Wikipedia:Reference desk/Archives/Science/2016 August 25

= August 25 =

Proxima Centauri b
How come that we have discovered thousands of exoplanets located very far away -- and even know some amazing details about them -- and yet had not discovered a planet orbiting the CLOSEST star to the Sun until yesterday (Aug 24, 2016)? --Qnowledge (talk) 12:18, 25 August 2016 (UTC)
 * Firstly, it wasn't discovered yesterday, our article states first hints were seen in 2013, and targeted observations started in January this year. Exoplanets get discovered though a number of different techniques. The 'simplest' is if a planets orbit puts it between its star and us, periodically dimming the host star (transit method). This is likely not the case of Prox B, and is indeed not the case of most exoplanets. It's only because of the transit finding Keppler mission that we have this larger number of new exoplanets, until that the number was a lot lower (see timeline in Methods of detecting exoplanets, only a 100 or so in 2013).
 * The way to find an exoplanet regardless of orbital plane is through the radial velocity method, or observing the movement of the star and seeing how the gravity of the planet affects it. This movement is proportional to the relative size of the planet and its distance. This is why a lot of the initial exoplanet findings were massive planets, orbiting very close to their star. Detection of roughly Earth sized worlds is still a fairly new thing. In the case of Proxima Centauri, this is a very dim star (0.1% of Solar luminosity). So although it's nearby, the amount of light we can observe will be lower than a bigger star that is considerably further away. I would imagine this hindered the detection, which as it is took I believe almost two months of near-continuous nightly observation by two massive telescopes. Fgf10 (talk) 12:47, 25 August 2016 (UTC)


 * How do we place a "date" on the discovery, anyway?
 * I used to hang out with some of the folks who made their professional careers out of studying Kepler scientific data. They mentioned observations, and surveys, and research.  Among our many conversations, we talked about Proxima Centauri, because it's a very interesting target for observation and analysis.  Surely they were thinking and hoping to find something noteworthy in our neighboring system.
 * Maybe the data that has now been published came up in casual conversation - who can remember?
 * But casual conversation about a few bits of tantalizing data do not meet the bar for good science.
 * It takes years of controlled, repeated observation before a professional scientist will publish a finding. For people whose professional reputation depends on being reliable and correct, they like to be very very certain before formally announcing a finding to the outside world.  This is a different culture than we have here on Wikipedia: scientists do not "boldly publish" in primary-source peer-reviewed journals and hope that someone fixes up any errors later.
 * This makes it difficult to ascertain "discovery date." Astronomers have studied Proxima Centauri for a very long time.  They have collected optical images and other data for years.  Legions of researchers have pored over the results.  If a planet was present, we've been "seeing" it for a really long time - but the picture, metaphorically, was always too "blurry" to say very much about it with strong certainty.
 * Finally, we have some certainty. That is why the research was announced now.
 * As exciting as this kind of discovery is, we have to be responsible scientists: we need to track down the official press release, the (main) peer-reviewed paper published in the journal Nature, read the findings, scour the data, read lots of good books and research on the topic, and maintain healthy skepticism, and let the entire community of expert astronomers and scientists weigh in. Here at the Wikipedia Science Reference Desk, we have lots of friendly people who can help you find excellent reading material to provide context at whatever technical depth you are interested in.
 * This is big news - and in context, it's a pretty important finding. It will hopefully motivate policy-makers in the United States and elsewhere to continue supporting (with money) the scientists and the equipment that enable these discoveries.  It will hopefully inspire a large segment of the population, who are not professionally involved, to keep looking up.  But let's not get ahead of ourselves: science moves slow.  If you formally study astronomy, a big part of your education will be grokking and re-grokking that in space, things are big.  Really big, just immensely vast.  When we concern ourselves with cosmic length and time scales, which are inherent to the search for extrasolar planets, everything happens very slowly.
 * Nimur (talk) 15:51, 25 August 2016 (UTC)


 * Basically it comes down to the High Accuracy Radial Velocity Planet Searcher. The way this planet was found was by slight changes in the frequency of the light from the star (Doppler effect).  This ever so slightly changes what angle the light passes through a prism or bounces off a diffraction grating, such as the HARPS echelle grating.  The HARPS instrument is extraordinarily sensitive, kept mechanically isolated, incidentally developed the most accurate spectrum of thorium emission known etc.  All this sensitivity, which was improved only recently, gave it the power to find Earth-sized planets around small stars (still not big ones).


