The shortcut to doubling solar efficiency?

November 5, 2010 9:28 PM

Super Soaker inventor proposes a solid state heat engine that resembles a fuel cell. The idea is to reach a new level in heat generating efficiency, or at least replace heat pumps, with no moving parts. "Johnson has opened up a fundamentally new pathway to generate electricity from heat," says Paul Werbos, program director for power, control, and adaptive networks at the U.S. National Science Foundation (NSF). Werbos, an IEEE Fellow, says the NSF is funding Johnson's heat-engine research because of the strong chance that it could cut the cost of solar power in half.
posted by Brian B. (49 comments total)

This is fascinating stuff. By all rights, Lonnie Johnson ought to be a household name for Super Soaker alone. If he can help bring us cheaper, more practical solar power, he could be responsible for a new age of green energy. Great post!
posted by Strange Interlude at 9:45 PM on November 5, 2010 [4 favorites]

This is really incredible. I'd go so far as to say it completely vindicates Super Soaker for making The Oozinator.
posted by mullingitover at 9:47 PM on November 5, 2010 [11 favorites]

This is good news!
posted by flippant at 10:08 PM on November 5, 2010

I wonder if this technology could be used to boost the thermal efficiency of microprocessors.
posted by delmoi at 10:14 PM on November 5, 2010

delmoi: Elaborate?
posted by phrontist at 10:18 PM on November 5, 2010

I'm guessing delmoi meant something like recapturing the heat given off when microprocessors are running to be re-purposed into something useful.

This is fantastic news. Can't wait to track the progress of this thing over the next few years; it seems like he's really close.
posted by titantoppler at 10:25 PM on November 5, 2010

I always wondered who invented the super soaker, because I also "invented" it. Years before they hit the market I was wondering why you couldn't go into a store and buy one - it would be so awesome and was so obvious and it was frustrating no-one was doing it. I was a kid at the time, so I couldn't act on it, all I could do was be frustrated and wonder "why not?".
(For years beforehand, my dad had had a pump-up pressure sprayer, that worked on the principle, and I just knew it would own in any watergun fight, but because it was used for spraying chemicals, I wasn't allowed to play with it. But I understood how the mechanism worked and I wished that someone would take that mechanism and sell it as a water gun. Years later, someone did.)

Reading about the guy, I'm really glad he was the guy that did it. As a kid, I guess I kind of resented knowing that someone made millions out of making pressure sprayers for kids, without even doing it early enough to get a product on the shelves before my childhood ran out, but this guy is totally the guy I want to see reaping those rewards and busy applying himself to even greater things. This is awesome.

Rock on dude. You make the world a better place.


I'm a grown-up too now. So I get to work on developing things too now. :)
posted by -harlequin- at 10:34 PM on November 5, 2010 [4 favorites]

The applications for this thing are staggering. Maritime use, for instance; put half the device in the sun, and half under the surface, and there's a good hefty temperature differential. Floating colonies with ample power anyone?

Space use is an obvious one; put one side towards the sun, the other in the ship's shadow, and you're going to have a massive differential to work with.

I like the way the future is looking right now.

Thanks, Super Soaker guy!
posted by MrVisible at 11:09 PM on November 5, 2010

Sounds like a catalytic converter on either side of a proton conducting membrane, performing opposing functions.
posted by five fresh fish at 11:13 PM on November 5, 2010

I'd like to have a beer with Lonnie Johnson. Fascinating man, fascinating work. I hope he can shepherd this one to production!
posted by Goofyy at 11:18 PM on November 5, 2010

Well this sounds a little better than the usual 'guy in a garage invents technology that will SAVE THE WORLD!". The folks at PARC are a smart bunch at least at science, so if they've really looked at it there is probably something there. Still, I wonder about a company who's web page is accepting paypal donations to help with research. Billions of investment dollars have been thrown at alternative energy ideas over the last few years, many that are far far wilder then this. So either the guy is completely unwilling to give up any ownership in return for development cash (which is an impression I get from the first article) or he's a pretty poor entrepreneur. Given that he's now essentially broke after making a fortune with the super soaker that may well be the case.

He got $850K (of tax payers money) from the military, I would thought that should have been enough to produce a small scale prototype, which is really what you need to raise some big bucks. However it seems he only has computer models and possibly some work on the proton exchange membranes that a key to the idea. Now he's looking for 'strategic partnerships' and more government money to take it further. I really want it to be legit, but my BS detectors are tingling just a little.
posted by Long Way To Go at 11:33 PM on November 5, 2010 [3 favorites]

neat; thanks for posting it.
posted by LobsterMitten at 11:46 PM on November 5, 2010

Now he's looking for 'strategic partnerships' and more government money to take it further. I really want it to be legit, but my BS detectors are tingling just a little.
posted by Long Way To Go at 11:33 PM on November 5 [+] [!]

