taste. I'd be surprised if they managed to mess that up.
Moving to California recalibrated my understanding of lettuce/salad greens. Not sure if you meant just lettuce, but in California a typical salad has little to no lettuce, and instead: mizuna, arugula, radicchio. What lettuce is does have is Romaine and butter lettuce, not iceberg.
The result is a very flavorful salad- intensely bitter with much less crunch.
I'm not optimistic here. At scale, cost always wins. Look what Hershey did in the US to its brand also Cadbury when they licensed it: that pasty brown wax is just not appealing compared to the real thing.
Taste is a really difficult thing to optimize for due to the subjective nature. There are certainly specific chemicals that appear in really tasty food that can be analyzed and used to guide the breeding process but it's incredibly complicated.
Optimizing for anecdotal data is one of the things I'm working on, it's amazing the depth you need in selecting specific traits to breed into plants.
>Taste is a really difficult thing to optimize for due to the subjective nature.
No one tastes locally grown produce and thinks, "this is OK, but I really prefer the flavor of the produce that's genetically engineered to not bruise as easily in shipping."
Or maybe I have and I can recognize that lettuce is one of the blandest greens around. Treviso, watercress, frisée, oak-leaf have way more taste and also offer better texture.
There is a lot that can be done to optimize plants genetically, including a ton that can be done with selective breeding. I'm working on a start up doing selective breeding software and it's amazing how much variety is available that isn't being tapped into yet. We're starting with cannabis breeding before moving into other crops but there is a ton of room for improvement in genetic stock.
Hey, I'd be very interested to hear more about this! Do you have a newsletter or something like that?
I had the idea that there is so much untapped potential in selective breeding of cannabis after reading about various terpenes and their medicinal properties.
The market's just been breeding for ridiculous THC levels (I daresay harmful levels - unfortunately I know several heavy smokers who have been institutionalized) without the CBD to balance it and without exploring the role of the other chemicals.
So I'm happy to hear that this is changing! (Like that news of a 15% CBD 1% THC strain from a few years back)
To clarify on the last bit, I was thinking it would be great if the genetics from the new high CBD strains could be brought into the mainstream by crossbreeding with popular strains.
My interest in this has to do with the idea that in the 1970s, CBD and THC levels were both around 1%. THC has skyrocketed but CBD has not.
THC is known to trigger psychosis and paranoia in sensitive individuals, while CBD reduces anxiety and was found to be as effective as other anti-psychotics.
Hey, we don't yet could you email me patrick@growflora.co ? I'd love to talk about it more. We're onboarding some of our first breeders this upcoming week and have a grow journal beta out in the open.
nice to hear but more relevant to the thread would be info on precise photo active radiation spectra. The absorption is easily meassured putting a leaf under a through light spectroscope, but narrowing down the actually active spectrum is a different matter.
They do not seem to have accounted for the fact that the "waste" heat from current solutions reduces the overall heating bill. Of course, electrical heating is not usually as cost effective as other methods.
Greenhouses in NL use natural gas to heat and they are getting a serious discount compared to homes and offices, compared to what they are paying for gas heat any conversion from electricity to heat is a loss to them.
Greenhouse farming really could be a good way to help tackle the waste/runoff problem. I think Greenhouse farms could easily be closed systems with no need to push polluted water back into the world. The dead zone in the gulf of Mexico (all ocean dead zones actually) HAS to be dealt with pretty damn soon as I understand it...
Ceramic Metal Halide bulbs, at much lower wattage than HPS, but with more power than LED seems to be the immediate compromise. A 315w Phillips Ceramic MH will deliver a huge spectrum to the plants with less than half of the heat of a comparable 600w HPS/MH. The cost is getting very close to traditional HPS lights. LEDs do work, but they still use plenty of power, and are largely unproven. Try to find a greenhouse/grow store that stocks ANY led fixures. They have been cheap China crap for so long, that the better ones cost quite a bit of $.
Solar cells capture a wider spectrum than plants (plants are green ergo not green eaters).
At what levels of efficiency do solar cells and LEDs have to be before you receive a net gain by having the solar cells in the sun instead of the plants. How far away from that are we now?
Underground farming in arid regions seems like the best scope for something like this.