 * So this is a procedure where the color of the light is absolutely critical, yet its intensity (and the distance of the star) is not so much so. Wnt (talk) 18:34, 25 August 2016 (UTC)
 * You're still going to have to get photons through your grating to measure them. Luminosity and collecting area are still important parameters. One photon isn't going to be enough. Fgf10 (talk) 18:48, 25 August 2016 (UTC)
 * Of course. But the point is, given existing telescopes, the image can be made bright enough to get that data for stars much further away nearly as easily as for Proxima Centauri.  The improvement in this search instrument was the rate limiting factor for getting this result. Wnt (talk) 19:18, 25 August 2016 (UTC)
 * The Nature paper actually addresses the question: "The Doppler semi-amplitude of Proxima b (approximately 1.4 m/s) is not particularly small compared with other reported planet candidates. The uneven and sparse sampling combined with the longer-term variability of the star seem to be the reasons why the signal could not be unambiguously confirmed with pre-2016 data rather than the total amount of data accumulated." --Wrongfilter (talk) 19:43, 25 August 2016 (UTC)
 * "Space is big. Really big. You just won't believe how vastly, hugely, mind-bogglingly big it is. I mean, you may think it's a long way down the road to the chemist, but that's just peanuts to space." —The Hitchhiker's Guide to the Galaxy --71.110.8.165 (talk) 03:58, 26 August 2016 (UTC)

Thanks for the answers, really helpful! --Qnowledge (talk) 11:23, 26 August 2016 (UTC)

Why soap foam is white irrespective of the color of the soap?
Sandeep Bhadkamkar (talk) 16:35, 25 August 2016 (UTC)


 * Your assumption is not entirely true. Soap bubbles can have color. Usually, they don't - if you look down at them. Adding color to bubbles turns out to be a rather difficult thing to do. Kids like to blow soap bubbles. Imagine a product that blows orange, blue, or purple soap bubbles instead of the normal clear bubbles. They exist - now. They didn't until some people put the hard work into identifying ways to color bubbles. What you normally get are clear bubbles with all the color pushed into a tiny dot on the bottom (due to gravity). If you have a lot of tiny bubbles (foam), they look white. That is just because you are seeing a lot of light reflected off a lot of little bubbles. If you were to look at them from the bottom, you'd see the color. I did this as an example. I put green dye in soap bubble solution and used a bubble machine to blow bubbles all over a glass table. From above, they began to look like a whitish foam on the table. From under the table, you could see a little green dot on the bottom of the bubbles. Where did I get the idea? This is a much better story...
 * I knew a guy named Tim Kehoe who worked for years on making colored bubbles. He got the soap and dye just right and made red, blue, and green bubbles. He got people to back his product. He took it public at a huge presentation. The crowd was huge. He told the public about it. The kids were excited. The parents were ready to buy all he could produce. He turned on the bubble machines. As the bubbles popped, the crowd noticed the red, blue, and green color splotches on everything. They acted like they were being sprayed with mustard gas. Kids were gathered up and the crowd quickly evacuated. He tried and tried to get the people to hear that all you had to do was rub the dye stain and it would go away. That was the trick. The dye that would remain in the bubble happened to be a dye that broke down in very slightly warm temperatures. So, just rubbing it would create enough heat to break down the dye and the stain would go away. It was a failure. Crayola stepped in and branded the product to turn the failure into a mediocre success. You can now purchase Crayola colored bubbles. Just remember to rub the stains to make them go away. 209.149.113.4 (talk) 17:09, 25 August 2016 (UTC)
 * Here's an RS for the IP editor's claim. μηδείς (talk) 17:39, 25 August 2016 (UTC)