Do I really need to even say it
posted by dubitable at 12:36 AM on November 6, 2010 [3 favorites]

Thanks for the post, one of my pen-pals sent me a diagram for a better way to plug up the oil leaking hole in the gulf, weeks before the oil people finally closed that hole, my friend does not have access to the internet, so i'm going to send him a copy of the article, then wait to see what he says about the idea.
posted by tustinrick at 1:09 AM on November 6, 2010 [1 favorite]

If only cost was the issue, this would be great news.
posted by falcon at 1:23 AM on November 6, 2010

Wow. Is this as wow as it seems? It seems pretty enormously wow.
posted by From Bklyn at 2:20 AM on November 6, 2010

Long Way To Go: "He got $850K (of tax payers money) from the military, I would thought that should have been enough to produce a small scale prototype"

Not if the basic research on developing the membranes required to actually make the device work is yet to be done. As I understand it, he's at the point where he can say "if we can make a material like *this*, then we can make this awesome device".

The unwillingness to share his invention is probably killing his progress though. I suspect there are far better materials scientists out there who would be able to push forward that research than the people he has access to, but if they can't share in the profits why would they bother putting any of their time into it?
posted by pharm at 2:41 AM on November 6, 2010

I wonder if this technology could be used to boost the thermal efficiency of microprocessors.

How about laptops that give off tons of heat, which funnel that back into a longer battery life?
posted by Blazecock Pileon at 3:21 AM on November 6, 2010

>if they can't share in the profits why would they bother putting any of their time into it?

I know, right? Why would someone contribute, say, 9 posts, 712 comments, and 44 favorites to something without sharing in the profits?
posted by Joseph Gurl at 3:29 AM on November 6, 2010 [6 favorites]

Wow, this is exciting!
posted by Splunge at 3:59 AM on November 6, 2010

Here's a video of a talk he gave at PARC about the JTEC, and a nice animation of the operation of the device.
posted by Jakey at 6:27 AM on November 6, 2010 [2 favorites]

That Werbos guy is a visionary on his own. Check his extraordinary website.
posted by oneironaut at 6:29 AM on November 6, 2010 [1 favorite]

Man, if you replaced your Segway's Stirling engine with one of these, you'd really have something.

Seriously, is the magic required to make a JTEC work the proton exchange membrane? Where can a novice read up on the current state of the art of those membranes? Wikipedia is pretty thin.
posted by Nelson at 7:12 AM on November 6, 2010

It occurs to me that this device could reduce the cost of coal fired electricity by half as well.

This is a very neat hack to get around Carnot limits.
posted by MikeWarot at 7:26 AM on November 6, 2010 [1 favorite]

My bad... it doesn't get around the Carnot limit... but gets you closer to it. The audio levels are way low on the video, but he makes a good case.
posted by MikeWarot at 8:09 AM on November 6, 2010 [1 favorite]

This looks really cool from a mech.eng point of view. The trick is that in his cycle the expansion happens over a very thin membrane, so that it's almost isothermal. This brings the efficiency closer to the maximum achievable - the "ideal Carnot cycle". He claims a good target is "85% of the Carnot efficiency" which is about the same as existing combined cycle generation plants, but much better than simple (Rankine) cycle generating plants. The advantage is that his cycle is much simpler than a gas or steam turbine. It should also scale down better both in terms of size and temperature. That's probably why he's focussing on solar power (that, and the green investment cred). But I'm most interested in seeing a J-tec engine replace the steam turbine in a cheap combined cycle plant.

The devils in the details though - Proton Exchange Membranes are notoriously flakey. He still hasn't found one that conducts efficiently while holding the pressure differential. Many great ideas have withered for lack of a magic material (space elevators, fuel cell cars, molten salt reactors . . . I could go on).
posted by Popular Ethics at 8:09 AM on November 6, 2010 [6 favorites]

about the same as existing combine cycle generation plants

I just realized that my linked article doesn't really explain this clearly. The most efficient power plants in the world are currently 60% efficient. They operate at a hot temperature of about 1500K and a cold temperature of about 300K. The Carnot efficiency is 1-TC/TH = 80%, so these plants are running at 0.6/0.8 = 75% of the Carnot limit, only 10% worse than the J-Tec's best-case scenario.
posted by Popular Ethics at 8:20 AM on November 6, 2010 [1 favorite]

