# Physical Insights

An independent scientist’s observations on society, technology, energy, science and the environment. “Modern science has been a voyage into the unknown, with a lesson in humility waiting at every stop. Many passengers would rather have stayed home.” – Carl Sagan

## “Breakthrough car only needs water to go”

“Petrol pricey? Japanese invent car that runs on water”

TOKYO (Reuters Life!) – Tired of petrol prices rising daily at the pump? A Japanese company has invented an electric-powered, and environmentally friendly, car that it says runs solely on water.

Genepax unveiled the car in the western city of Osaka on Thursday, saying that a liter (2.1 pints) of any kind of water — rain, river or sea — was all you needed to get the engine going for about an hour at a speed of 80 km (50 miles).

“The car will continue to run as long as you have a bottle of water to top up from time to time,” Genepax CEO Kiyoshi Hirasawa told local broadcaster TV Tokyo.

“It does not require you to build up an infrastructure to recharge your batteries, which is usually the case for most electric cars,” he added.

Once the water is poured into the tank at the back of the car, the a generator breaks it down and uses it to create electrical power, TV Tokyo said.

Whether the car makes it into showrooms remains to be seen. Genepax said it had just applied for a patent and is hoping to collaborate with Japanese auto manufacturers in the future.

Most big automakers, meanwhile, are working on fuel-cell cars that run on hydrogen and emit — not consume — water.

(Writing by Chika Osaka, editing by Miral Fahmy and Chang-Ran Kim)

You can’t run a car on water.

It’s bogus, baloney, BS. Anybody that tries to tell you otherwise is trying to scam you.

I suspect that I don’t at all have to convince most of the regular readers of this blog of this fact, but I just needed to state that, for the record, if you will.

You can use electrolysis to generate hydrogen from outside the car, and have a tank of hydrogen in your car, but that’s not a water-powered car, it’s an indirectly electrically powered car, using hydrogen as an energy storage medium. No such technology is especially novel, and it’s certainly far from impossible.

There’s also the option of using the sulfur-iodine cycle to disassociate water into hydrogen, using thermal energy – however, in practice, this disassociation is, overall, not dissimilar to the electrolysis of water, other than in that it employs thermal energy directly as opposed to electricity, and can be considerably more efficient than electrolysis, especially when thermal energy is used directly from a heat source – such as a nuclear reactor – as opposed to driving a heat engine to generate electricity from the nuclear reactor, followed by electrolysis.

(Aside: The SI cycle was invented at General Atomics in the 1970s, and the Japanese Atomic Energy Agency (JAEA) has conducted successful experiments with the process with the intent of using high temperature Generation IV nuclear fission power reactors to produce hydrogen on a large scale.)

For that reason, I will focus on electrolysis for this post, but just remember that in terms of stationary generation of hydrogen, the SI process could be substituted anywhere where electrolysis is mentioned.

The fundamental claim of the water-powered car is that electrical energy from the car’s electrical system is used to generate hydrogen via electrolysis of water in situ, which is burned in the internal combustion engine, generating energy.

Alternatively, it might be claimed that water can spontaneously be disassociated into hydrogen, without any energy input.

It’s a ridiculous claim. It’s simply a matter of basic physics.

$\mathrm{Anode: 2H_{2}O (l) \rightarrow O_{2} (g) + 4H^{+} (aq) + 4e^{-}; E^{o}_{ox} = -1.23 V} \\ \mathrm{Cathode: 4H^{+} (aq) + 4e^{-} \rightarrow 2H_{2} (g); E^{o}_{red} = 0.00 V}$

Thus, the standard potential of a water electrolysis cell 1.23 V at standard pressure and temperature. The positive voltage indicates the Gibbs Free Energy for electrolysis of water is greater than zero. The reaction cannot occur without necessarily adding energy. It is just not thermodynamically favorable.

Electrolysis is not perfectly efficient, of course. It does not convert 100% of the electrical energy into the chemical energy of the resulting hydrogen. The process requires higher voltages what would be expected based on the cell’s total reversible reduction potentials. This excess potential accounts for what is known as electrochemical overpotential. The extra energy supplied, corresponding to these overpotentials, is eventually lost as heat. The reaction overpotential for the oxidation of water to oxygen at the anode is the dominant overpotential in the process, and an effective electrochemical catalyst to facilitate this reaction does not exist.