 * Some foams (using a detergent base) used in firefighting are orange -- but I can't find a source from a cursory search. 2606:A000:4C0C:E200:1821:CD59:E35A:CB68 (talk) 20:03, 25 August 2016 (UTC)
 * None of you are anywhere near the answer, you can't see the Forrest for the trees. If you have kids and have ever used food coloring at bath time, you will know that "suds" are still white, no matter the color of the water or the soap. The reason is because a tiny bubble acts more like a mirror than a "droplet" of water. In fact most of what you are actually "seeing" when you look at soap foam is actually internal reflection. Therefore when viewing soap foam under white light, the foam will appear white. If you put them under a red light or a blue light, the bubbles will appear red or blue. Vespine (talk) 23:30, 25 August 2016 (UTC)
 * Sort of--the situation is a little more nuanced still, and sort of between you perspective and one voiced above, imo. There are substances which can produce bubbles that will selectively reflect different wavelengths of light; even the average bubble formed of a saponified material could have some surface irregularities in this regard.   But you're explanation I feel is incomplete.  The reason the bubbles appear white, even though in reality they are largely transparent, is because the bubbles are so small and numerous and they are all refracting the ambient light from all sources and directions at a multitude of different angles.  Your analogy of a mirror isn't quite apt because most of the light actually passes through the bubble, but it's passage is altered as a result, and only a small part is outright reflected so that it passes not at all through the surface of the bubble. Those different colors blend to a white for us because the density of the photoreceptors on our retina (and our ability to process the stimuli they receive and then relay) are limited by biophysical reality.  If we had theoretically "perfect" (but physically impossible) resolution to our eyes and visual cognition centers, such that we perfectly perceived and mapped each and every photon which reflected from those bubbles and struck our retinas, then we would see a beautiful cacophony of color.   But the reality is that we receive that stimuli as "white light"; that is, light that is roughly distributed across the visible spectrum, with a little bit of tint from the base color of the soap when it is in it's solid state, since some of the soap "froth" is not composed of bubbles.  It's always worth remembering that colour does not exist as a physical force; light traveling at different wavelengths exists, but "colour" is purely a matter of perception and qualia, and therefore there are many cases where what we experience does not map 1-to-1 with physical analogues.  S n o w  let's rap 05:43, 26 August 2016 (UTC)


 * A couple more points:


 * 1) The surface of a bubble is very thin, so any dye on it is spread very thin, making it difficult to see.


 * 2) Bubbles can also refract light, leading to rainbow colors, like here: . StuRat (talk) 19:00, 28 August 2016 (UTC)

Question about wire
How many volts and amps can BT telephone wire take safely when uses as a power supply rather than a telephone wire? I read the article but it doesn't seem to mention it in plain English, and I'm too stupid to divine meaning out of the complicated mathematical equations listed there. I've currently got 12v DC at 2 amps going through a 30 meter cable powering a fan from a car battery and it seems okay. Can I ramp it higher?

RomanBirdy (talk) 17:10, 25 August 2016 (UTC)


 * You need to know the gauge of the wire. Telephone wire tends to be 24 gauge. Then, look at a chart like the one on this page to see what the limits are for that gauge. If your telephone wire is 22 or 18 gauge, it will have different limits. 209.149.113.4 (talk) 17:15, 25 August 2016 (UTC)
 * See this datasheet (page 4, under "ESP TN/JP") for a typical set of ratings. This particular manufacturer specifies 296 V (as it has to take the voltage for a mechanical bell ringer) at 300 mA, so using it at 2 A is outside the specification.  The problem will be the voltage drop rather than any safety-related issues, but using a properly-specified (automotive) cable would be better. Tevildo (talk) 22:48, 25 August 2016 (UTC)


