Waste Heat to Electricity?

Darwin_Frog writes: "Recent advances in thermionics at MIT lets waste heat generate electricity, thus pushing entropy one step further down the chain. These devices work at a temperature around 250 deg. C, instead of around 1000, so cars can augment the alternator by using the waste heat in the exhaust system to produce power for onboard electronics and A/C."

Comments

Matrix style

[King of Caffiene]

soon they'll be able to use excess heat from humans...matrix style.

Hmmm...

[caseydk]

I think it might make the EPA happy if companies had these in their smokestacks... maybe reduce their power draw a bit...

less power required= less pollution

Re:Hmmm...

[wass]

I used to work at an MIT laboratory that was sponsored with DARPA funding. I left 2 years ago to go back to school to get my PhD in physics. I'm not sure of the exact details, but here's the basic scoop as far as I see it.

DARPA essentially funds research laboratories to perform research projects that will further advance technology related to DARPA interests. In my case, the research was unclassified, and our group was able to colloborate with other groups and colleages, present our research at conferences, as well as publish our methods/systems/data in scientific journals.

The laboratories that DARPA funds are either university laboratories, FFRDCs (Federally-Funded Research and Development Centers), and commercial laboratories (ie, IBM or Motorola research labs, for instance). It is usually standard practice for employees of all the above labs, upon the beginning of employment, to sign contracts handing over patent rights to the employer (ie, the FFRDC or the company). Actually, I'm not sure about students, as I haven't signed any patent forms yet. But did when I was an employee of MIT. So did Richard Feynman when he worked for Los Alamos (FFRDC).

So, essentially, DARPA has certain technological goals it wants to achieve, and funds a variety of sources to help achieve them. Usually for each specific project, DARPA funds a variety of research labs, and has them compete for further funding. The research labs in turn present their results at least annually for funding renewal. Eventually, DARPA gets it's results (or lack of them), and gets what it needs in terms of advanced technology, and then cna use that technology within more advanced systems.

I do not know specifically what kind of strings come attached with DARPA funding. However, I would imagine that most likely the research labs themselves get some significant percentage of patent rights as a bonus for conducting DARPA research. Otherwise there is no incentive for, say, Boeing to research a new type of stealth aerofoil if DARPA holds on to all patent rights. I know my boss at MIT had his share of patents, but of course, MIT essentially owns said patents.

Note that DARPA's ultimate purpose is to get better technology into Defense-related projects. They advocate using COTS (Commercial Off-The-Shelf) hardware/devices whenever possible. That is, don't waste $$$ designing your own op-amp if Analog-Devices has one that's within your specifications. Of course, you must roll your own if the COTS op-amps don't meet your bandwidth/linearity/bias/power/etc requirements. So, DARPA doesn't care about who gets the patent rights for that op-amp, they want the research that makes use the op-amp. So, in this example, your tax dollars are already going to Analog Devices and helping their own patent processes.

Your concerns about tax dollars funding university patents are either too narrow or too broad. Realize DARPA funds commercial entities as well as FFRDS too, which have similar patent processes. However, DARPA's fundamental purpose is to fund advanced research projects to further American defense interests. That's what it does, and it will support commercial, government, or university research labs to achieve this goal. It's a government agency, so obviously it is funded with tax dollars. I don't think DARPA cares about patents, as long as it can utilize the fruits of the research.

Introducing...

[Iamthefallen]

Introducing Athlon XP 5000 - Now self powered!

Re:Introducing...

[mother_superius]

In this universe, we obey the laws of thermodynamics!

Nice but not the end of entropy

[SysKoll]

According to the article, this "breakthrough" is a reverse Peltier junction with about twice the efficiency of current semiconductor thermoconverters. Nice, but nothing revolutionary.

I think it's quite excessive to claim this will reduce entropy. Although I agree that if it's economically deployed in, say, cars, it will supplement the alternator.

Could this new junction actually replace the alternator for producing electricity in a car? Let's see: assume a car has a 100 HP internal combustion engine. That's 75 kW. Two third of this is wasted in heat. Typically, the radiator gets about half of this heat (the other half is dissipated away in radiant heat or through the exhaust. Assume further that 20 percent of this can be recovered and converted to electricity (for a really efficient semicon pile). That's 75 * 2/3 * 0.50 * 0.20, or 5 kW. That's more than a good SUV alternator. So this could actually work, provided it's reliable and not too expensive.

You'll need a battery for the short runs, though.

