Chip to convert excess heat into energy?
- Thread starter London Geezer
- Start date
Well, they might use an on-chip cascade, but that in itself would be quite an accomplishment on many levels.
Wow! This is fantastic! Finally the heat from our CPUs can power our GPUs! 
I read some of the Slashdot comments about this article and it cracks me up how many people make uneducated remarks. This is not a perpetual motion machine and I don't believe the developers are claiming such. I've been casually following this type of technology for a while now and I've been waiting for someone to get it right. I agree with hupfinsgack that Eneco is taking advantage of the Seebeck effect. Their improvement seems to be in efficiency and their "solving" the challenge of maintaining a heat differential. Power Chips claims to have even better efficiency than Eneco.
The Seebeck effect devices I've read about perform better when there is a large temperature differential between the hot and cold sides of the device. As we all know it's difficult to keep something cold if there is a heat element near it so I'm curious to see how Eneco's device performs in practice.
The Seebeck effect devices I've read about perform better when there is a large temperature differential between the hot and cold sides of the device. As we all know it's difficult to keep something cold if there is a heat element near it so I'm curious to see how Eneco's device performs in practice.
I had never heard of this tech before stumbling on this article (thanks to gmail and its links... err...).
If something like this really does materialise, even in 10 or 20 years, it could mean that the huge headache we're gonna have to energy in a few years time could vanish. No more waste of energy would solve many problems we will have, such as oil reserves drying up and even pollution.
Hopefully someone will get it right...
If something like this really does materialise, even in 10 or 20 years, it could mean that the huge headache we're gonna have to energy in a few years time could vanish. No more waste of energy would solve many problems we will have, such as oil reserves drying up and even pollution.
Hopefully someone will get it right...
That won't happen. The heat-energy conversion is not the point of this. Converting heat into energy via Seebeck is a bad idea, especially on the large scale of power plants etc. Moreover, the significance of this technology lies somewhere else. If that thing really works it makes for a more energy efficient cooling solution that is completely noiseless. So this would ideal for Laptops, etc. Using this to generate energy would mean that you're not cooling your chip (remember this depends on the temperature differential between hot chip and cold surrounding). Not a good idea.
I think turning heat into energy is most useful as a way to make other forms of energy generation or use more efficient. It will also be useful for generating power in situations that aren't practical with other methods. Initial approaches are starting small.
Thermo Life has a small generator that can be used to provide power in situations where replacing batteries can be difficult.
http://www.poweredbythermolife.com/
Thermo Life has a small generator that can be used to provide power in situations where replacing batteries can be difficult.
http://www.poweredbythermolife.com/
But... but... but....
I thought you could just put these things in your kitchen and they'd produce energy from all the heat when i'm cooking pasta... and stuff...
I mean, i could put some in my bedroom, temperatures reach alarming levels in here at times!
Some formulas:
Th = Absolute temperature at hot side
Tc = Absolute temperature at cold side
Qh = Thermal energy difference at hot side
Qc = Thermal energy difference at cold side
W = Electrical energy
n = efficiency
Theoretical max efficiency n of converting thermal energy to electrical form. Qh is taken from hot side , Qc is released at cold side, W is the electrical power you get out.
n = W/Qh
n = 1-Tc/Th
If you want to give temperatures in ºC, then the formula is
n = 1-(tc+273.15)/(th+273.15)
But all practical examples are usually a lot worse than this maximum.
Try a couple of hot and cold temperatures in that formula, and you'll see that the efficiency is usually quite low. And Qh is the energy that is transfered into the device, any heat energy that sneaks by it by other means (probably a lot in most cases) will further reduce the efficiency.
Extracting electrical energy from body heat will probably give an efficiency of a couple percent. And unless you've got a full body suite to pick up the energy, you'll only pass a small amount of your produced energy through the device. Ie, you could probably power a wrist watch, but not much more.