 * This isn't a great idea. Firstly you'll see a voltage drop across a long piece of thin wire, so your fan is running under-voltage.  Secondly phone wire (unlike alarm wire) is single core, so it's not happy with an impermanent installation or any movement. Expect trouble with cracking and ends falling off. Thirdly, the current question. Cable doesn't have a "current rating" as such, it has a temperature rating and the current flow raises the temperature. The 17th edition of the wiring regs (expensive and unreadable) gives ways to calculate this in a standard manner. More easily, just feel the cable - if you can't feel it getting warm, it's OK.
 * A car battery needs to be fused at the battery end (seriously!). They have an enormous capacity for firey mayhem otherwisem if anything goes wrong. A car battery _will_ start a fire, given half a chance with thin wire.
 * If it were me, I'd use "bell wire". This is cheap two-core stranded wire with a single insulation layer. It's cheaper than phone cable too. Andy Dingley (talk) 23:07, 25 August 2016 (UTC)
 * As long as the wire doesnt heat up, there is no limit to the amount of current that it can carry.--86.187.161.98 (talk) 00:27, 26 August 2016 (UTC)
 * That is true, but the practical limit is reached when the voltage drop from one end of the wire to the other due to Ohms Law causes the device at the far end to malfunction. That can occur long before the wire shows any sign of distress. Akld guy (talk) 02:55, 26 August 2016 (UTC)
 * The categorical claim by IP user 86.187.161.98 is not true because a) all wires that are not superconducting have Resistance so any current causes Joule heating, and b) superconducting wires lose their property of no resistance when the current causes a magnetic field exceeding a threshold (here called the second critical field strength Hc2). It is true that very high levels of Joule heating can be tolerated for short high-current pulses. AllBestFaith (talk) 11:21, 26 August 2016 (UTC)
 * IP 86.187 was not talking about wires in general. He spoke of "the wire", meaning the telephone wire that the OP is using. Therefore, your objection over superconducting wire is irrelevant. Akld guy (talk) 20:47, 26 August 2016 (UTC)
 * The wire can be cooled externally so that its resistance renains constant whatever the value of current.--86.187.175.218 (talk) 20:20, 26 August 2016 (UTC)
 * It's much simpler to use a thicker wire than to run a thin wire through liquid nitrogen or some similar cooling agent, and there would still be a limit.   D b f i r s   20:25, 26 August 2016 (UTC)
 * An intriguing question! I feel like there must be "only so many electrons that can fit" into the conduction band in a piece of metal of given size and shape, no matter what the cooling.  Yet I have no idea what that limit would be called, if it even exists as a concept.  Maybe at some point the voltage differential over a short space ought to become so extreme that pair production becomes possible and the energy disperses uselessly as gamma rays causing V=IR to fail?  But below that, is there a hard limit to how much cooling can theoretically be done?  I have no idea. Wnt (talk) 20:32, 26 August 2016 (UTC)
 * Well, for a real wire, the voltage (between the conductors) will eventually reach the dielectric strength of the insulation and arc over, giving us an upper limit to the current even with unlimited cooling. Another issue will be the ability of the cooling system to extract heat from the conductors through the insulation - even if the outside of the insulation is maintained at room temperature (or lower) by removing an arbitrarily large amount of heat from it, the temperature gradient across the insulation will increase with the power produced in the wire, and eventually the inside of the insulation will melt.  Tevildo (talk) 16:42, 27 August 2016 (UTC)
 * Dont use solid insulation. Use the cooling fluid as the insulating medium. V=IR. Keep resitance down by cooling and you dont need to raise your voltage that high.--86.187.165.85 (talk) 23:55, 27 August 2016 (UTC)
 * @User 86.187.161.98 your reckless claim has been disputed yet you persist in inventing extraordinary cooling schemes that are useless to the OP, only to make your claim seem right. When you so wilfully neglect this rule, you have lost all objectivity.  AllBestFaith (talk) 17:57, 28 August 2016 (UTC)
 * Combined cooling and insulating fluids has been traditional for oil-cooled transformers for over a hundred years. IBM (and Cray) also did it with computers in the '80s - although "insulator" here just meant "avoiding it being a conductor", rather than "an essential part of maintaining the breakdown voltage between windings".
 * Induction heating machines (very, very low impedance, so a huge current and a tiny voltage) use copper pipes as conductors, with flowing water as a coolant. Andy Dingley (talk) 18:40, 28 August 2016 (UTC)
 * "BT Telephone Wire" is too vague to offer a definitive answer as there are several cable specifications that are used in places that a consumer might get access to the wire. If this is CW1308, the white insulated twisted pair used for wiring between sockets, the individual solid copper wires are 0.5mm diameter. They have a maximum resistance of 12.4 ohms/km which would make your 30 meters come out at a resistance of 372 milliohms, which would drop 744 mV at your 2A load and about 1.5V for the whole round trip. That's going to pump about 10mW per metre into your cable which isn't likely to cause a significant temperature rise in the cable.
 * What nobody else has said is that it is imperative that you insert a fuse in this circuit - you are powering it from a car battery that will happily deliver hundreds of amps into a short and whatever the rating of the wire actually is, it isn't enough to safely handle this current under fault conditions. If the circuit isn't already fused at the battery end you are, potentially quite literally, playing with fire. Using the same basis as above (CW1308) a 100A fault on this cable will throw out 7.44 kilowatts of heat, a 500A fault 186 kilowatts - either would incinerate the cable. The latter is 6.2 kW per metre, more than a classic electric bar fire element. A 500A fault current is well within the possibilities for a well charged car battery. — Preceding unsigned comment added by 2001:8B0:1625:41F:0:0:0:36 (talk) 20:04, 28 August 2016 (UTC)
 * The above are good warnings in place of nonsense about "no limit" current carrying. The OP's link for "BT telephone wire" to British telephone sockets suggests they may be using a common Cat 5 telephone signal cable. This typically has tiny 24 AWG or 0.51054 mm conductors for which this table advises no more than 0.577 maximum amps for power transmission. By greatly exceeding this rating the OP courts a failure sooner or later, likely due to a breakdown inside an Insulation-displacement connector. The wire itself should act as a sacrificial fuse to limit fault current. AllBestFaith (talk) 22:21, 28 August 2016 (UTC)
 * It is unusual to see a UK domestic telephone installed over Cat5 - however the cable core specs are similar enough. I would also be surprised if the OP is using IDC connections here. As to the need for fusing a car battery connection, see the "firey mayhem" comment a few days ago. Andy Dingley (talk) 22:56, 28 August 2016 (UTC)
 * Unable to find that comment. Please provide a link. AllBestFaith (talk) 13:02, 29 August 2016 (UTC)
 * Assuming this is a serious question, this is the diff for Andy's addition of the comment. Which is still visible above. Tevildo (talk) 14:29, 29 August 2016 (UTC)
 * I found it, good comment. Thank you Andy and Tevildo. AllBestFaith (talk) 18:18, 29 August 2016 (UTC)
 * The original Q was about current limits on a piece of wire. No mention of sockets or insulation was made. A single wire conductor has no current limit until it melts due to the I^2R losses. If it is prevented from melting by sufficient cooling, there is no limit.--86.187.166.181 (talk) 23:41, 28 August 2016 (UTC)
 * Suggest WP:DENY instead of wasting space on this repetitious troll. AllBestFaith (talk) 13:02, 29 August 2016 (UTC)
 * I was just answering the Q as it appeared, not how some people would like it to have appeared with connectors, insulation problems, CAT ratings and, oh, and nobody mentioned the golden rule: 1000A/sqmm max. THe fact remains that conductors are not limited in their current density.--86.187.171.175 (talk) 19:18, 29 August 2016 (UTC)