--SysKoll

Re:Nice but not the end of entropy

[GMwrench]

I don't think so. First your 100 HP engine will only produce 25-35 HP most of the time. Peak power is only produced during hard accerlation during cruse it's much lower and at iddle almost nonexistant. This is 99% of the time. Also an alternator only produses 1-1.5 KW. And the battery cannot be replaced it's needed to start the engine and supply power at low speed when your charging device is insufficent.

Re:Nice but not the end of entropy

[ZxCv]

Even at 25-35 HP, according to his math, that still makes 1.6-2.3 KW. More than an alternator, according to you.

Also, there was no mention of replacing the battery. In fact, I believe it was: You'll need a battery for the short runs, though.

Maybe read the post a little harder next time before responding in a such a know-it-all tone?

Use on Hybrid cars?

[BlueJay465]

My question is how much more gas mileage could this technology squeeze forth given an array of these attached to the heat producers of a vehicle, like the engine or the brake pads.

Another thing is how do these "thermal diodes" compare to a Peltier Element in heat conversion to electricity?

Anyonw know how much they cost.

[argoff]

It'd be great if we could use this for cheap solar cells. Regular solar cells are pretty expensive. (I'm almost convinced that other industries are screwing with the market to make them cost so much). Anyhow, does anyone know how much this new stuff would cost? PS: nuclear's my favorite, but it's too easy for the govt to regulate.

thermodynamics, and entropy, and all that

[StandardDeviant]

here's the layman's formulation of the things that give chemistry students the cold sweats, the rules of the game as it were:

1. You can't win.
2. You can't break even.
3. You have no choice about playing.

Any closed system ends up in the state of most disorder, and all systems are closed if you look at the boundaries carefully. No matter how hard you try, no matter what ingenous things you do, in the end, the dealer wins and everything is dust. Cold dust, at that. The more energy you expend enforcing order, the more chaos you cause. There are no wins in technology, only a prolonging of the inevitable loss. So while I'm sure this new doohickey is neat, somewhere, Carnot is laughing and his cycle is tapping you on the shoulder snickering to itself.

Re:thermodynamics, and entropy, and all that

[dragons_flight]

I am a physicist and have studied entropy, though it is not my specialty.

At a fundemental level, entropy is a measure of the number of accesible states of a system for a given energy distribution. Presumably you know that temperature is really just a statistical measure of average kinetic energy in a substance. In the simple case of a uniform temperature gas, it's possible to compute the entropy directly, by (a process analogous to) counting the possible ways to arrange the molecules and distribute their kinetic energy such that you still have the same temperature. (Okay it's not really counting cause there is [usually] a continuum of positions and energy values, but the idea is there, only with more integrals.)

Roughly speaking a system is "ordered" or "disordered" based on how much freedom it has in distributing the energy in it's heat. For instance, in highly complicated and stable configurations (e.g. DNA) you can infer that the heat gets distributed only in ways that don't break down the basic structure. Of course with enough heat it will no longer be stable, but that's a different case.

While the number of accesible internal configurations for the heat energy is the basis for entropy, very few people actually use this. What is actually used is a set of laws mathematically derived from this which can be directly applied to macroscopicly measurable quantities. Chemists know more about these areas than I do, but I'll cover a few of the basics.

The most important is known as the Second Law of Thermodynamics, stated simply "Entropy always increases (or stays the same)." Whenever you do anything that moves energy (such as heat) around, the net entropy will increase (except in those rare cases when it stays the same). It is possible to locally decrease the entropy of one system, but you are guaranteed to increase the entropy of everything else by at least the difference.

There is another important trick about entropy. It tells you that it's impossible to transfer energy from heat to any other form with 100% efficiency. Not only that but you can't even do it with arbitrarily close of 100% efficiency unless you have something who's initial temprature is arbitrarily close to 0 degrees Kelvin. Heat engines, any device that changes heat into other forms of energy, depend on having a difference in temperatures available (for instance, cool river water versus hot steam pipe). If you just have a box sitting at room temperature, it can't work.

There is an interesting caveat here. The Second "Law" and most of how we typically apply entropy are based upon something called the Fundemental Assumption of Thermodynamics. Roughly stated: "All possible energy configurations are equally likely". As it turns out this is rarely ever exactly true, but it is so nearly true in almost every concievable macroscopic situation that it makes no difference. Entropy always increases is a mathematically certain law derived from the fundemental assumption and mathematical definitions of temperature, etc, but it is still concievable that their might be systems where the fundemental assumption doesn't apply and entropy might decrease. Over the years there have been a few suggestions for how to build such a thing (mostly at a quantum mechanical level), but no one has ever succeeded.

