It's not really fair to compare this, which takes heat at 2000°C, to a steam turbine that takes heat at say 550°C.
A turbine system would have much better efficiency than 40% if its heat was available at that temperature. For example a closed Brayton cycle gas turbine + steam turbine system. Certainly complex and expensive, but could get much better than 40% efficiency.
If a steam turbine could operate at those higher temperatures it would be more efficient. But it cannot do so under any reasonable condition. If you want the higher efficiency from storing high grade heat energy, it's not feasible to use a steam turbine.
Think of the steam turbine as a baseline. Like rating a vehicle in horse power.
Also I wonder if, using similar principles to a heat pump's operation, you could still get juice out of stored heat at lower temperatures. Surely you can have this hot graphite sitting at under 2000°C, heat some fluid/gas to say 1000°C, and then compress the gas to increase its temperature? Surely that would be the ideal solution anyway, since you don't want your hot graphite to become a chunk of useless heat simply for dropping below temp briefly.
On the topic of heat pumps, you could also use a TPV for geothermal power. Since there are no moving parts and presumably no huge steam engine installation, it would be more feasible to have one of these in your back yard. The grid powers a heat pump, you compress the fluid till it hits 2000°C, and your TPV extracts power. The heat pump itself is >100% efficient, so overall you can steal a fair bit of electricity from the ground. Right?
I just want to point out that compressing gas takes a non-zero amount of work. Compressing gas to increase temperature to use temperature to make energy would be better spent using the compressing work directly to make electricity.
The best theoretical efficiency of a heat pump is 1/CarnotEfficency. So the best you can do is to have a round trip efficiency of 1. See Carnot batteries. So using a heat pump to increase the temperature so that a thermal engine is more efficient can’t possibly increase the overall efficiency.
Heat pumps are different than heat engines... I don't think Carnot efficiency applies. A heat pump can move several units of heat energy per input unit of electrical energy (thanks to awesome compressor / refrigerant / phase-change technology). ...They're >100% efficient.
For instance, take a Brayton thermodynamic cycle. It's motor cycle, but you can literally just reverse the arrows on this cycle, it becomes a reversed Brayton cycle, which corresponds to a heat pump (or a refrigeration machine, depending on which inlet of the machine you're interested in). In practice, the reversing of the arrow means the reversing of the fluid flow, which means compression processes are now expansion processes and vice versa.
Carnot theory does still apply. The COP (Coefficient Of Performance) of a Carnot heat pump is 1/n, where n is the usual Carnot efficiency of a heat engine.
With n=1/3, you get a COP of 1/1/3 = 3 = 300%
Carnot efficiencies consider ideal machines. So ideal Carnot batteries have a n * 1/n = 100% efficiencies. It means you can't increase the efficiency of a heat engine by putting a heat pump in front of it. In an ideal case (ignoring all losses and inefficiencies), you would get the same efficiency. In practice, you can only be worse.
Are you sure you aren't confusing the coefficient of performance with the efficiency? If a heat pump was >100% efficient, I could cool my house by leaving my refrigerator door open.
Correct, that's what refrigerated air conditioners do. But that's really just playing a trick with the system boundary in the problem; it doesn't mean the global efficiency has exceeded 100%.
To be clear, I don't think the term is imprecise. Efficiency is always (work out)/(work in). The numbers get wonky when we aren't fully accounting for the "work in". In the case of refrigeration, we have to account for the work it takes to compress the fluid and the resulting waste heat.
I'm not sure if there is any way to make heat pump operating in this scale of temperatures, and keep in mind that current heat pumps can only "pump" over about 50°C max until it's no longer profitable.
That’s not really true at all. Home heat pumps are stuck there because of the refrigerants used. Ammonia heat pumps can go way hotter (though still nowhere close to those temps - more like 150c iirc), it’s just that ammonia requires safety controls that are impractical on a small scale.
Combined cycle power plants take heat at 2000°C using a combination of gas and steam turbines. Their conversion efficiency to electric power is 52%, see Wikipedia. This is proven technology, at least 30 years old, and quite a few of those exist out there.
Steam turbines alone can be operated at higher temperatures by using mercury instead of water. Some plants using this approach were built in the 1920/30s.
Not a steam turbine, a combined gas turbine (for the high temperatures) - steam turbine (for the low temperatures).
The waste heat from the gas turbine would power the steam turbine. Something like a combined cycle gas turbine, but with a closed loop for the gas turbine (say, helium) rather than combustion.
High temperature, though I'm confused myself about how a 40% efficient conversion done twice can match a battery. Probably will need to read the actual paper, I thought batteries were on the order of 95%. Maybe it is more energy dense or something.