Speed of sound and quantum entanglement
For the concept of "substance" you need a local speed of sound, i.e. a traveling pulse of energy. Isn't this an excited state above the ground state of the quantum entangled atoms? If the collection hasn't settled on a ground state then you can't have this collective action of a well defined wave. So is "substance" defined by a quantum entanglement of the atoms comprising the substance? Hcobb (talk) 17:59, 25 August 2016 (UTC)


 * Substance, stuff, matter, mass. I don't think any of them traditionally and canonically depend on quantum entanglement, but that doesn't mean your idea hasn't been explored. SemanticMantis (talk) 18:21, 25 August 2016 (UTC)


 * Quantization sound results in the phonon, which is as you say an excited state. Graeme Bartlett (talk) 10:36, 26 August 2016 (UTC)

Restating the question a tiny bit, does existence of a Phase_(matter) require a quantum entanglement that extends over it? Hcobb (talk) 04:41, 28 August 2016 (UTC)

Distribution of mass estimates from Doppler spectroscopy
The news of Proxima Centauri b brings up a basic question with Doppler spectroscopy - how much more massive is the planet than the minimum value? In theory you can put up a sort of histogram curve with probability on the x-axis and mass on the y-axis, which starts at 1.27 Earth masses on the left and goes up to infinity at the right. Of course, other factors intervene to make some of those possibilities not so possible, but it is a start. But in reality, trying to go from the integral of a sphere by theta as an indication of the distribution of unit vectors to an actual estimate of what proportion of orbits are measurable is... really freaking confusing me right now. And I can't even use it in the article if I get it. Can someone provide a source I can use for what this curve comes out as? Wnt (talk) 18:13, 25 August 2016 (UTC)
 * It is not Abel transform? Ruslik_ Zero 20:06, 25 August 2016 (UTC)