Platinum alloy electrodes are the default state of the art for this oxidation. The reverse reaction, the reduction of oxygen to water, is responsible for a significant contribution to electrochemical inefficiency in hydrogen-oxygen fuel cells.

Developing a cheap and effective electrochemical catalyst for this reaction would lead to increases to efficiency in both water electrolysis and in hydrogen-oxygen fuel cells – but efficiency can only be increased incrementally towards the ideal 100% – it cannot ever exceed 100% efficiency, of course.

Catalysts are pretty cool, and extremely useful, but they don’t violate the laws of thermodynamics.

The formation of hydrogen gas at the cathode can be electrocatalyzed with almost zero reaction overpotential by platinum, for comparison. Platinum is, of course, an expensive metal, but it is widely used in these applications.

The theoretical maximum efficiency of the electrolysis of water is between 80–94%, whilst in real world, practical contexts, the efficiency achieved is typically around 60%.

It should be obvious to everybody, then, that you cannot generate hydrogen in an electrolytic cell and convert it back to energy in a fuel cell which you then use to generate more hydrogen.

If this could happen with perfect efficiency in all steps then you would have a perpetual cycle from which no useful work can be extracted. Since it can not, does not and will not occur with perfect efficiency in all steps, then it’s basically nothing at all, besides nonsense.

In real-world terms, with an efficiency of around 60% for the electrolysis, and efficiency of 30% or less for energy conversion in the automobile engine, you’re considering an overall efficiency of well under 20% – it simply does not and cannot work.

We have yet another company basically claiming to have invented a perpetual motion machine.

It’s bogus. No, I don’t consider it “unscientific” to come right out and say that, with 100% certainty.
Thermodynamics-violating scams are at least as old, probably older, in fact, as the understanding of the laws of thermodynamics.

Should we wait and see if more data, results, demonstrations or peer review become available? No. I, for one, wouldn’t bother.

Look at Steorn. Remember them, making headlines a year or two ago? They were going to have peer review of the technology, and public demonstration, and all sorts of proof to convince the naysayers. Have we seen any of it delivered? Nope!

As gas prices rise, and energy becomes a central political topic, more and more people are suckered into taking a second glance at “free energy” peddlers.

There’s no need to be “open minded” about any possible wonderful hitherto unknown scientific discoveries here – it’s entirely just the same old snake oil.

By all means, let’s be open minded, but no so open minded that our brains drop out.

“The law that entropy always increases holds, I think, the supreme position among the laws of Nature. If someone points out to you that your pet theory of the universe is in disagreement with Maxwell’s equations — then so much the worse for Maxwell’s equations. If it is found to be contradicted by observation — well, these experimentalists do bungle things sometimes. But if your theory is found to be against the second law of thermodynamics I can give you no hope; there is nothing for it but to collapse in deepest humiliation.” – Arthur Eddington.

However, let’s just assume, for a moment, that you can do it.
Let’s assume, just for the sake of argument, that water can magically be disassociated into hydrogen with no energy input.

1 litre – or 1 kilogram – of water will yield 119.2 grams of hydrogen. Combustion of hydrogen yields 143 kJ per gram – therefore, the energy content of water in this form is just over 17 MJ per kilogram.

Gasoline has an energy content of about 47 MJ/kg – therefore, the “fuel economy” of the hypothetical car that can somehow do what we’re describing here is intrinsically limited to being about 2.8 times worse than a conventional petrol engine.

Bright readers will note that the story referenced above refers to a speed of 80 km (50 miles). Yeah, sure – I’m not sure who’s to blame here – the journalists or those people making the claim. I think they’re both as bad as one another.

Maybe we’ll operate on the assumption that they mean 80 kilometers per hour. That means they’re claiming the “fuel efficiency” as 1 litre of water per 80 kilometers. That’s 1.25 L/100 km, corresponding to the equivalent of 68 miles per gallon or 3.46 L/100 km in normal petrol consumption terms. Well, that doesn’t seem so bad at all – aside from the fact, remember, that it’s impossible.

There is, however just one way, in principle, that a water powered car can conceivably work, without violating any of the laws of nature.

This would have to involve liberation of hydrogen from water via electrolysis, followed by energy generation by means of nuclear fusion of the hydrogen.

Suppose your car’s tank is fueled with 1 litre of heavy water. As for light water, I’ll come back to that in a moment. 1.1056 kg of heavy water corresponds to 55.17 moles of molecular deuterium.