"Evaluated herein is one E-TES concept, called Firebrick Resistance-Heated Energy Storage (FIRES), that stores electricity as sensible high-temperature heat (1000–1700 °C) in ceramic firebrick, and discharges it as a hot airstream to either (1) heat industrial plants in place of fossil fuels, or (2) regenerate electricity in a power plant. … We report that systems of 100–1000 s MWh may be cycled daily, and discharged at a constant heat rate typically for 70–90% of the storage capacity. Traditional insulation can reasonably limit heat leakage to less than 3% per day. Preliminary cost estimates indicate a system cost near $10/kWh, substantially less expensive than batteries."
I think it's fair, if we're talking about capabilities of the designs. Since a turbine can't take it's steam at 2000 degrees in practice, that's a limitation of the system, and that limitation limits it's efficiency potential.
That said, we don't really know what this new design can take in practice, so it's probably a comparison drawn too soon.
Heat supplied at high temperature is more valuable than heat supplied at low temperature. Look up Carnot efficiency, and "Exergy". There are limits on how efficient a heat engine can be, defined by the "hot" and "cold" temperatures. A hotter "hot" side allows higher efficiency.
That sounds like room for innovation, rather than an unfair trick. If you can find ways to achieve high temperature cheaply and easily, you get better efficiency.
I think the point they're making is that 40% efficiency is amazing when you have a 1500° difference between the hot and cold side. But at 2000° the theoretical efficiency limit is higher, for all heat engines.
It's like if I compared the top speed of your Porsche on a perfectly straight road, clear weather, no wind, to the top speed of my sedan, but I've got it pointed toward a very high cliff and I'm claiming a top speed of 400 mph (straight down). It's not a fair comparison, because I am cheating by exploiting a greater 'height'.
If we actually compared the same thing, yours would still be at least 50mph faster than mine (at a slight angle from straight down)
Solar concentrators, yes. Or even cut out the steam completely and use the hot exhaust gas from the fire directly. Or combine the element into existing powerplants extracting high temperature and use the residual heat to run steam turbines.
The university where I did a post-bacc had a research project back in the late 1990s to build a TPV-powered hybrid electric car that used compressed natural gas to heat an emitter surrounded by water-cooled PV cells.
With the technology at the time they weren't able to get the efficiency to be competitive with an internal combustion engine, but something like this probably could've made it competitive. I'm not sure if there's any need for it with today's battery performance/price, but maybe as a range extender or something.
Toyota has had a 41% efficient engine in production for five years. If it were set up as a range extender it would likely be even more efficient (fixed RPM range and load can be optimized for.)
Yep, it's amazing what you can do when you've got ten million dollars per motor to spend, run on an extremely tightly controlled / very high quality fuel, and need to only run for about 1500 miles.
Much like space technology, a lot of tech that would never have been considered if there wasn't excess capital available eventually filtered down to "regular" applications.
There are numerous examples of tech developed for F1 that ended up in road cars.
Those are all amazing vehicles. I played a bit with water cooled solar cells, I found that the hard part wasn't the cooling but to be able to draw that much current from a single cell, the attachment points for the wiring typically really weren't up to the job and having multiple attachment points became a necessity. You also need a pretty beefy pump and radiator to get rid of the excess heat, though I guess in a vehicle you would use an active system with a fan.
What I don't understand about the Viking 29 article is how an 8 KW generator is going to power a 75 KW motor, is there something I'm missing?
OK, so perhaps it's not the de-carbonization future we all wanted, but this could be an absolute game changer for planetary exploration, where RTG (nuclear-decay-driven thermal engines) are common. The existing efficiency for those is abysmal, which is actually OK since heat is such a useful product in itself.
I'm looking at Dragonfly, specifically, where an RTG provides the electricity and heat to keep everything alive. Imagine what 10x longer flights would do for that mission.
The advantage does appear to be higher efficiency compared to themocouples, and the TVP should in theory require less cooling relative to the input due to the higher efficiency - However the disadvantage is that TVPs generally require a far higher operating temperature for effective power production, which may actually require more sophisticated cooling.
This particular one requires around 2000C, which appears to be above the critical temperature of most RTGs (though not all!):
The temperature will rise until either the rate of loss of heat balances the rate of its generation, or the container loses structural integrity.
I wondered whether 'critical temperature' meant a temperature above which the junction fails to generate an EMF - which would not be an issue here, even if this were the case, as we know this material works at 2000C - but the article seems to use the term to mean certain temperatures that have engineering relevance, with regard to the design of the device, as opposed to fundamental physical limits to feasibility.