 * Doppler spectroscopy is sensitive to $$M_{obs} = M_{true} \sin(i)$$, where $$M_{true}$$ is the true mass of the planet and $$i$$ is the inclination of the orbital plane relative to us. The minimum mass occurs if we assume the observed system is edge-on relative to us, $$\sin(i) = 1$$, and in general $$M_{true} = {M_{obs} \over \sin(i)}$$.  In principle, every inclination between 0 and $$\pi/2$$ is equally likely.  This would imply that the average true mass is $$<{M_{obs} \over \sin(i)}> = {2 M_{obs}\over \pi} \int_0^{\pi/2}{1 \over \sin(i)} di = \infty$$.  In practice though, if $$M_{obs}$$ is too small then we will never detect it, and if $$M_{true}$$ is too large then it couldn't be a planet and would need to be a second star.  Astronomers sometimes say things like, there is a 90% chance, $$M_{obs} \le M_{true} \le {M_{obs} \over \sin( 0.1{\pi \over 2} )} \approx 6.5 M_{obs}$$.  Dragons flight (talk) 07:45, 26 August 2016 (UTC)


 * I don't think every inclination is equally likely though. There's only one [circular] orbit [of a given size] that has an inclination of 0, but there are many orbits that have an inclination of 90 degrees (their axes could point in any direction perpendicular to us).  I think there are an intermediate number with an inclination of 45 degrees (axes around a smaller circle) but I wouldn't bet my life on it.  There are a lot of little gotchas like that that have made me resolve firmly not to even kick at WP:OR on this one; if I add anything to Doppler spectroscopy I'm gonna need a decent source. Wnt (talk) 19:40, 26 August 2016 (UTC)
 * Yes, there is an error in the above calculation. Correct averaging is
 * $$<{M_{true}}> = {M_{obs}} \int_0^{\pi/2}{\sin(i) \over \sin(i)} di=\frac{\pi}{2}M_{obs}$$,
 * though the dispersion is infinity. The distribution function is
 * $$F\left( M_{true}\right) =\sqrt{1-\left(\frac{M_{ods}}{M_{true}}\right)^2}\quad M_{true}\in[M_{obs},\infty]$$,
 * whereas the corresponding density is
 * $$f\left( M_{true}\right) = \left(\frac{M_{obs}}{M_{true}}\right)^2\frac{1}{\sqrt{M_{true}^2-M_{obs}^2}}$$
 * Ruslik_ Zero 21:04, 26 August 2016 (UTC)


 * If I understand this correctly, the mode is that the real mass is the observed mass; the median is that the real mass is 2/sqrt(3) = 1.1547 times the observed mass, and the average is that the real mass is pi/2 = 1.5708 times the observed mass. Additional limitations placed on the distribution because the star doesn't noticeably orbit the planet would affect the average most, the median less, and the mode not at all.  One of the more interesting features here is that it is 50% likely that the mass of Proxima Centauri b does not exceed 1.446 the mass of Earth; if we assume equal density (big assumption) the radius goes up as the cube root of that and the gravity decreases as the 2/3 power, so there is a 50% chance that gravity there is 13.6% or less greater than that of Earth.  For some of us that's a whole lot of weight, but still less than we've handled before.  It is indeed no word of a lie then that this is an Earthlike planet in mass ... probably.


 * Presently our article says that there is a 90% chance that the mass is less than 3 Earth masses. To check this, Mobs/Mtrue = sqrt(1 - (0.9)^2) in the F distribution above, or 2.294.  Checking the source ... well by golly, it says 2.3 -- which gives 3 Earth masses for the 90% value as said.  Chalk one up for Ruslik Zero! Wnt (talk) 14:38, 27 August 2016 (UTC)

Why are there zero calories in water?
Water is something you ingest, but the body's cells can obtain no useful energy from it in the form of ATP. In fact, I would guess that there is not one organism on Earth that can obtain energy from the oxidation of water, not even lithotrophs such as the hydrogen bacteria. It also doesn't seem to be possible to use water as a fuel in a machine, such as a water-fueled car. I know that there are many energy-rich substances from which the body can extract no calories, such as gasoline, but it seems to me that nature would have found a way, given the abundance of water on Earth, if it were really possible to oxidize water and obtain useful energy from it. I know that water is an energy-poor substance that is a byproduct of the oxidation of carbohydrates, fats, and proteins in cells. Its chemical bonds are very stable and difficult to break. We can measure the calorie content of foods in a calorimeter and water doesn't burn because it's fully oxidized hydrogen. However, I have a couple of questions. First, how did we find out that we can't get any calories from water? Was it something that we could tell just by observing that water doesn't burn? Or were there studies, perhaps ones showing that there is no increase in body temperature after a person drinks water? Second, is there any substance on Earth that reacts with water to release energy? I'm talking about an electron acceptor here, not an electron donor. I would appreciate any information you can give about the energy content of water. — Preceding unsigned comment added by 174.131.55.2 (talk) 21:58, 25 August 2016 (UTC)