I’m not sure, quantitatively, how much of an effect the isotope kinetic effect of deuterium will have on the electrolysis of heavy water. Probably a measurable effect indeed, but for the purposes of argument we’ll ignore it. The energy input required for the production of one molecule of hydrogen via electrolysis is equal to E = 1.23 V * 2 * e / 60%; where 60% is the efficiency, as previously mentioned, or 4.1 electron volts per molecule of hydrogen.

That’s a significant amount of energy relative to the heat of combustion of hydrogen, which is 3 eV per molecule of hydrogen, however it is completely insignificant relative to the 3.27 MeV liberated from deuterium-deuterium fusion to yield helium-3.

Of course, essentially all anthropogenic nuclear fusion technologies, in existence or under development, at present utilize the much easier ignited deuterium-tritium reaction – however, I will consider DD fusion only, since we’re considering a water powered car, not a car powered by water along with tritium or lithium.

That corresponds to 17.4 terajoules of energy output from a litre of heavy water – assuming that, say, 90% of the energy output is required to sustain the confinement conditions for nuclear fusion to occur in a plasma, then we have 1740 gigajoules of energy content.
The equivalent of something around 50,000 litres (13,000 gallons) of petrol.

Imagine being able to drive your car 1.76 million miles on a gallon of heavy water. (or 750,000 kilometers per litre.)

It can’t be done, of course.

There is, at present, no working, mature technology available that can usefully generate energy via nuclear fusion (with the exception of a Teller-Ulam bomb, and that’s of debatable usefulness.), let alone a fusion reactor capable of fitting within and operating within a car.

But there is absolutely nothing, in terms of the natural laws, that precludes it. It’s not at all physically impossible – it’s an engineering problem. Water-powered cars – real water-powered cars are a fantastically interesting thing to dream about, but we won’t be driving them in the foreseeable future.

Written by Luke Weston

June 15, 2008 at 2:22 am

Posted in Uncategorized

### 5 Responses

1. I LOVE “it runs on water” scams. They are way up there with 1000 mpg carburettors, and better living through cold fusion. There was man in California that claimed he made a device with a dipolar oscillator (I think he meant a microwave oven) which vibrated the molecules of water at their resonant frequency, causing liberation of hydrogen with less that the energy that bound them and the creation of oxy/hydrogen fuel which would then be combusted and the heat converted into energy through a thermionic emissions (the superheated exhaust passed through a hollow filamented vacuum tubes). Some of the current generated would go to the dipolar oscillator but most could power the car. Total nonsense, but entertaining, nonetheless.

truthwalker

June 16, 2008 at 4:02 am

2. Luke:

About fusion and the scalability and breakeven problems, am I correct in stating the following causes affecting accelerator-driven and electrostatic confinement systems:
1. Low energy density in terms of fusion events per m^2, chiefly affecting ADS (Farnsworth’s fusor wasn’t actually that bad in this department).
2. Bremsstrahlung, or more specifically, that the path of the bremsstrahlung couldn’t be predicted because collisions were happening all over the place and in every direction (which applies both to an ADS and the fusor); it’s been pretty conclusively demonstrated (Todd Rider) that losing the bremsstrahlung inevitably causes both systems to go below breakeven.
3. Poor control of the beam; in both cases, the beam thermalizes soon after leaving the accelerator (in an ADS system); my understanding is that this is because the field is acting on only one component of the particles’ velocity vector, leaving the other two with Maxwellian distributions of velocities, and the resulting collisions in the plane perpendicular to the beam rob it of a significant amount of kinetic energy and thermalize the other component (is that the case?)–the corresponding problem with the fusor, it would seem, is identical but stated in polar coordinates.
4. Most of the recovered energy is in the form of neutrons, which is a good thing because they can’t be confined by a magnetic field, but quite a bit of the energy released by the fusion reactions is in the form of the kinetic energy of the products, and the confinement system spends (as I understand it) a decent amount of energy slowing them down–a total waste. This also implies that aneutronic above-breakeven confined fusion is physically impossible–might it be time to stop trying to do complete confinement?

Also, do I understand the physics behind bremsstrahlung correctly to say that the bremsstrahlung is emitted orthogonally with respect to the velocity vector of the particle that emits it?

I ask these questions for a reason–if the postulated causes are correct, and there aren’t any other issues with these specific systems of a similar magnitude, something occurs to me as a possible way around a few of them. But, if I’ve overlooked something (as I suspect I have, given that this is a 60-year-old physics problem), there’s no reason to even bring it up.