To be clear, wouldn't, say, the atmosphere of Titan provide ample cooling opportunity? Heat distribution is, itself, a feature of RTGs, where planet-side things are usually very, very cold.
"OK, so perhaps it's not the de-carbonization future we all wanted, but this could be an absolute game changer"
I hate to throw water on the situation but we constantly read articles about tech that will be a game changer only to never be seen again mainly because it can't be scaled to the size needed and provide the advantages we need.
Yes, it sounds good. But what we need now is a proof of concept rather than theories on how much of a miracle the tech is. My question is, "How can we help to move it forward to a point where we can see actual advantages?"
Apparently not. This 1991 document [0] says the GPHS [1] heat source's iridium cladding has an upper operating temperature limit of 1,300 °C (1,573 K). There's a suggestion this could be safely raised to 1,500 °C. (This is the structural part that contains the Pu-238 fuel; the plutonium itself could be hotter than this, and the thermoelectric junction is much colder).
(Note the reason given for this limit is to guarantee that, if this capsule accidentally falls to Earth, the cladding has the mechanical performance to stay intact on impact. If iridium gets too hot, its strength is permanently degraded (so it says). I suspect if you voided this requirement -- if you created a separate of class of spacecraft "no longer capable of impacting earth" -- the limit could be raised much higher).
I also suspect that thermal ranges are limited due to the challenges of radiating away heat.
Cooling things in space turns out to be hard. There's no convection or conduction, so all cooling occurs via radiation. I believe the ISS radiators are rated at about 1 watt per m^2, which means that you can calculate the radiator area (and mass) given your power budget.
To a first approximation, the ISS radiators are about the same size as its solar panels.
This is a factor missing from virtually all sci-fi representations of space ships. I had this realisation a few years back talking about 2001 --- the Discovery should have had a huge set of radiator fins given its nuclear propulsion (ion drive I believe, in the story universe). Wikipedia states that there were radiators in the book, though I don't recall that element.
That's a constraint on thermal power, not on temperature (?). Hotter objects radiate faster (T^4 scaling even), so it's *easier* to cool them in a vacuum, all else being equal.
MMRTG [0] dissipates 2,000 W, in a package that only has several m^2 of radiator fins. That's >100x more efficient than the ISS (1 W/m^2). ISS' radiators have to operate below room temperature; MMRTG fins [1] are in the 100 °C - 200 °C range. Higher temperature -> higher heat dissipation per fin area.
(By the way, the comment we're replying to asks about nuclear quadcopters on Titan [2], which is in dense atmosphere not a vacuum!)
Bear in mind that, according to TFA, this work is noteworthy specifically because of the high temperature - past work has achieved the same results, albeit with lower efficiency, at lower temperatures.
I think the idea is that you basically have a giant pit of lava under your neighborhood or whatever, and which you then put a bunch of these things over to generate electricity from the heat. Other energy sources (wind, solar, coal, shakeweight power, whatever) are used to keep the lava molton and replace whatever heat was recently converted to electricity.
If you had a pit huge enough and hot enough, seems plausible to me.
Radiation, if you're in a vacuum. It causes no end of problems getting this right. But if you're on a planet you've probably got an atmosphere to help.
If I'm reading this right, this actually has a larger usable band gap range than traditional photovoltaics -- the article talks about capturing the high-energy photons -- in the first pass. This means that the cell can actually capture the electrons knocked out by those high-energy photons, which is something we haven't been able to do.
Silicon's band gap (1.11 eV) corresponds to 1110 nm (NIR), and any photons with more than 3 eV energy (413 nm) are lost (and all the excess energy in photons in between is lost as heat). Newer cells are around 0.6-0.7 eV, but I don't know their maximum capture energy. That's all the violet light and UV. There's a startup that makes a polymer film that can create two lower energy photons in the band gap from a high-energy photon to capture some of that wasted light. This would seem to be a cell that could capture it directly. Very very cool; that's a lot more energy captured per photon.
What I'm unclear on is why you need to heat it up so much to get to those efficiency levels, and it wouldn't just work as an ordinary solar cell.
The article is confusing, but I read it the same way as you: these guys have figured out a way to capture a larger part of the blackbody spectrum than a normal photovoltaic cell. I don't think they are saying that the cell needs to be heated up. I think they are saying that they capture 40% of the power of a blackbody at 2000K. I am not sure why they don't mention what happens for a blackbody at 5600K such as the Sun.
Replying to myself and GP, I found the answers to our questions in the Nature paper linked from the article.
The TPV is distinct from a solar cell because the TPV is meant to reflect energy back to the source. A solar cell also reflects some energy back to the sun, but that energy is effectively lost. In the TPV case, the reflected power helps keeping the source hot, thus effectively recycling photons in a feedback loop involving the source and the TPV.
