It’s getting on time to have another try at explaining how we can recycle the energy we currently throw away, since with a bit of luck I should be fairly close to a Proof of Concept device. The paradox that bugged me for a very long time was that (a) energy is conserved and (b) we use energy up in doing work. When we’re talking about motors and mechanical stuff, there are energy losses at each conversion along the process-chain, and we look at the efficiency graphs to get the best bang for the buck in the design. We look at the energy lost in heat as just being a problem we need to overcome so that the motors etc. don’t overheat. When we’re talking about atomic process, however, there are no such losses and the total mass/energy we start with is the total mass/energy we end with. No gains and no losses. If it looks like there’s a gain or loss, then there’s something unaccounted for and the sums are wrong.
I’ve of course covered this subject over the last few years in talking about 2LoT (last one http://revolution-green.com/science-by-the-sceptics/ ), but this time I’ll approach from a more mechanistic viewpoint. My baseline axioms here are (a) Mass/energy is conserved and (b) the fields we know of are conservative, which means that you can only get energy out of them that you’ve previously put in. As the late Mark E put it, a rock only drops once, and in order to get more work out of it you need to lift it again which will absorb the energy you got out of it. To put it another way, our environment is rigged so that we can’t gain energy from *nowhere* but it always has to come from *somewhere*. The potential energy-level in a conservative field depends purely on the position within that field, and when you have returned to the same spot you will have exactly the same potential energy. The conservative fields I’m talking about are gravity, electric, magnetic and nuclear (of various types). I don’t know of any non-conservative fields to compare them against, though of course a non-conservative field would allow energy to appear from nowhere and disappear to nowhere.
Often, kinetic energy is defined as an amount of energy with a specific direction. This is misleading. Kinetic energy is a scalar – it has only a magnitude and no direction. Since it is moving, though, it does have a momentum vector associated with it. The two parts of kinetic energy may be separately manipulated – you adjust the energy level by adding energy or taking it away, and you adjust the momentum by exchanging momentum with something else. Though you can generally vector-add kinetic energies together, you should really be adding the momentum vectors and the energy scalars separately. In nuclear calculations, we add up all the mass/energy we have and it must equal the total mass/energy after the event, and we also need to be aware that both spin and linear momentum will also be conserved – these rules limit what is actually possible as a result. Still, as regards the energy, we do not consider its direction at all in these calculations – it’s simply a scalar quantity with only a magnitude.
Given a conservative field, what can we do with it? It cannot change the sum of the kinetic and potential energy of our object, after all – during transit of that field, that sum will remain constant. Let’s have a thought-experiment with some tennis-balls in a gravity field. Chuck the balls in the air in any direction and what happens? Unless you’re superhuman, all those balls will come down again. The conservative gravitational field cannot change the total of the KE and PE of those balls once they’ve left your hand, but it can (and does) reverse the upwards direction to downwards. The balls will exchange momentum with the source of that field (here, the Earth) and the direction of the kinetic energy will be changed without changing the magnitude of that energy. Though in the real world there will be energy and momentum lost to the air, and this loss can be significant in ballistics, the field itself has nothing to do with that.
The simple example of the tennis balls illustrates a basic property of conservative fields, in that they will exchange momentum with the object subject to them and thus change the direction of that object. If the field is strong enough and of large-enough extent, then all the random-direction objects we put into it will end up going in a known general direction down the field. The initial momentum-vector of that KE can be 4.pi steradians, whereas the final momentum-vector is limited to 2.pi steradians.
We use this property of an electric field in a solar cell, where the electric field is produced by the PN junction, thus sweeping charged particles (electrons and “holes” which are actually the absence of an electron) in one direction, obviously different for the electrons and “holes”. Photoelectrons are emitted in random directions (leaving a “hole” or unoccupied orbital behind), and exchange momentum with the electric field so they only go towards the +ve side (upfield). Similarly, electrons are attracted from the -ve side to fill a hole nearer to the +ve side, and thus the “hole” moves towards the -ve side (downfield). If you want a gravitational analogy, consider that we have a device that produces weights attached to Helium balloons. The balloon wants to go up, the weight wants to go down, and if we put a small generator on the string connecting them we can produce power from that increasing separation. Of course, that work we get out has been put in (and a lot more) in producing the ballons and the weight, so this isn’t really particularly useful except as an illustration. The separation is however done by gravity (and gravity can’t do any net work) and energy is produced by that separation.
What work does the field do in the solar cell? The answer is that it can’t do any work – it’s a conservative field. This is really the crux of the matter. A conservative field can change the momentum vector but not the total energy level, so there will be no energy-exchange between the field and the object. We know however that in the case of the solar cell then we do get energy out of the cell, and this is simply the translated energy of the random-direction incoming photons. The PN junction electric field doesn’t disappear, since it’s a property of the construction. You can easily test that a solar cell does actually work, even if you’re not too sure of the internal details.