Oh, I forgot to mention that water passes out of the body in an, ahem, unchanged form. Was that how we knew? I'd love to know the history of calories and how scientists figured this all out.174.131.55.2 (talk) 22:12, 25 August 2016 (UTC)


 * Hydrogen peroxide is oxidized water. Sagittarian Milky Way (talk) 22:59, 25 August 2016 (UTC)
 * Water is "burned hydrogen". It would be like trying to start a fire using the ashes from a previous fire. Vespine (talk) 23:05, 25 August 2016 (UTC)
 * So should I never say oxidized x to mean "x but extra oxygen"? Well if it can be made from x, oxygen, and less energy than is released I'm sure you still can. Sagittarian Milky Way (talk) 00:02, 26 August 2016 (UTC)
 * Not if you want to use the term correctly. oxidation has a specific chemical meaning which doesn't actually rely specifically on adding "oxygen" atoms, but the change in oxidation state, which is specifically related to electrons, not oxygen molecules. Water has a higher oxidation state than hydrogen peroxide (+1 vs -1) so I believe it is incorrect to say hydrogen peroxide is "oxidized water". Vespine (talk) 00:32, 26 August 2016 (UTC)
 * This is a somewhat pointless question, unless one gives the context.  Liquid water on Titan certainly hasn't zero calories on the Moon Titan. 00:14, 26 August 2016 (UTC)
 * We're talking about chemical Food energy, a completely different question to how much energy it would take to keep water liquid on Titan. Vespine (talk) 00:35, 26 August 2016 (UTC)
 * Here's the short short explanation as to why you cannot extract energy from water by digesting it. When you digest food, what you're doing is a series of chemical reactions during which chemical bonds are broken and reformed in new ways.  Roughly speaking, you can only extract energy from the food if those chemical reactions are exothermic, that is if the products of the reaction contain less chemical potential energy than the reactants.  Your body basically takes food and turns it into waste products, extracting energy so long as the waste products have less potential energy than what you started with. For carbohydrates, the reaction is roughly
 * (CH2O)x + xO2--> xCO2 + xH2O
 * Now, as long as the products (water and carbon dioxide) have less chemical potential energy than the reactants (the carbohydrate and oxygen) your body can extract energy from the reactions. Why can't we extract energy from water?  Because water has very low chemical potential energy.  There isn't any product you could make from the water that would have less chemical potential energy, so there's no way to extract excess energy from any reaction with water as a starting material.  Thus, you cannot get food energy from it.  -- Jayron 32 13:19, 26 August 2016 (UTC)

Great answer. Thanks!174.131.35.50 (talk) 13:37, 26 August 2016 (UTC)


 * It's not strictly impossible to get energy from water... but close enough. From the first source I saw I'm getting that you get 94.64 kJ/mol out of breaking down hydrogen peroxide into water with catalase, which is a lot.  But formally, the Gibbs free energy depends on entropy, and Le Chatelier's principle applies.  Which is to say, if someone hands you a bottle of absolutely positively impossibly pure water, and you put it under an oxygen atmosphere, then some small fraction of the molecules (calculated via this relationship) will get converted to H2O2.  In theory, you could even have some variant of catalase that produces ATP when you convert water and oxygen to H2O2, and you could use it to extract a couple of ATP molecules out of pure water, thereby proving "there is free energy to be extracted from pure water".  But this is only in theory given perfectly pure water that may not be possible even to obtain, and certainly would be impossible to preserve, since otherwise, even lacking an enzyme, over time the water has probably found a way to do that to itself already.  For example, at some rate it splits to H* + OH* and those OH*s find each other.  But they do the same in reverse, and so there's only the appropriate miniscule amount of peroxide in normal water.  Of course, all of this is really minor as an energy source compared to something like getting pure water and seawater and using osmotic pressure produced by allowing them to come into contact across a semipermeable membrane, etc. ... and that isn't a practical energy source either. Wnt (talk) 19:22, 26 August 2016 (UTC)
 * But not a single one of those is "getting energy from water". Every one of those is "do a chemical reaction to water to make it not-water, and then get energy from that".  -- Jayron 32 01:53, 27 August 2016 (UTC)
 * Yes and no. The trick here is what is "water"?  Is water defined as a collection of pure H2O molecules and nothing else, or as the collection plus such reactive oxygen species as are found at equilibrium under a certain set of circumstances such as temperature and exposure to the air?  Yes, this is basically a mind game - nothing here is going to really turn an engine - but it's just funny to think about.  The thing about philosophy is that you can look at something like water and realize that you're not sure what you really mean when you talk about it. Wnt (talk) 14:45, 27 August 2016 (UTC)


