Indulge me in a bit of science fiction for a moment, for a very simple product that could revolutionize all lighting. Today, if you wanted eco-friendly illumination, you would have solar panels generate power during the day to run your T8 fluorescent bulbs at night. But what if you could just store daylight itself and save it 'till later?
All lighting today ultimately uses power from the sun, whether it be direct solar power, wind power, or fossil fuels. An excellent fluorescent light (one giving 100 lumens per watt) converts electricity to light at about 15% efficiency . If it's powered by a solar panel that is 15% efficient (which is on the high end of affordable these days), you get a total efficiency of 2.25%. That's not even including the losses in the electrical system (converting from DC to AC, and storing in a battery or sending power to the grid and getting other power back). Terrible, right? That performance should be easy to beat.
Light in a Box
What if we could avoid the losses we get in converting sunlight to electricity and back into light? Can't we just store daylight in a box or an optical fiber and release it later? Unfortunately, no, not for hours on end. Every time light bounces off a surface you lose a bit of it, and since it travels at about one foot per nanosecond, it will bounce around a lot in any box you can reasonably fit on a continent. You can get 100% reflection in a prism, but then the material of the prism absorbs a little of the light as it passes through. Likewise with optical fiber, every foot of material that the light passes through absorbs some of the light, and again, the distance it would travel in several hours is big. (It only takes eight minutes for light to get to the Earth from the sun.)
You might think the new 'metamaterials' with a negative index of refraction could be the solution, but I doubt it; they are very lossy and dispersive. Likewise, photonic crystals are new and hip but unlikely to help. However, optical amplifiers have been around for decades, and are widely used in lasers and fiber optic networks. The ones that exist now only work on narrow bands of wavelengths (single colors) rather than a broad spectrum of visible light, and they require a fair amount of power, but perhaps these problems can be solved. Color should be easily dealt with: white-light LEDs have been made by combining red, green, and blue LED's; the same trick could make RGB optical amplifiers. As for power, significant gains need to be made, but if you can use less than you would in converting light to electricity and back (at 2% efficiency, remember), you're still ahead of the game. Even with perfect optical amplifiers, you still need your fiber loop or light box to have the least losses you can get--otherwise the light that gets absorbed will become massive amounts of heat that you need to deal with. But it could be used for cogeneration of heat for air and water.
You might point out that even if you did have a box holding a day's worth of light, you would need a system that could release it a tiny bit at a time instead of all at once as one blinding flash. This is fairly plausible, though--you just need a good switch that can turn on and off extremely fast, and the telecom industry have already been working on switches like this for years, as have optical computing researchers. Again, they require power, but with further development might be able to create an adequately efficient system.
There are rocks which will absorb light and re-emit it later: that's what phosphorescence is. How would you like to have the sun shining through windows onto your walls during the day, and have the walls keep shining through the night? Or maybe you'd rather install a window shade that cuts the brightness of the sun during the day and brightens your window at night? Yes, we're ultimately talking about glow-in-the-dark paint here, but not like the ones that exist today. Efficiency is already competitive--the best minerals re-emit almost 10% of the input light. The main problem is that currently, "long lifetimes" in phosphorescent materials are measured in seconds; we need ones that last hours. Then there's the question of color. Some people already don't install compact fluorescents because their light isn't "warm" enough; the best glow in the dark paint today provides colors that are Vincent-Price-like at best. Clearly a great deal of material science clearly needs to be done.
The semiconductor industry has done amazing things with doped silicon, getting it to hold high-energy electron states like you would need for long-lasting glow. The physicists among you might say "wait, storing light energy for that long is like a battery storing electricity for that long." But it's not. An electrical battery must both conduct charge into and out of itself easily, and store large amounts of charge without letting it leak out easily. A phosphorescent material uses incoming photons to kick electrons into higher energy states, but they stay in the atoms they started in. Granted, it has difficulties of its own, but there are many other marvels of material science today that we would not have thought possible even ten years ago.
So all you optical engineers, physicists and materials scientists out there, you've got some product development to do! It might be a wild goose chase, but it might be a breakthrough.
This proposal raises a problem somewhat similar to Maxwell's Demon. Can humans, through engineering and creation of very effective machines, get things done without spending energy. True, 2.25 percent is pretty inefficient, so there's at least 97.75 percent of the energy usually lost with which we could fuel a light capturing machine.
I just wonder, from more scientifically-minded readers, is this something that theoretical physics says shouldn't work? What I mean is, would this innovation prove a theory wrong, or is it truly new ground?
What of these techniques that have managed to slow and even stop light?
