> Solar panels are maybe 30% efficient. So 10-acres of glass-roofs need to be replaced by 30-acres of solar panels

Not remotely. Plants are also extremely inefficient, converting only about 1% of the solar energy that falls for their use. [1]

Most of this inefficiency is from solar energy being in the form of frequencies that the plants can't use, but solar panels can. So the panels can capture this energy, then funnel into the red and orange lights that are most efficient for plant growth.

I've read a bunch on this, and haven't been able to find an authoritative source for what the efficiency conversion is -- how many acres of solar panels power how many acres of vegetables, and is it greater or less than 1:1? -- but it's certainly not as simplistic as "solar would need 3x more land because they are 30% efficient."

1. https://phys.org/news/2012-01-energy-conversion-solar-cells....

The article's numbers are that 1m^2 of wheat needed 20m^2 of solar panels.

That implies a 5% efficiency in the "total solar captured area" of the solar panel + LED lights compared to just sticking the plats out in the sun, which is totally independent of the plants own efficiency in photosynthesis.

(I guess coming from a site about Low Tech, their slant on the numbers might be questioned, but they certainly hold up to initial scrutiny from here...)

> I've read a bunch on this, and haven't been able to find an authoritative source for what the efficiency conversion is -- how many acres of solar panels power how many acres of vegetables, and is it greater or less than 1:1? -- but it's certainly not as simplistic as "solar would need 3x more land because they are 30% efficient."

I own a townhome, so my only real ability to grow plants is through a grow-light connected to electricity.

As such, I've spent some time calculating the PAR values of a decent grow-light, as well as the amount of PAR that natural sunlight gives. Plants need a ludicrous amount of PAR (basically blue + red lights, green not needed cause green just bounces off of plants).

Sunlight is mostly broad spectrum: broader than plants need and therefore a source of inefficiency (green light is wasted) that LEDs can somewhat replace.

Grow-lights have a benefit that they can be placed very close to the plant (maybe just 1-foot away) to "focus" the energy a bit better. Nonetheless, the amount of PAR / PPFD from a typical day sun (or even a cloudy day) far exceeds what you'd get from 500W or even 2000W grow lights.

--------

Its just a hobby of mine, and I'm not growing anything especially hard (just Basil, which is really easy to grow... but Basil is a summer plant that really wants sunlight).

Still, once you start calculating PAR and actually mapping out how much electricity your "emulated sunlight" needs, you'll realize how grossly inefficient that "solar panel -> electricity -> LED" plan really is.

EDIT: Natural sun is like 2000 PPFD or something FAR in excess of what most plants need. Still, a good growlight solution might hit ~1000 PPFD constantly. Lets take this 650W LED and think about it: https://allgreenhydroponics.com/collections/american-made-le...

You'll get ~500 to ~1000 PPFD across a 4'x4' or 16-square foot area from that 650W LED (and most of that light is focused on the center: you'll want to overlap your lights a bit for more consistency).

Then think about how much solar panels you need to power a 650W LED for the 16-hours / day your typical plant would want (to account for the lesser PPFD indoor plants get, you run the lights for a bit longer than sunrise-sunset).

Just some napkin math. Nothing serious here: just guestimating the area in my head.

----------

EDIT: Now it should be noted: I've heard of good hydroponic greenhouses that have the "do both" approach: glass roofs to let the sun in most of the time, and LEDs to augment the natural sun (cloudy / rainy days, as well as winter-settings when you have fewer hours of sun). The sun isn't nearly as consistent as we'd like, but... that means that you need something aside from solar power powering those LEDs.

But the concept of building a all-LED underground (or "inside a building") without any natural light just... seems grossly inefficient to me. Such a setup only seems useful to those growing contraband IMO.

We can also increase yield and desirable traits by providing the right wave lengths at the right point of the plants lifecycle, we can make the basil more intense, the lettuce bushier and more red. And we can control the environment to easier prevent vermin without pesticides, bad weather, etc. We also do large scale installations in the desert under ground where we previously couldn’t farm at all.

I was an engineer at a company called Heliospectra, we specialized in this and I personally built systems that on an industrial scale enabled above light/environment control down to umol of individual wavelengths every minute based on variables such as sensor feedback, algorithms for specific types of plants and traits as well as learning to adapt based on the environmental daily patterns from sensor feedback over time. It was actually quite fun to work on, I do still keep in contact with the company.