Some of the energy incident to the TPV is transformed into electricity, some is reflected, and some is dissipated as heat in the TPV. The 40% efficiency is measured as (electricity+reflected)/incident. Thus, it is not true that the TPV converts 40% of its input into electricity. Instead, the TPV converts some fraction, reflects (40%-fraction) to be recycled later, and dissipates 60% into heat. Using the TPV as a solar cell would reduce the efficiency because the reflected energy is wasted. Part of the magic of the TPV is to design a really good mirror that reflects energy back to the source, which is not a concern for solar cells.
The current design operates best for a blackbody source around 2000K and there is probably no point in using it at 5600K.
I understood it that the efficiency of the TPV depends on the photons wavelength which is determined by temperature on emission. So the band gap probably has an efficiency curve.
Yeah, not to mention that these semiconductors are probably not particularly transparent ? (Not counting the photoelectric photons.) And you're going to have to cool them "before" and "after" the light hits the mirror behind them.
There's a silicon-valley based company that's building a combined thermal storage and TPV system: https://www.antoraenergy.com So these ideas may be get to market sooner than a typical university press release.
> can generate electricity from a heat source of between 1,900 to 2,400 degrees Celsius [...] plan to incorporate the TPV cell into a grid-scale thermal battery. The system would absorb excess energy from renewable sources such as the sun and store that energy in heavily insulated banks of hot graphite. When the energy is needed, such as on overcast days, TPV cells would convert the heat into electricity, and dispatch the energy to a power grid.
Heating graphite based thermal batteries to >1900C using the Sun for long term storage? I'm not sure why the article is refraining from being explicit, but i'm guessing the intended application here is to replace the steam turbine usually found at the centre of high temperature solar thermal collectors.
I wonder how feasible and cost effective it is to insulate a battery well enough to maintain over 2000C for multi day periods without substantial loss? The heat storage strategies used for steam turbines doesn't require such high temperatures.
I attended a paper given at MIT around 1980 which discussed using photovoltaic cells collecting energy from incandescent graphite in a solar collector. There is not really any advance here. The paper discussed using multiple bandgap stacked PV with backside gold reflectors, as discussed in the linked Nature paper.
One of the difficulties is keeping the semiconductors cool. They definitely don't operate at 2000 degrees. One envisions a solar power tower[1] with this technology, but with more graphite mass at the collector.
It's not that hard. All you need is a lot of mass surrounded by a good insulator. The heat has nowhere to go and leaks out at a rate controlled by how efficient the insulator is. But only at the surface area of the mass which relative to the volume is quite small.
There's a basalt based thermal storage system in the Netherlands that is used for seasonal storage. It's heated up over the summer using solar energy and during the winter they pump water through it for heating. It's Basalt in a metal box surrounded by wool and it stays warm through the winter. The temperatures for this system are much lower (500 degrees) but the principle is about the same. https://materialdistrict.com/article/battery-natural-stone/
So, if you can heat the mass to the desired temperature, keeping it there is relatively easy. It would slowly cool over time but not at a rate that is problematic and only very slowly. A large enough system might store heat for months/years.
The Carnot efficiency of a heat engine favors large delta T! This is the theoretical limit of efficiency. For steam turbines this efficiency should be around 60-70%, theoretically.
While cool (1,900 to 2,400 degrees C cool) the Carnot effiencies should be closer to 86%.
The idea that heat engines get more efficient as you increase delta T has been around for a while. The problem is constructing a delta work extraction loop that doesn't have more losses as a result of the delta T increase, ie the practicalities of extracting work energy.
Warning: I am assuming they are working with an approximately room temperature cold side, as article doesn't say. The practicalities of allowing for the delta-t is usually where the efficiency losses are made.
The title is somewhat overstatement, EPA says turbines have efficiencies up to 90%: "Multistage (moderate to high pressure ratio) steam turbines have thermodynamic efficiencies that vary from 65 percent for very small (under 1,000 kW) units to over 90 percent for large industrial and utility sized units. Small, single stage steam turbines can have efficiencies as low as 40 percent."
Yes, it's comparing apples to oranges, or the efficiency of the whole steam turbine plant (electricity/fuel) vs just a single component (electricity/heat).
Is the comparison fair? Well it's tricky but I'd say yes, the generation of just heat from fuel is normally highly efficient so electricity/heat should be similar to electricity/fuel in the second case. However in a turbine generating plant it's not plain heat but it needs to be converted to steam and moved around, so there are many more loses.
That paragraph in [1] is referring to: "Isentropic steam turbine efficiency refers to the ratio of power actually generated from the turbine to what would be generated by a perfect turbine with no internal flowpath losses using steam at the same inlet conditions and discharging to the same downstream pressure."