Well, I hope this works.
posted by Xoebe at 8:38 AM on November 6, 2010

This changes everything. Let's make a dyson sphere out of our dinky little solar system, ASAP.
posted by mccarty.tim at 9:28 AM on November 6, 2010

So... how does it work? The articles are mysterious and vague, with phrases like "On the
high-pressure side of the MEA, hydrogen gas is oxidized resulting in the creation of protons and electrons". How? Ionizing hydrogen is not easy -- it takes 14 eV (visible light photons are only around 2 eV). Thermal dissociation needs temperatures in the 5000 K range. It's some secret-sauce ceramics that magically do this?
posted by phliar at 11:37 AM on November 6, 2010

He got $850K (of tax payers money) from the military, I would thought that should have been enough to produce a small scale prototype

The salaries of eight engineers will burn through that in a year. The equipment they need to do the work could burn through that in a week if you're not careful. It sounds like they're basically trying to find a way to build a material with the desired exotic specs, I can see that taking a while :-(
posted by -harlequin- at 12:02 PM on November 6, 2010 [1 favorite]

Space use is an obvious one; put one side towards the sun, the other in the ship's shadow, and you're going to have a massive differential to work with.

Not sure that would be the best use, MrVisible. In space, there's only radiative cooling; that's a lot less efficient than the conductive and convective you can use on Earth.

Still, exciting stuff.
posted by IAmBroom at 12:14 PM on November 6, 2010

In space, there's only radiative cooling

Also, lest we forget, no one can hear you scream.
posted by Naberius at 12:21 PM on November 6, 2010 [3 favorites]

It is a spectacularly interesting idea with fantastic potential. There are still some misconceptions in the reporting though.

1. Steam turbines are not complex and unreliable, they work great.
2. There are moving parts in the new scheme. Not mechanical parts and bearings and such, but the hydrogen does move.

Also, the membrane may turn out to be unreliable or lossy or what have you.

All that doesn't stop this from being very interesting.

Also, I need to learn more Thermodynamics :P
posted by Chuckles at 1:33 PM on November 6, 2010

And it might be better not to think of it as a moving part exactly.. But it is a "wet" process, like a battery or whatever, not a dry process like a semiconductor. That is a really important difference, because semiconductors mostly have near infinite service lives, were a battery may only last a couple of years or a few hundred cycles. That is after the battery technology has been commercialised.
posted by Chuckles at 1:43 PM on November 6, 2010

ut it is a "wet" process, like a battery or whatever, not a dry process like a semiconductor.

Hydrogen is tricky stuff too. It will embrittle or otherwise damage materials it comes into contact with over a long period.
posted by atrazine at 2:42 PM on November 6, 2010

Anyone else noticing the parallel to a mitochondria? Or am I not understanding the process very well?
posted by EtzHadaat at 3:17 PM on November 6, 2010 [1 favorite]

Next Big Thing: Engineer's Life Mission (CNN Video)
posted by Brian B. at 4:29 PM on November 6, 2010

2. There are moving parts in the new scheme. Not mechanical parts and bearings and such, but the hydrogen does move.

Chuckles, that's like saying that there's moving parts in a hot water kettle, because the water circulates as it heats.
posted by IAmBroom at 10:21 PM on November 6, 2010

Joseph Gurl: ">if they can't share in the profits why would they bother putting any of their time into it?

I know, right? Why would someone contribute, say, 9 posts, 712 comments, and 44 favorites to something without sharing in the profits"

Putting in a few minutes here and there to a web discussion forum isn't quite on the same scale of time donation as committing several years of your life to fundamental R+D :)

I don't think an hour here & there from a materials science postdoc over a decade is really going to cut it here...
posted by pharm at 1:35 AM on November 7, 2010

Chuckles, that's like saying that there's moving parts in a hot water kettle, because the water circulates as it heats.

Yes, exactly. And hot water kettles have wear lives commensurate with this fact. Unlike solid state electronics, which have essentially no wear life whatsoever.

And ya, you'll tell me that kettles last a long time. The only reason they last so long in practice is that they are built very robustly compared to the amount and type of use they get.
Fact is, calcites build up, metals corrode, components of plastics get leached away. All of those effects are accentuated because the material inside is circulating. If you could stop the circulation, the deterioration could be greatly slowed/stopped.