 * To look at it from another POV, if water was brimming with chemical energy waiting to be released, then it would have been, by organisms or other natural processes, leaving us with the less reactive products and very little water. This is pretty much what happens to hydrogen peroxide.  It soon gives up it's chemical energy to become water and O2, so we don't find oceans full of hydrogen peroxide. StuRat (talk) 19:08, 28 August 2016 (UTC)


 * Practical use of water as fuel?

I was reading that desalination requires a theoretical minimum of 0.76 kWh/m^3. I believe that means that if you happen to own a seaside retreat where a rivulet of water passes just one cubic meter each hour into the sea, you can get 760 watts of continuous power out of the reverse process .... in theory. That isn't counting any mechanical or thermal energy you might extract from the rivulet.

In practice, of course, you can drink the water and cut your desalination costs. But if your rivulet happens to be treated sewage or cooling water from an industrial plant, you might want another way. I can picture, rather disgustingly, using some kind of semipermeable membrane to desalinate sea water drawn through tubes against the outflow, either allowing wastewater and hopefully no solute to pass through the membrane, or if you had some kind of sodium-chloride symporter in your membrane. But that's not serious engineering, and besides, I want some electricity that can spin a motor for emphasis. I can picture some kind of electrochemical gradient but that's usually seen in biological systems and I don't know how you'd extract very much energy with it. Is there any existing device that can tap the osmotic power of fresh water flowing into the sea? Wnt (talk) 21:36, 27 August 2016 (UTC)


 * User:Wnt, maybe it's been a while since you've been to an estuary. I think I see what you're getting at from a physics perspective, but in real life, I think it's exceedingly rare to find purely unsalted water flowing into the ocean the way it would in a cartoon schematic. Streams tend to not be able to enter the ocean at high slope, and so they basically can't avoid long stretches of mixing, and that would considerably smooth the gradient that you're picturing as sharp. See e.g. the "long profile" graphic here . Estuarine_water_circulation also hints at some of the complicated mixing that goes on. SemanticMantis (talk) 15:04, 29 August 2016 (UTC)
 * I don't think so. Sure, estuaries are common on flat sediments, but there are many rugged coastlines where fresh water pours down into the sea.  I don't think it's uncommon to find a small freshwater creek with this sort of flow rate pouring directly into salt water.  Besides, even if there were an estuary, an ecologically irresponsible person might still be tempted merely to line the lower reaches of the stream in plastic and create a small dam to keep the water pure before the energy extraction. Wnt (talk) 18:09, 29 August 2016 (UTC)
 * Sure, there are small streams that have low mixing at the mouth/waterfall/cascade, but nothing with much size, and it's going to take a large scale to be worth developing, right? As for putting in dams, if you want to build a dam to get electricity, I'd think that traditional hydroelectric is the way to go. Still, an interesting idea, maybe you could start a new question about it. SemanticMantis (talk) 20:05, 29 August 2016 (UTC)
 * Well, this question was about whether water could be a fuel to run a motor. And in truth, well, it does depend on the environment.  If you lived on a planet where the soil was lithium and the air was cesium, a bottle full of water would be a powerful fuel indeed.  But even on Earth, pure water has some energy that can be tapped by osmotic reaction, I think; enough in such a small creek contains chemiosmotic energy to turn a motor and burn some bulbs and in all ways to seem like fuel in a generator provided the sea is its background environment.  I think.  It might almost be another question to see if this can really be done, but I fear if I haven't gotten a response here so far, I'd be wasting everyone's time to try again. Wnt (talk) 20:50, 29 August 2016 (UTC)


 * If you have both a river and seaside available for your use, you might consider hydroelectric, tidal energy or even wave power. StuRat (talk) 15:20, 29 August 2016 (UTC)
 * True, but that wouldn't be using the purity of the water as fuel, only its kinetic or potential energy. Wnt (talk) 18:11, 29 August 2016 (UTC)