It specifically says not to confuse them: "Turbine efficiency is not to be confused with electrical generating efficiency, which is the ratio of net power generated to total fuel input to the cycle."
Typical steam turbines have efficiency up to around 50%. I think the EPA is referring to isentropic efficiency[0], which is "efficiency" compared to the theoretical maximum for the delta T involved.
This would be interesting for powering spacecraft because it doesn't have any moving parts. For the most part, spacecraft are not serviced ever, so it's best to minimize things that can break.
Nuclear powered spacecraft have been hard to develop because of the need for moving parts. NASA's cancelled Advanced Stirling Radioisotope Generator, which was supposed to be a more efficient radioisotope powered generator than the thermoelectric generators previously used, had trouble because the moving part based generator wasn't very reliable.[0]
In addition to eliminating moving parts this is also interesting for nuclear powered spacecraft because it may be possible to pass light through a radiation shield to prevent damage to the converter. The problem is that the reactor would need operate at extremely high temperatures. While this probably isn't high enough to melt the fuel, the fuel might not be structurally stable. Although liquid uranium nuclear rockets are being considered[1].
Those thermocouples are less than 10% efficient, usually less than 5%.
This means you have to dissipate more than 10 times more energy as heat than the energy you actually want to use as electricity.
Dissipating heat in space is not easy because unlike earth there is no surrounding fluid to dissipate heat into through convection. This means you have to spend precious mass budget on huge (compared to your energy budget) direct heat radiation systems that cannot leverage convection efficiencies.
Thus, a significant energy efficiency increase would be a big deal for RTG powered spacecraft design. It is curious that the article above does not consider this.
Bonus thought: consider what this lack of convection problem means for "Hyperloop" type vehicle systems that operate in a vacuum tunnel, as most conventional trains dissipate excess braking energy through convection from resistive heating elements atop the roof.
Conventional trains that run under wires can also dump excess braking energy into the overhead. That seems feasible even if you had to have your resistive heating elements somewhere outside the tunnel system.
Yup, Voyager 1 and 2 are still running off RTGs right now since 1977.
However the thermoelectric converter used in those RTGs exploit a different effect, and appear to have quite different properties as a result... What I can tell from a quick dig:
- Thermocouples ("Seebeck effect") operate on a temperature difference, and are more commonly used as temperature sensors. They have the advantage of working at a larger range of temperatures, but the disadvantage of needing to maintain a temperature gradient for power production... Any inherent efficiency is likely negated by the requirement for constant cooling. [0]
- Themal Diodes ("Thermophotovoltaic") is more like a solar panel for a different wavelength (infrared). The principle they operate on suggests no temperature gradient is required, not sure about cooling for other reasons though, but the clear disadvantage is the requirement for a high operating temperature for effective use in power production. [1]
Historically thermal diodes don't appear to have been particularly efficient compared to thermocouples either, obviously this particular one changes that.
> The principle they operate on suggests no temperature gradient is required
A temperature gradient is absolutely required. Ignoring materials problems (such as melting) a photovoltaic cell ultimately does work off a temperature differential because it has to absorb photons. It cannot do that if its temperature is the same as the black body temperature of the lightsource illuminating it, otherwise they would be at an emission/absorption equilibrium.
This is also why PV cells could theoretically work in reverse mode at night, emitting IR into space. They'd just do that paltry power ratings because ΔT(earth, cosmic background) is much smaller than ΔT(photosphere, earth). And they'd have to be made of a material with a much smaller bandgap.
You're right, I kinda knew it needed cooling, but I guess your point is even though the effect being exploited doesn't require a temperature gradient, in practice the heat dissipation requirements are the same if the TVP is to survive - or are they different... surely for the purposes of satisfying the first law the temperature dissipation requirements are dependent on the efficiency of conversion?
Which would make the cooling requirements of this TVP lower (relative to input) due to it's higher efficiency? but still substantial.
If you point concentrated solar at a vantablack object, heatsinked to this 40% efficient TPV, do you get an easy 39.94% efficiency, easily outpacing mass produced photovoltaic or am I missing a loss?
This is just a multi-junction photovoltaic cell optimized for temperatures of the radiant body between 1900 and 2400 Celsius degreees.
Such multi-junction photovoltaic cells, but optimized for the higher temperature of the Sun, have existed for many years and efficiencies over 45% are well known.
So there is no point in heating anything, the concentrated solar light must be directed to an appropriate multi-junction photovoltaic cell, for the best efficiency.