I don't know, maybe the distinction isn't that important fundamentally. Once you start building semiconductors near the scale of single atoms, they might start to have similar problems. It seems important to me though, because the article implies that this approach to a heat engine would be inherently more reliable than the existing alternatives. Absolutely not true.
posted by Chuckles at 1:11 PM on November 7, 2010

Ok, seriously cute device. But the 'green energy' angle is completely bullshit. Even if it does everything promised, it will have little use outside of deep space exploration and perhaps some weird military stuff.


How it works:

The JTEC device is essentially a counter-current heat exchanger with proton-permeable membranes at either end, forming a loop:

       _______________________________________
    / |      ____   high press H2-> _____     | \
COLD  |-PEM-|____|||||EXCHANGER|||||____|-PEM-|  HOT
    \ |            <-low press H2             | /

At the hot end, the pressure differential forces H+ ions through the membrane, generating electricity. This is basically how a fuel cell works, except standard fuel cells produce an H+ gradient by using oxygen to strip the hydrogen from the cathode side of the membrane.

At the cold side, the other PEM works like a fuel cell in reverse, pumping hydrogen from low pressure to high pressure. This process consumes electricity. But because of the lower temperature on the cold side, less electricity is consumed than is produced on the hot side.


Big problems:

1) JTEC depends on a high efficiency low-temperature fuel cell. This is a very difficult material science problem. Current solutions are very expensive, inefficient, and unreliable. There's been a huge amount of work done on PEM fuel cells, but they have yet to see significant use outside of research labs and technology demonstrations.

On the plus side, JTEC's closed cycle avoids the problem of catalyst poisoning. It also gains a bit of efficiency normally lost in the oxygen reaction.


2) JTEC depends on a high efficiency high-temperature fuel cell. Good news: high-temp fuel cells have been successfully commercialized and they work great! Bad news: solid oxide and molten carbonate fuel cells both rely on oxygen transport, not hydrogen. So they're completely inapplicable to JTEC. Only phosphoric acid cells actually move hydrogen ions, and they won't work above 200 C (and have pretty crap efficiency, too).

On the hot side, I think JTEC is pretty much on their own.
For normal fuel cells, it doesn't matter whether you're moving hydrogen or oxygen. At high temperatures, ceramics are favored because they're chemically stable and—this is key—ceramics are made of oxygen. I suspect hydrogen-permeable high-temperature materials are a lot harder. They've certainly seen less commercial research.


3) JTEC losses = thermodynamic losses + high-temp fuel cell losses + low-temp fuel cell losses.

Even if you solve all the engineering problems, this thing is still less efficient than either a Stirling engine or a fuel cell by itself. And I seriously doubt the real-world efficiency of this thing will beat a combined-cycle steam plant.



Ok, so there are several multi-billion-dollar engineering problems to be solved before this thing can be built, and a 'worst of three' theoretical efficiency. Why the hell would I want such a thing, what's the benefit?

"No moving parts!"

Ha, good one. Since when is turbine maintenance a significant cost in power plant operations? In a solar-thermal generating plant, I expect most of the maintenance goes to the thousands of mirrors. They break in windstorms, the motorized mounts fail, and you have to wash them. The central steam turbine, on the other hand, is probably the most reliable device in the plant.


The Atlantic article and Mr. Johnson himself are seriously overselling the JTEC concept. JTEC isn't a green energy solution, it's a long-term research project on basic material science and thermodynamics. It's not going to "make solar power competitive with coal" or "fulfill the promise of renewable solar energy". Granted, it's a pretty cool device. But that PayPal button? Decidedly less so.
posted by ryanrs at 6:45 PM on November 7, 2010 [5 favorites]

It should also be noted that significant improvements to heat engine efficiency will likely discourage progress in renewable energy. Increased thermodynamic efficiency decreases the running costs of fossil fuel plants, making renewables less competitive.

If we really want to "make solar power competitive with coal", we need to decrease the overall efficiency of fossil fuel plants (perhaps by taxing emissions).
posted by ryanrs at 7:10 PM on November 7, 2010 [1 favorite]

Why the hell would I want such a thing, what's the benefit?

"No moving parts!"

Ha, good one. Since when is turbine maintenance a significant cost in power plant operations?