My first thought was the poignant "Light of Other Days" by Bob Shaw...
creative idea, but at least need tens of years to make it happen, even more.
your idea illuminates my another idea: in modern city, almost all the offices or working areas turn on the bulb in the daytime during the working day. so I am wondering if we can guide outside sun light into the inside of the builing, we don't need to use bulb any more and we can directly use outside sunlight. therefor, it will save huge electric power. is anyone doing such research work? I am interested in this.
How about bioluminesence? Bacteria that flourish in the sun by day, and luminesce at night. Phil - thanks for posting that link - what a great story.
Well, they've recently made some breakthroughs understanding the process of photosynthesis which is nearly 100% in it's energy conversion. Perhaps researchers will apply some of those findings...
What you are referring to is a branch of architecture and lighting design called Daylighting, and there are whole companies devoted to controlling natural light, Lutron and MechoShade for example.
Most of the current office building stock is from the 1970's and 80's, uses glass that blocks the entrance of most sunlight due to heat and glare problems, and have floor plans that are square and blocky in proportion. Unfortunately buildings must be designed to use the sun sensibly from the beginning: with regards to overall depth of floor plate and compass orientation on their site (north faces of buildings get no sunlight in the Northern Hemisphere). Direct Sunlight will only travel about 20' max. usably to the inside from the skin of a building on average (less as you move closer to the equator), double that for both sides and you get a building only 40' wide, which is quite narrow in practice. These issues would make retrofitting existing building difficult.
Some countries, especially Germany, have begun legislating standards for maximum distance a desk can be from a window in effect mandating long narrow buildings; interestingly as part of worker's health standards rather than an ecological effort. In the US there is a market-based approach known as LEED for new buildings that pushes the same goals.
However neither of these efforts do anything for existing buildings. This is a looming challenge for architects in the next few decades as buildings from the last few decades reach their mid-life refit periods.
thanks very much for your detail information.
but my idea is a little different with those companies you mentioned. I want to design a fiber-like tunnel which can total reflect light, then the light can be totally guided into a bulb-like container in the house. if we can integrate as many tunnels as possible into our house, the light guided into the container should strong enough to be a lamp.
In response to the first comment by Graham, no, Jer's idea would not prove any existing theory wrong or violate the laws of thermodynamics (as Maxwell's Demon apparently does, which just means you have to use energy to power the demon). Light capture would violate the first law, conservation of energy, only if it produced more light than it absorbed from the sun, and it would violate the second law only if it produced exactly as much light as it absorbed, taking no additional energy input to do so. One need not make either of these claims to beat the 2.25% efficiency of solar electricity. (Trust me, when it comes to physics, Jer knows what he's talking about.)
There is no new physics involved in Jer's idea, only the engineering challenge of making a device or material which is more a efficient means of storing sunlight than solar electricity, perhaps by storing the light in raw form or by converting it to chemical energy and then back.
I wonder what the efficiency is for a hand-dipped beeswax candle? I bet it beats solar electricity by a wide margin, provided you take into account the other energy-storing activities of flowering plants, bees, and beekeepers.
Oops :-) turns out Jer has already answered my question elsewhere. Turns out fire isn't a very efficient or practical light source. I doubt beeswax is as bad as kerosene, and since it involves no fossil fuels it has no greenhouse gas problem, but this article seems to indicate that candles are horribly inefficient in general.
Thanks for the elaboration.
I think you are talking about a Solatube; Velux also makes a device like that. The tubular skylights are limited to low-rise buildings due to the diameter of the "ducts" and the reduction of light due to bouncing inside them, but Big-box stores are starting to catch on, Wal-mart is using them in their "Eco-store" prototypes (Dallas and Denver).
Another variation probably closer to what you had in mind is replacing the coated duct on a tubular skylight with fiber optics that can have a run into the hundreds of feet. I saw some prototypes in the mid-nineties, but the telecom boom at that time sent fiber prices to the moon.
The other drawback is light off the end of a fiber is directly proportional to the diameter of the fiber. To get any light that you could read a book from (which is how lighting fixtures are rated) you need a pretty thick ($$$) fiber, and the attendant space ($$$) to run that fiber. This is why we have mainly seen fiber optics limited to decorative applications in buildings.
great! you are so informed. that's what I want to know. thanks
I have been thinking the same thing for many years and have even mixed up some phosphorescence materials of different colors to get a better color index. But the problem remains that the chemicals do not store enough concentrated light to be useful and they fade quickly.
But I am still workin on it!
Brilliant, and I too remembered the old sci fi story about "slow glass" - glass windows were placed out in a field for 10 years to absorb light - and light traveled so slowly through the glass that you could then put the window in your home and get that 10 years of visions through the "window"
As soon as I saw above, yes: it was The Light of Other Days...