Despite their very high efficiency, the multi-junction photovoltaic cells are seldom used for solar energy, because they are expensive, so they can only be used together with light-concentrating mirrors, to achieve a reasonable cost.
Even with mirrors, the price is still much higher than for normal solar panels, so they might be chosen only when space constraints would prohibit the use of a larger area with solar panels.
The exact price is unknown, because you will not find them at retail.
They have been used for solar panels in satellites or space probes, where maximum efficiency is more important than the price, and in experimental solar plants with movable mirrors that concentrate the solar light from a very large area onto a small photovoltaic cell.
In both applications, the complete systems are very expensive and the cost of the photovoltaic cells is a very small part of the total.
Well any area is enough area for solar power, you just don't get much out of it. I think for a car I remember some calculation at some point was that if you covered the whole thing in panels and let it soak in the sun for a day you'd get a few miles of driving out of it.
Having more efficient cells would give you a few more miles I presume. Not too useful for the everyday user, but if you're doing an off grid trip it would be very useful if you camp somewhere and let your car charge for a few days. I guess might as well roll out some proper panels in that case though.
Planes do have a reasonably large wing surface area that could be panelled up if they're not too heavy (you can make quite large RC planes fly perpetually in the sun even with regular monocrystalline panels) so there would definitely be some fuel savings from it if you had like a hydrogen powered jet that already uses electric propulsion.
What you suggested is more similar to what Solar Thermal systems do (https://en.wikipedia.org/wiki/Solar_thermal_energy). They are known to be more efficient than solar photovoltaics. They have their own downsides of course.
They are more efficient because converting anything to heat is trivial. Heat is just losses basically. A process consisting of 100% losses is great for heat generation.
However, extracting exergy (electricity is pure exergy) from a flow of energy is the tricky part and will always be associated with efficiencies way below unity, based on fundamental principles.
Yes, but labs have also produced 40% efficient multijunction solar cells that work directly from sunlight without the intermediate heat absorber. Off-the-shelf multijunction PV for space applications is I think 36% efficient.
Also, you don't need Vantablack, a regular cavity absorber would be fine.)
The article says, if I'm not mistaken, that the heat source must be between 1900 and 2400 degrees (Celcius), and I would bet that Vantablack loses its blackness at such temperatures?
Wikipedia indicates the melting point of vantablack being 3000C. I would think a black coloring is pretty heat resistant. It looks like it needs to be 500-750C to create it as well
It sounded like it wasn't the heat that got converted to electricity, but rather the photons emitted by the hot object glowing, and since your vantablack object is not glowing you would expect to get nothing.
But maybe not, the glowing comes from black body radiation, so the vantablack material would presumably glow as well (ironically). As long as the heatsink coupling did not block the visible "white" light produced, or glowed itself, then at least the photons from the back would get used. I expect that getting a heatsink paste rated for 2200 C is ... challenging, but, conveniently, you'd do better if you just skipped the paste.
The vantablack material would indeed "glow", as everything in the universe glows. It just does so outside the visible spectrum. The difference between this thermo-voltaic cell and a photo-voltaic cell is that photo is visible spectrum and thermo is IR. It's all just photons!
> I expect that getting a heatsink paste rated for 2200 C is ... challenging, but, conveniently, you'd do better if you just skipped the paste
Liquid metal is some of the best performing thermal paste around. In computer applications that's normally an alloy made from Gallium, Indium and Tin, but at 2200C the majority of metals should work. Maybe Gold to reduce oxidation.
Apart from the vantablack heating issues sibling comments have already mentioned, you'd also need to take into account the energy used by the cooling system for the cold end of the TPV. In practice you'd need to pump either water or air past some form of heatsink and the energy consumption of the pumps would reduce the efficiency below that of the 40% of just the TPV.
This would be a heck of a lot of waste heat to deal with. You could probably boil water with the leftover energy to turn a steam turbine to power the cooling apparatus. A dual stage solar plant.
The reason for the high level of waste heat is that the system has to operate at thousands of degrees C. There is still a huge potential above room temperature. Most systems work closer to room temperature so there is space to squeeze them in after this system has extracted all of the energy it can.
Yes, but the pumping power is about 0.2% of the total power, so this is not a significant consideration in practice. If it was, people would use solar chimneys instead of mechanical fans and pumps.
One minor issue is diffusion due to clouds, similar to all concentrated solar power systems it needs direct sunlight. Normal solar panels can produce some power even under a thin cloud layer.
I’m guessing the vantablack would be destroyed by the concentrated heat or UV. Or is there an industrial formulation that could withstand the high temperatures?