If someone does build a box that sits on a desk and converts ambient heat into energy output, then they are almost certainly guaranteed a Nobel prize. On the other hand there may be something better than the fundemental assumption, which is exactly true and excludes all possibility of such a wonderful, energy giving black box.

Hmmm...

[jimhill]

I couldn't help noticing that within a few paragraphs the writeup mentioned that (1) the research was partly sponsored by DARPA and (2) patents have been applied for with one already issued. Color me bitter, but as one of the taxpayers who funded the research I can't say I'm overjoyed at the prospect of paying licensing fees to MIT through the eventual commercial implementors.

I'm all in favor of government-sponsored research. They have the resources to investigate stuff with great benefits but staggering R&D costs. I'm all in favor of universities conducting the sponsored research. Grad students are cheap (I know, I was one for many years) and the brainpower is not less than one finds in industry. However, when the government pays a university to do something new, the university's benefits should be the equipment bought for the research and the prestige that comes from doing it first/best/cheapest.

is it more efficient than turbines?

[Pyromage]

this truly is the fundamental question: can this be made to be more efficient than a turbine/generator combo?

If this can be more efficient than a turbine, we can have solid-state power plants. Nukes are nothing more than a complex method of boiling water to push a turbine: if we can replace the water, we have an order of magnitude less waste! Not to mention that the core stuff is much easier to deal with than heavy water. Plus, with no pumps or pipes to break, it becomes even safer than it already is.

Or other things, say laptops? PDAs? Naturally all these kinds of applications are XYZ years off, but just imagine what would happen when we get the effiency of these things up? I'd bet that boiling water to turn a turbine is real low efficiency: if we cut out the turbine step alone, that should increase effiency by a whole lot.

This is truly cool shit.

Re:is it more efficient than turbines?

[leucadiadude]

You are confusing reactor waste with waste heat.

The waste comes from the approximately 65% of the original heat pumped into the primary circuit being lost to the river. You have to condense the steam coming out of the turbine so you can pump it. It takes a *lot* of energy to condense this steam back to water. You may not be raising a particular gallon of river(or ocean) water by more than a few degrees (usually less than 5-8F) but you are moving a whole pisspot full of cooling water through your condenser. So the total energy rejected to the environment is quite large. Real world example, the plant where I work is 34.2% efficient, which is actually pretty good for a large steam cycle power plant. The reactor core pumps about 3400MW of heat into the primary circuit and we get about 1175MW of electricity out of the turbine generator, the vast majority of the rest (2225MW) is transferred to the 1,000,000 GPM of ocean water used to cool that pesky steam back into water so it can be pumped.

Now if you could design an economical steam pump (or better yet a two phase pump - steam in and water at higher pressure out) your billions of $'s would be waiting for you. You would be able to knock the stuffing out of the Rankine Cycle and increase plant efficiency into the 50-60% range overnight.

Not enough information yet

[Animats]

Not too much info yet. In particular, there's no indication of how much such devices will cost per watt. This is a basic problem with things like Peltier-effect devices and solar cells; they work fine, but you need an awful lot of them to get serious power levels. If this requires something like a wafer fab to make, it will be a niche device for years to come.

Cold Fusion Redux

[Baldrson]

Peter L. Hagelstein was the guy at MIT who had MIT's lawyers churning out cold fusion patents like there was no tomorrow at the same time that MIT's official position was that cold fusion was an illusion -- and making official recommendations against its funding.

Re:Hmmm... - might ruin smokestack effect

[victim]

Sapping heat from the smokestack contents will probably cause it to not work correctly.

The goal of a smokestack is to get the harmful exhaust away from the ground long enough that it disperses sufficiently before touching down.

This is done with convection. The hot gas in the tall stack creates the draw that powers it and blows the plume up after it leaves the stack, the hot plume continues to lift itself until it bleeds off too much heat, then it starts coming back down, but presumably dispersed enough to not be too noxious.

The smoke stack was designed with a known gas temperature and dispersal requirement and a desire to minimize masonry. If you take away heat from the gas you will reduce your plume altitude and cause it to come down in a more concentrated region.

I doubt you can use the thermo-generated electricty to run blowers to compensate. The `no free lunch' law of thermodynamics will probably forbid that. (Unless blowers are much more efficient than convection.)

Now, if you are just bleeding off waste steam then it would work, but most of the energy in steam is the expansion from water to steam, there is relatively little left in the puffy clouds.