Taking the concept a bit more generally, it can be seen that if we have an object with kinetic energy that is in a conservative field that can act on that object, then the exchange of momentum with the field will ensure that the momentum of the object will end up only in one general direction. If many such objects are put into that field in random directions, then they will all end up going down-field and will not have totally random directions after the action of the field. Using standard Cartesian axes, we put in objects that have total freedom +ve and -ve in the x, y, and z axes, and we get out a population that still has freedom +ve or -ve in the x and y axes, but can only go -ve on the z axis with the +ve excursions on the z axis being disallowed by the field.
We thus have a way of changing a random collection of kinetic energy to one where there is a directionality, and we don’t need to put in work or energy in order to do that. It just happens naturally because the conservative field will exchange momentum with the particles we’re using. What’s more, we’ve had the techniques of how to do this for half a century, and in principle for over a century. We just haven’t applied the principles to the random collection of kinetic energy we call environmental heat, so that we can re-use that energy rather than throwing it away once the energy-differential becomes too small to use with a heat-engine. Generally, this is believed to be impossible because of a misconception of a directionality in the way heat moves, whereas in truth the movement of heat is simply random in the absence of a field that could act on the particles involved. There’s also the problem of applying the mathematics of totally-random systems to the case where we’ve introduced a non-random perturbation, as with the photoelectrons in the electric field of a solar cell. In such non-random situations, it seems crazy to try to apply that mathematics, but it seems that very few people can see that.
I’ve previously explained that the sort of energy we can actually do work with must have both a magnitude and a single specific direction for its momentum vector. If the direction-vectors of a collection of such energies were totally random, then of course we can’t apply a net force and the object will not move. What we do with heat engines is to produce an energy-gradient and thus select a direction for the energy to move (from higher-density to lower density). The amount of work this will yield is well-described by the Carnot limit. The waste heat from this process will however still have a certain amount of kinetic energy, but it’s simply moving in random directions. If we de-randomise those directions using the momentum-exchange with the right sort of conservative field, we can use that energy to do work, too. The Carnot limit is only a limit for devices that use random-direction kinetic energy, and does not apply if we can de-randomise the momentum vectors of that kinetic energy.
The principle of how we recycle the energy we already have is thus depressingly simple. We just need to find particles that will respond in the way we want in a conservative field, and find a way to manufacture a device that utilises this. Put in random-direction energy, and get out energy in a single direction. It’s not actually even necessary to use a conservative field to de-randomise those momentum vectors, and there are several other ways in which we can selectively operate on one of the 3 degrees of freedom (x, y and z directions) to reduce the randomness of a system. Still, the conservative field is a lot easier to explain and to give examples that can be visualised, so that’s where I’ve started.
Of course, I don’t really expect that many people will actually get the point I’m making. It goes too much against our experience of what temperature is and the way heat moves, and most people will automatically think of the temperature of something as a single figure that is an adequate description. It isn’t. The single number of “temperature” actually describes one of a set of Maxwell-Boltzmann curves of the statistics of the actual energy-level of each particle, and in real life even those curves won’t apply during a change of temperature. The measured temperature is thus really an average number, anyway, and averages hide details.
When it comes to radio waves, we’re quite happy to accept that they are there and that we can rectify them with a Silicon diode to get power out. We’re also used to the idea of using an antenna to increase the signal strength by resonance. Much the same applies to IR and visible light (just another section of the EM spectrum), and we’ve covered a few experiments here where nantenna arrays have been made to work in IR and I think we covered the one where visible light (a green laser) was converted to an electrical output using a very small nantenna. The diodes needed must have a very low forward voltage and be very fast. The interesting point is that even at room temperature, everything will emit IR according to its temperature – otherwise you wouldn’t be able to use an IR camera to measure temperatures remotely. Convert those photons to electricity and you’ve got power. All that we need is something that will de-randomise the momentum-vectors of the energy that is around us as waste heat in the environment, and we’ve got enough power to do everything we want to.
When it comes to Electric Vehicles, I’ve pointed out the problem of generating enough power to be able to charge them all unless we start right now in installing new power stations. If we can however sort out recycling environmental energy, then those EVs will self-recharge and it becomes a non-problem. I’ve previously detailed a few possible routes to actually converting environmental heat into electricity, and as usual the devil is in the details and actually making them is a little difficult. Still, we’re probably fairly close to being able to demonstrate something that actually works. They’ll need a lot of money to equip the factories, so I’m not going to give precise details of what I’ve found so far, but you do have the principles and, if you understand them, you can design and make your own. The more competition the better, after all, since the aim is to get electricity dirt-cheap and monopolies aren’t a good way to ensure that.