To everyone in this thread commenting about the Carnot efficiency: yes, the Carnot efficiency grows with T, but you must use a Carnot engine for that! The Carnot theorem is an upper bound not a property of a given engine. It’s trivial to build a heat engine whose real efficiency (measured efficiency / Carnot efficiency) approaches 0.
If the stored heat is on the level of getting the emitter into the white-hot, how is the "battery" turned off, to hold the "charge"?
The article mentions a mirror layer as part of the cell, for retaining the energy of out of band photons. Would that be the "off" solution, just bounce them back when they are not needed? Cell moved out of the path? Somehow that triggers "too simple to be true" heuristics in me, but on the other hand... yeah, mirrors (or just very white surfaces, precise direction is not needed) can be quite capable of not getting heated by incoming photons, and that must mean bouncing them back.
You have a heat reservoir, i.e. a well-insulated and very hot object, that stores energy as heat. If you insulate it with mirrors, that can look like bouncing photons back into the reservoir.
When you want to generate energy, you open the insulation and let heat out to hit this chip.
It’s like opening an oven door to let some hot air out.
This thermophotovoltaic cell is tuned to operate with radiant energy from a body heated to between 1900 and 2400 degrees Celsius. That is much hotter than any current reactor core. It wouldn't work with a traditional nuclear plant; it would require (at a minimum) developing reactors that operate at a much higher temperature. The most common power reactor design, the pressurized light water reactor, heats water to around 315 degrees Celsius:
I think the problem isn't that we can't achieve 2000 degrees with fission, but the entire design would need to be re-thought to handle such temperatures.
You need a completely different sort of materials and need to consider new types of risks for a reactor that's supposed to operate at such temperatures.
Most semiconductors don't like ionizing radiation. In addition, the bottom-end of the operating range for this (1900C) is well above what google suggests is the typical coolant temperature (300C).
Does anyone know if we already use things like this to increase the efficiency of existing energy usage in applications which require high degrees of heat output by recycling energy that is otherwise lost as excess thermal output?
The example that springs to mind for me is a steel mill. The temperatures required there easily meet and exceed what is required to generate and store energy with this device.
Probably not in a steel mill, because otherwise you could have just heat water to steam and use it to run a generator. Since this is not done, there is a catch.
An example where this idea works is a condensing boiler where the burned gases heat is used to increase efficiency by 10-30%.
You might be keying on the absence of any advertising whatsoever. So many media sites are so bogged down by advertising that it would be funny if it wasn't so sad.
Beyond that, it does have good fundamental design. Hierarchy of typography is simple and clear, they font used for the article copy is legible and sized nicely. And, because of he lack of advertisements and attempts to pull you further into the site, there's enough white space to give it a more relaxed feeling.
If you want to learn the fundamentals of design, pick up the Non-Designers Design Book. Quick read with great explanations of the core principles of good information design.
Be sure to get the paperback. The design doesn't carry over into the Kindle version.
I found this amusing bio on the Amazon author's page:
Robin Williams spent twenty-five years writing computer and design books and was one of the top three best-selling computer book authors in the world. But the world changed and no one bought computer books anymore, so she went to London and formalized her long study of Shakespeare with an M.A. and Ph.D. in Shakespeare studies.
It requires "a heat source of between 1,900 to 2,400 degrees Celsius, or up to about 4,300 degrees Fahrenheit" -- so your Nvidia gpu should work great for this purpose :)
As others have pointed out, 40% at this temperature isn't really an impressive in itself. Combined cycle power plant have a yield over 60%, while still operating at a much lower temperature.
That doesn't means it's useless (it can improve yields for thermodynamic solar) but it means there's a really wide margin of improvement: the theoretical maximum for such temperature is 87%, and if it could be made as effective as a combined cycle[1], you could a yield up to 70-75%.
[1] by that I means being as close as a combined cycle from the Carnot cycle at the same temperature.
> The researchers plan to incorporate the TPV cell into a grid-scale thermal battery. The system would absorb excess energy from renewable sources such as the sun and store that energy in heavily insulated banks of hot graphite.
The article doesn't provide the efficiency of this insulated hot graphite battery. Given that this device is meant to be paired pretty exclusively with such a battery, it seems like a narrowly applicable solution. Or are there other renewable sources of heat between 1900C and 2400C that could be used for this purpose?
The efficiency of heating things up is pretty much 100%. The efficiency of insulation can be arbitrarily high (up to 100%) or arbitrarily low; heat loss is proportional to surface area and storage time and inversely proportional to insulation thickness. If your heat engine is only 40% efficient you might as well design your insulation to be 80% efficient or so over the expected time span, which would be a few hours for grid storage systems.
I was thinking more about the efficiency of the mechanism that transfers the heat from battery to the TPV engine. The article said that the device has to be exposed to photons coming off a white-hot source.