Mostly unrelated note: I used to live in Pittsburg in a community where all the houses were required to have slate roofs, stone or brick exteriors and no wood trim. Even the window frames were metal. It was a fire-proof community from the days when the steel mills spewed lots of solids including hot cinders. The plume was powerful enough to carry those large distances fast enough that they were still hot enough to start a fire.

Thermodynamic efficiency limits

[Animats]

The law of thermodynamics that's relevant here is that the maximum efficiency of any heat engine is

• (T1 - T2)/T2

where T1 is the temperature at the hot side, and T2 is the temperature at the cold side. Both of these temperatures are measured from absolute zero.

This is why extracting energy from something that's just a little warmer than its environment is very inefficient. With the hot side at 100C and the cold side at 20C, you're limited to about 20% efficiency in theory, and will be lucky to get half that. Power plants generate steam at upwards of 600C, not just above the boiling point, for exactly this reason. Gas turbines run even hotter. Solar plants for power production typically focus enough energy on a target to reach the 600C level, as Solar Two in Mojave does.

You just can't extract much power from things that are merely warm. They have to be really hot.

Re:More info

[beable]

Dude, it doesn't "reclaim heat". It uses a generator as a brake, thus avoiding using brake pads to convert kinetic energy into heat. From the link you posted:

* When decelerating or braking, the electric motor turns into a generator to charge the batteries automatically. It's a unique hybrid feature called regenerative braking. Normally when you brake, all that energy is converted into heat into the brakes. Toyota's Prius actually recaptures about 30 percent of that energy to recharge the nickel-medal-hydride batteries in the back.

To stop the car, it needs to remove kinetic energy from the car. In normal braking, the energy is absorbed by the brakes, which radiate the energy away later. Regenerative braking instead uses a generator to convert the kinetic energy into electricity (and heat), storing it in the car's batteries. Electric trains have been doing this for years.

Re:This would be useless in automotive..

[NevarMore]

I beg to differ. Being an ex-geek, now a car guy, I'd love to use the heat my engine throws off.

If the heat is being converted to electricity then there will be less heat. Lower heat in the exhaust alone means lower engine temperatures because the exhaust sytem radiates the most heat near the engine at the headers (the part where the exhaust comes off of each cylinder for you non-car types). Since thats where the exhaust is hottest thats where the devices would be mounted. A lower exhaust temperature means a lower overall engine temperature.

Secondly, the big step is going from 1000 degrees down to 250 degrees. Taking that 250 down to 180 or 160 would likely allow these devices to draw heat from the engine itself. Having these devices draw energy would reduce the work a typical liquid cooling system needs to do, allowing it to be reduced in size.

Newer cars and performance cars are replacing belt driven components with ones powered electrically, most notably fans and water/coolant pumps. Elimiating belts allows the engine to put more power to the wheels rather than turning an accesory. The catch is that these devices need more power from the battery and alternator. Alternators are presently limited to about 150-200 amps, enough for a stripped race machine to run its accesories, but not enough for a street driven car with lights, music systems, and long continuous driving. These thermocouples would add more electrical power to the system and use more of the energy produced by the combustion.

The automotive example is a bit advanced for the time, but in todays science community a potential commercial use is the best way to get money for new ideas.

Sorry if that went on too long, or was too automotive for you slashdot geeks. ;-)

Thermal diode := Peltier Element

[wowbagger]

A thermal diode IS a Peltier element. This has been covered in EE Times among other trade journals. All they've done is take the standard BiTe diode, which is very thick, and thinned it down by creating the layers with standard chipmaking techniques. So, instead of one diode junction being about 1mm thick, they make a device that is 0.1mm thick consisting of many tens of layers.

You forget, it runs on waste heat

[Spamalamadingdong]

First your 100 HP engine will only produce 25-35 HP most of the time. Peak power is only produced during hard accerlation during cruse it's much lower and at iddle almost nonexistant.

Which doesn't make much difference, because the engine's waste-heat output doesn't change nearly as fast with throttle opening as the crankshaft output does. Even at idle (zero power) you are still burning fuel and still pumping heat out the exhaust pipe. If you can force that waste heat to do some work for you instead of just being diluted to uselessness in the atmosphere, you've accomplished something.

A hybrid vehicle would probably shut down the engine at idle and eliminate that waste-heat stream, so the thermal converter would be more useful as a way to increase the general efficiency level of the powertrain. If you can get an extra 10% off the 40% of the heat which is rejected through the exhaust, that's 4% of your fuel value; added to a 30% engine thermal efficiency, you've gained 13%. That's nothing to sneeze at.