If you want to save energy when you're cooking your pasta, it's a lot more efficient to simply not generate so much heat. With a pot that is thermally insulated on all sides (except the bottom :smile, you could turn down the power a lot and save much more energy than you could ever reclaim. I know that cooking pasta with the lid on can be tricky, with high risk of boiling over or the pasta sticking to the bottom. But that's a different story. :smile:
If you want a good efficiency, you'd need some real temp differences. To get the claimed 30% efficiency, they probably need the 600ºC they said it could handle. But if they actually can get that kind of efficiency, and build a device where most of the heat from a fire pass through the device (ie, the exhaust is close to ambient temperature), then it would be in the same range as the efficiency of a good combustion engine. That could actually be useful.
The followup question is, I wonder how much power those chips can handle, and what they cost?
If they can be made reasonabely cheap, they could be useful as a portable electrical generator.
Hmm.. what temperatures can you get from a camp fire? Sitting by the camp fire in the middle of the woods, and powering the laptop from the fire could maybe add a dimension to playing Doom3 (or your favourite scary game).
Th = Absolute temperature at hot side
Tc = Absolute temperature at cold side
Qh = Thermal energy difference at hot side
Qc = Thermal energy difference at cold side
W = Electrical energy
n = efficiency
Theoretical max efficiency n of converting thermal energy to electrical form. Qh is taken from hot side , Qc is released at cold side, W is the electrical power you get out.
n = W/Qh
n = 1-Tc/Th
If you want to give temperatures in ºC, then the formula is
n = 1-(tc+273.15)/(th+273.15)
But all practical examples are usually a lot worse than this maximum.
Try a couple of hot and cold temperatures in that formula, and you'll see that the efficiency is usually quite low. And Qh is the energy that is transfered into the device, any heat energy that sneaks by it by other means (probably a lot in most cases) will further reduce the efficiency.
Extracting electrical energy from body heat will probably give an efficiency of a couple percent. And unless you've got a full body suite to pick up the energy, you'll only pass a small amount of your produced energy through the device. Ie, you could probably power a wrist watch, but not much more.
If you want to save energy when you're cooking your pasta, it's a lot more efficient to simply not generate so much heat. With a pot that is thermally insulated on all sides (except the bottom :smile
, you could turn down the power a lot and save much more energy than you could ever reclaim. I know that cooking pasta with the lid on can be tricky, with high risk of boiling over or the pasta sticking to the bottom. But that's a different story. :smile:If you want a good efficiency, you'd need some real temp differences. To get the claimed 30% efficiency, they probably need the 600ºC they said it could handle. But if they actually can get that kind of efficiency, and build a device where most of the heat from a fire pass through the device (ie, the exhaust is close to ambient temperature), then it would be in the same range as the efficiency of a good combustion engine. That could actually be useful.
The followup question is, I wonder how much power those chips can handle, and what they cost?
If they can be made reasonabely cheap, they could be useful as a portable electrical generator.
Hmm.. what temperatures can you get from a camp fire? Sitting by the camp fire in the middle of the woods, and powering the laptop from the fire could maybe add a dimension to playing Doom3 (or your favourite scary game).
Here's an interesting article from New Scientist magazine that I read a couple years ago. You need a subscription to read the entire article, but here's the summary. I don't remember if the article mentions any specific companies. It is ironic that the article mentions the technology has now arrived yet 2 years later there are still no products on the market.
http://www.newscientist.com/article/mg18324635.100-heat-wave.html
... u sure u're german? That was totally brit
I'm not german (nor brit) but I prefer english English
Mintmaster
Veteran
You sure? Even 30% of perfect conversion is not that much. For a 50 deg. CPU in 25 deg. ambient air, 30% of Carnot is something like 2.3% of heat transfer going to electrical energy. I don't see that making much difference, and that's under the assumption that the back side is room temp. What exactly did you have pictured?
The application to cooling seems to be the -200 stuff they're talking about, which actually makes things far less efficient. A 100W CPU can be cooled to -200 deg if you're willing to pump 100W*(323/98 - 1) = 342W of electrical energy into it, as long as you can also exhaust all 442W out at 50 deg. And that's assuming 100% efficient refridgeration.
Nope, extreme refridgeration ain't pretty.