Thanks.

Stewart Peterson

June 21, 2008 at 4:52 pm

3. Wow, what a comment. This will be a challenge to respond to, I’m not really an expert on it all, but here goes.

2. Bremsstrahlung, or more specifically, that the path of the bremsstrahlung couldn’t be predicted because collisions were happening all over the place and in every direction (which applies both to an ADS and the fusor); it’s been pretty conclusively demonstrated (Todd Rider) that losing the bremsstrahlung inevitably causes both systems to go below breakeven.

That certainly seems true; it’s not an intractable problem for deuterium-tritium, but it seems that way for the exotic aneutronic fuel systems like p-11B fusion. The Bremsstrahlung losses almost certainly make a useful reactor using these fuels with a quasineutral, anisotropic plasma impossible – I just downloaded Rider’s paper on fundamental limitations on plasma fusion systems, and it seems like a fairly useful review of the topic and an interesting paper.

4. Most of the recovered energy is in the form of neutrons, which is a good thing because they can’t be confined by a magnetic field, but quite a bit of the energy released by the fusion reactions is in the form of the kinetic energy of the products, and the confinement system spends (as I understand it) a decent amount of energy slowing them down–a total waste. This also implies that aneutronic above-breakeven confined fusion is physically impossible–might it be time to stop trying to do complete confinement?

Well, if you consider a fission reaction, by far the majority of the energy is carried away as kinetic energy by the neutrons, since the neutrons are very low in mass relative to fission product nuclei – however, in the case of fusion, the relative mass difference between the neutron and the He-4 nucleus (Or whatever the product is, almost always He4) is much smaller, relatively speaking. The He-4 kinetic energy is basically just retained within the plasma – they just contribute to some of the energy that is recycled back into heating the plasma, which is OK, since you need that heating anyway.

You raise an important question – in an aneutronic reactor, how can the energy be usefully recovered? In all serious proposals for (generally Tokamak based) fusion power systems, like ITER, the neutrons heating the lithium blanket are the primary mechanism by which useful energy is recovered.

All serious proposals for fusion energy rely on neutronic reactions, such as D-T fusion – perhaps that’s why? The most notable counterexample is perhaps Robert Bussard’s fusion reactor, and I believe such proposals rely on unconventional mechanisms to harness the energy from charged particles in the plasma. Certainly, some kind of exotic mechanism is certainly needed if aneutronic above-breakeven fusion is to be accomplished.

These conversion techniques can either be inductive or electrostatic, based on making charged particles work against an electric field. If the fusion reactor worked in a pulsed mode, inductive techniques could be used.

A sizable fraction of the energy released by aneutronic fusion would not remain in the charged fusion products but would instead be radiated as X-rays. It has been proposed that some of this energy could also be converted directly to electricity, ecause of the photoelectric effect, X-rays passing though an array of conducting foils would transfer some of their energy to electrons, which can then be captured electrostatically, but it’s an extremely challenging idea.

Bremsstrahlung occurs due to the acceleration of a charged particle, and it’s the acceleration vector that determines the geometry of the radiation, which normally takes on a dipole form around the charged particle, but can be “beamed” by special-relativistic effects.

If the particle’s acceleration is parallel to velocity (like, say, if the particle is just being accelerated in some electric field), the expression for the angular distribution of the radiation is given here:

However, the Wikipedia description is pretty limited; it can also be determined for the case where acceleration is perpendicular to velocity (like where a charged particle is subjected to a magnetic field, and moves in a spiral path as a result, like in a cyclotron.)

There should be plenty of literature available where you’ll find details of the calculations of the angular distribution of the radiation, as well as pictures of the actual pattern corresponding to such arcane looking mathematics. The book I have in front of me at the moment is a good place to start – Classical Electrodynamics by JD Jackson.

Luke Weston

June 22, 2008 at 5:17 am

4. Ah–I see on re-reading my comment that I didn’t specifically say I was referring to bremsstrahlung emitted by the orthogonal collision of two beams of (say) deuterons; it makes no sense to imply that bremsstrahlung is dependent on the velocity vector! But it appears from the reference that the power is undefined at zero, zero at [pi] and maximum at [pi]/2 and 3[pi]/2–pretty much the opposite of what I thought. Should’ve looked it up myself; sorry, and thanks for the book recommendation.