Let's not forget the original cost of the turbine. I also seem to recall that in nuclear applications, especially at sea, a turbine presents plenty of problems to be designed around. I can certainly see why a developer of an electric airplane would not want a turbine. Don't get me wrong, though, some of my best friends wear turbines.
posted by Brian B. at 8:21 PM on November 7, 2010 [1 favorite]

You what else "presents plenty of problems to be designed around"? A ship with hundreds of 18-year-old boys, firearms, and a nuclear reactor. The fact that they made it back at all is a testament to the reliability of gas turbine powerplants.
posted by ryanrs at 9:16 PM on November 7, 2010

Come to think of it, there is one kind of turbine that is notably unreliable: wind turbines. Of course this has no application in that field :)
posted by Chuckles at 10:39 PM on November 7, 2010

(On preview: good fuel efficiency though)


As to gas turbine cost, here's what India paid:

GE Wins Indian Record $750 Million Turbine Order for Reliance Power Plant

The turbines, which GE valued at more than $750 million including service agreements, will add 2,400 megawatts to the Samalkot plant in Andhra Pradesh state and start operating in 2012. The so-called combined-cycle plant will convert exhaust from a gas-turbine generator to power a steam-turbine generator, allowing extra power production without more fuel.

The purchase consisted of three 9FA combined-cycle turbine systems. Per-system specs:
- two gas turbines
- one steam turbine
- 787 MW combined-cycle output
- 6308 kJ/kWh
- 57.1% net plant efficiency

For comparison, efficiency specs of the 9FA simple-cycle turbine (i.e. no steam turbine):
- 17.0:1 pressure ratio
- 9575 kJ/kWh
- 36.9% efficiency
- 602 C exhaust gas temperature

Above figures assume stock configurations described in GE Heavy Duty Gas Turbine Products.
posted by ryanrs at 11:02 PM on November 7, 2010

I've worked out some more efficiency numbers. As a reference, I've been using those natural gas turbines GE recently sold to India. I believe my earlier comments underestimated the benefit of the JTEC device. Previously, I was comparing the JTEC device to existing combined-cycle fossil fuel plants. In that context, the JTEC looked like a very, very hard problem with only marginal benefits. I think I was wrong.

The JTEC part won't be competing against the 57.1% combined-cycle plant, it would only replace the steam turbine portion. In the GE 9FA system, the gas turbines capture 36.9% of the combustion energy and send the remaining 63.1% downstream to the steam turbine. The steam turbine captures 32% of that 63.1%, thereby increasing overall plant efficiency from 36.9% to 57.1%.

Eff_total = Eff_gas + (1 - Eff_gas) * Eff_steam

To improve the overall efficiency of the GE 9FA system, JTEC needs to achieve greater than
32% efficiency from a 602 C heat source (the gas turbine exhaust gas temperature).

Assuming only carnot losses and losses in the two proton membranes, the JTEC efficiency budget is
Eff_carnot * Eff_pem_hot * Eff_pem_cold > 32%

Assuming equally efficient PEMs,
(1 - Tc / Th) * Eff_pem^2 > 32%

Assuming a cold side temperature of 15 C,
Eff_pem > sqrt(32% / 67.1%)
Eff_pem > 69%

That's actually achievable today on the low temp side. It's expensive and unreliable and hasn't made it out of the lab, but it's definitely doable. The 600 C high-temp PEM doesn't currently exist, as far as I know.



But look what happens if you plug in some optimistic numbers and work out the efficiency gains.

For example, let's assume 85% efficiency for both the PEMs:
Eff_jtec = 67.1% * 85% * 85% = 48.5%
Eff_ge_plant = Eff_gas + (1 - Eff_gas) * Eff_jtec = 67.5%
Eff_solar_plant = Eff_jtec = 48.5%

The GE plant goes from 57.1% to 67.5% overall efficiency, i.e. an 18% increase in output for the same fuel.

The solar plant, on the other hand, goes from 32% to 48.5%; a whopping 51.6% increase in output.

The difference in efficiency improvement occurs because solar concentrator plants supply low-quality heat (i.e. not very hot). The steam temps at the hot side of a solar concentrator are nowhere near the peak combustion temperatures inside a gas turbine. Since the JTEC is a low-temperature device, it is only applicably to a natural gas plant's steam cycle, which provides only which contributes 1/3 of the total plant output.

A solar concentrator plant, on the other hand, only has one cycle, the steam cycle. So the JTEC device can improve the efficiency of the entire plant. Therefore the JTEC device has much greater impact on solar concentrator plants than fossil fuel plants. Nuclear plants would also see very large gains, as they have notoriously shitty thermodynamic efficiencies.
posted by ryanrs at 6:07 AM on November 8, 2010 [5 favorites]

Electrochemical Hydrogen Compressor

That's the PEM hydrogen pump that sits at the cold end of the JTEC. I found several examples once I figured out what other people were calling it.
posted by ryanrs at 6:22 AM on November 8, 2010 [1 favorite]