Forgot your password?
typodupeerror
Earth Power

Recycling Excess Heat From the Data Center 121

Posted by kdawson
from the talkin'-'bout-my-generation dept.
itwbennett writes "A new data center being built in Helsinki, scheduled to go live at the end of January, will generate energy and deliver hot water for the city. The data center is located in an old bomb shelter and is connected to the Helsinki public energy company's district heating system, which works by pumping boiling water through a system of pipes to households in Helsinki. The recycled heat from the data center could add about 1 percent to the total energy generated by the energy company's system in the summer." The article doesn't say what the overall efficiency of the heat recovery is. Researchers at MIT are working on a new energy-conversion technology based on quantum dots that they say has already demonstrated 40% of the Carnot efficiency limit — 4 times what is achieved by current commercial thermoelectric devices. The researchers believe they can reach 90% of the Carnot limit.
This discussion has been archived. No new comments can be posted.

Recycling Excess Heat From the Data Center

Comments Filter:
  • by sjames (1099) on Tuesday December 01, 2009 @05:28PM (#30289152) Homepage

    It depends on the priorities and the source of the heat. Steam turbines have a considerable load of waste heat that has to be dumped somewhere. No sense heating a lake if it can heat homes instead. Since it is waste heat, the efficiency hardly matters, any usefulness beats the alternative.

    In other cases the choice is between many small boilers or one big one with exhaust scrubbers and more complete combustion. With good pipe insulation that can be a win.

  • by MillionthMonkey (240664) on Tuesday December 01, 2009 @06:06PM (#30289684)
    We've both made errors. I made a reference to "Swedes" and not Finns because I read too fast. However, you said:

    Carnot cycles are for calculating heat engines. This is not a heat engine. There is NO WORK being done.

    First of all theoretical efficiency limits apply to both types of reversible heat engines- steam engines which extract work and heat pumps (e.g. air conditioners) which require it. (In the heat pump case I guess Th would still be a temperature in the datacenter somewhere, but Tc would now be their hot water heaters' temperatures, not the temperature outside- another mistake I made. These Finns are confusing.) But RTFA- this looks like a heat engine:

    Helsinki public energy company Helsingin Energia will recycle heat from a new data center to help generate energy and deliver hot water for the Finnish capital city, it said on Monday.
    The recycled heat from the data center, being built by IT and telecom services company Academica, could add about 1 percent to the total energy generated by Helsingin Energia's system in the summer, according to Juha Sipilä, project manager at Helsingin Energia.

    You DO have more to stand on here if the water were flowing (say) into some hot water heater before being heated to a desired temperature, so the heater would has less work to do- but what's this then in your post about "transmission lines"? It's clear that work is being done here if they're generating electricity in the summer.

    I put electricity into a chip. The chip uses X amount of energy and ~100% of the (electrical) energy is converted to heat energy.

    Yes that's conservation of energy. But notice that the energy has been converted into a less useful form, even if the flow goes directly off that chip into someone's bathwater. There isn't enough water being heated by that chip, or if there is, it's too cold for anyone to notice. Finns want water to come out of their hot water faucets at temperatures greater than if it were just pumped through datacenters.

    If I run water over the chips (lets assume fully submerged for simplicity) and maintain their temperature at 20 degC. ~100% of the heat is removed by the water.

    Whoa whoa slow down there, Tex. You're still thinking in terms of conservation of energy. But there are entropy losses- how will you maintain the temperature at 20 deg C unless you're pumping loads of water across the chip- so much that the heat is dissipated across a volume of water too large to make any difference? Or if you're not- say you pump a low flow across the chip so it maintains a temperature at 30 deg C. A heat engine running on that potential difference (back to generating electricity for transmitting over these transmission lines) you won't be able to extract all the electrical energy that you needed to run the chip. Most of it will heat the great outdoors. But that's not what's going on- either in the article, or an imagined situation involving generation of electricity to go across power lines.

    I realize there are line losses in the electrical wires. There are heat losses from the pipes piping away the heat. There are also radiation losses in the form of light and other waves that transmit through the water.

    Yes yes I realize that too, but we're both waving that stuff away.

  • by MillionthMonkey (240664) on Tuesday December 01, 2009 @06:29PM (#30290032)
    Yeah I read too fast. Although when I was a kid I used to have a Finnish pen pal in Turku. She stopped writing.
  • by Doc Ruby (173196) on Tuesday December 01, 2009 @06:48PM (#30290274) Homepage Journal

    Common commercial devices for homes already use heat exchangers to recover 66% or more of heat from vented air, heating the incoming fresh air with it. During heating seasons, machinery's inefficiency generating heat can replace heat that would have consumed more energy. Bathroom fan vents cost under $300.

    What we need is good heat storage devices. If a lot of heat can be stored during the cooling season, and released during the heating season, these electrical devices become close to 100% efficient. Places like Helsinki have much longer heating than cooling seasons, so they're good places for datacenters that can recover heat for use.

    The problem is that water is about the densest heat storage material we have, but it doesn't store very much. And even the most cutting edge insulators, aerogels, are only about 2x as insulating as the current common top performers, closed cell foams, and only about 4x as insulating as the earlier common stuff like fiberglass and cellulose. If we could store in similar volume the energy that fuels like oil store in chemical covalent bonds instead in physical materials like high specific heat fluids that don't get that hot, we'd have a lot more options in engineering efficiency. If we could regenerate chemical fuel from heat at very high cycle efficiency, we'd have something of a miracle cure for many of the worst of our industrial ills.

FORTH IF HONK THEN

Working...