I was thinking of two identical cones of deuterons, or one cone of tritons and one cone of deuterons, with the vertices touching each other, so that, assuming that the particles are moving along the possible generatrices of the cones toward the vertices, a high proportion of the collisions would be orthogonal compared to a normal beam-beam ADS. Given the above angular distribution, it would seem like the bremsstrahlung projected on the wall of a cylinder surrounding the cones would be a “shorter” cylinder, the length of which would be inversely proportional to the aspect ratio of the cones. So, it would seem that if the aspect ratio were maximized (i.e., if the cones were lengthened to the point where you start to lose beam coherence), the projected “cylinder of radiation” might be short enough to use a waveguide–I’m assuming the bremsstrahlung is x-rays here–to create a “loop” of x-rays going from the reaction area at the center of the cylinder out to the waveguide mounted in the wall and back (as I understand it, the angle of incidence has to be less than 5 degrees). With that in place, I would think the reaction products would follow the path of least resistance and expand in a toroid centered at the vertices of the two cones as well as inside the hollow part of the cones where there’s no beam. At that point, assuming enough clearance between the bases of the cones and the cylinder as well as enough space inside the cones for the reaction products to pass through, a solenoid coil around the cylinder could direct the reaction products to the ends of the cylinder (obviously, there would be bremsstrahlung produced here, too, but that’s output energy and it’s recoverable with a lead blanket, so I don’t think it’s a problem). IIRC, beryllium is porous enough and the particles would be moving fast enough that a reservoir of liquid lithium (acting as a moderator) behind beryllium caps on the ends of the cylinder would capture the products and convert their kinetic energy to heat. As for the neutrons, a prismatic carbon fiber insert could be placed in a larger reservoir of liquid lithium surrounding the cylinder, so that any neutrons that didn’t thermalize immediately would at least stay in the container until they did; obviously, this would breed tritium, which would be recovered (presumably using a vacuum off-gas system) and used in one of the cones.
Now, assuming that’s not totally wrong, the issue is how to create the cones. An obvious (but perhaps wrong; I don’t know) way to do it would be a toroid in which the particles would accelerate from the inside to the outside (poloidally; looking at a cross-section of the toroid from the side, the circle on the left is moving counterclockwise and the circle on the right is moving clockwise) until the particles reach “escape velocity” (for lack of a better word) and exit from a slit cut toroidally around the toroid near the bottom. The cone then would “point” down; I guess the toroid would be copper or silver magnet wire wound toroidally, with discrete coils (i.e., not a continuous coil but a stack of loops, to prevent weird field effects originating from the slight angle of the wire in a continuous coil; the “loops” aren’t necessarily only one complete orbit of the toroid’s axis, just that all loops in a given layer around the toroid are separate–in fact, each loop must complete at least two orbits to allow one end of the power supply to be connected to each loop without putting gaps in the field, which would require breaking the loop, obviously, while the other end of the wire can be spot-welded to the presumably metal toroidal structural bus, completing the circuit). The acceleration process is also going to produce bremsstrahlung, obviously, and if it would be a good idea to recover that, I guess you could place waveguides at the appropriate poloidal spacing (cross-section: imagine the brush on a DC motor–the waveguides would take a decent fraction of the bremsstrahlung and direct it around the toroid until it exits at a defined point that would result in the formation of a cone of x-rays that would intersect with the vertices of the cones; whatever wasn’t absorbed would continue in a divergent cone until it encountered another waveguide placed above and outside of the opposite accelerator, which would direct it back into the reaction area, from which it would go to the identical waveguide placed on the original accelerator, etc., in a loop).
This would also make the heat recovery system a minor pain in that there are two points–both ends of the cylinder–generating heat, as well as the lithium blanket, and the proportion of energy recovered from neutrons vs. products changes with the fuel mix, which changes the heating pattern, but that seems solvable.

Is there something flawed about the concept, or do you think it would be worth it to do the mathematics and see if (a) the cones can be created, (b) the “bremsstrahlung cylinder” projected on the inside of the vacuum container (above referred to as the “cylinder”) can be made short enough for a waveguide to work, and (c) if the possible efficiency improvements add up to Q>20 (viability) or even Q>1? It seems like solving, minimizing, or mitigating these issues should make Q>>20, but given that it hasn’t been done in 60 years of trying to do it, there has to be something wrong with my reasoning.

Stewart Peterson

June 23, 2008 at 7:51 am

5. […] Posted by Luke Weston on July 14, 2008 An update on the latest “breakthrough car that runs on water!”: […]