I suppose that if the TPV devices are closely coupled with the white-hot source this should be minimal but it's not clear from the article whether there is an intermediary step, or how they plan to control the TPV's exposure to the heat source to discharge the battery.
Given that this works via radiation (vs convection like a traditional heat engine), the path between the heat source and the TPV would have to be fairly direct, and not lose much heat to any adjacent non-TPV material.
Maybe they have some kind of massively insulated opening/closing heat shield that they can use in a manner similar to the gates used to control water in hydroelectric plants.
> The researchers plan to incorporate the TPV cell into a grid-scale thermal battery. The system would absorb excess energy from renewable sources such as the sun and store that energy in heavily insulated banks of hot graphite.
It's clearly not ready for production yet, but storing energy as concentrated heat seems like one of the plausible proposed grid-scale storage solutions to me. I'm interested to see where this goes.
Well now. We can have a heat pump with a COP of 3.5 or so. Let's say we put in 1kW of electricity, and take out 3.5kW of heat energy. Now let's take 2.5kW of that heat energy, push it through this gadget at 40% efficiency to get 1kW electricity out. Push that back into the heat pump and, violà, 1kW of free heat left over!
> We can have a heat pump with a COP of 3.5 or so. Let's say we put in 1kW of electricity, and take out 3.5kW of heat energy.
A heat pump COP of 3.5 or higher or happens only at relatively low delta-T between the source and destination temps of the heat pump [1] - like the delta-T typical for space or water heating. The COP degrades exponentially with increasing delta T as it has to work ever harder to pump heat against an ever higher temperature/pressure (assuming thermal storage with a fixed volume with few losses).
The refrigeration cycle can't raise temperatures even hundreds (much less thousands) of degrees C - otherwise we'd already have heat pump stoves and ovens.
This is coincidentally also one reason why all else equal, heat pump clothes dryers (which are great) take a bit longer than conventional technologies to dry clothes: they only reach about 50C (vs 70C-75C for standard gas or electric resistance dryers).
> Push that back into the heat pump and, violà, 1kW of free heat left over!
Because the COP degrades with the delta-T, it bottoms out at 1 (an electrical resistance heater), and in your scenario, you end up with 1kWH in, and 400Wh out, so a theoretically 40% efficient battery.
Minus the electrical generation, and at much lower temperatures, your scenario with a heat pump + thermal storage does however describe how the new domestic thermal heat batteries can work with heat pumps [2].
Entropy. You cannot look at a system purely from an energy conservation stand point. Take Carnot efficiency for instance. That’s the extreme case where it teeters on the fully recoverable entropy. For your example, the entropy price has to be paid somewhere in the system (usually in the heat to work ratio).
You are missing that any heat engine (including photovoltaics !) is limited by the Carnot efficiency. And a heat pump is just a reverse heat engine. And you are trying to couple their inputs and outputs in a loop. I'm not sure that this would even qualify as a perpetual motion machine of the second kind ?
P.S.: A heat pump is usually able to "cheat" an "efficiency" (aka Coefficient of Performance) over 100% because (for a refrigerator) it's defined as the inverse of the usual Carnot efficiency.
When you want heating, you artificially get up to +100% to the CoP from the spent "mechanical" power being 100% converted into heat.
Is "engine" the right word for this? I'd call it a generator if anything. I thought an engine took in power and output mechanical energy.
I understand there are other kinds of engine, such as a gaming engine etc, but in this context it seems like there are better descriptors. "Heat to electricity converter"?
Sometimes I like to take a step back and imagine a world with nearly free energy. Converting sea water to fresh water to irrigate the deserts is always the first thing that comes to mind, but that's mundane compared to the changes nearly free energy will bring, which is the direction we seem to be heading.
As long as proof-of-work cryptocurrencies exist, free energy is impossible. Any time the price of energy falls below a certain point, the energy consumption of the network will increase to compensate. PoW is a floor on the price of energy.
While this is far away from being a production device it's pretty exciting to see 40% efficiency.
How would you go about converting the energy stored in a thermal battery to a high enough temp for this to work? Some kind of heat pump? It has never been clear to me how to concentrate heat like that.
I think they're planning to heat up big chunks of graphite to thousands of degrees C with electric heaters powered by the grid and then attach these thermocouples when the renewable sources start to flag (windless nights) to cover the gap.
Heat radiation losses (scaling with the fourth power of the temperatures) would obliterate such a storage solution. I could only imagine it as very short-term, at which point heating using electricity and converting back shortly after quickly becomes pointless.
I don't know how they mitigate the radiation losses (huge amounts of insulation?), but it appears that graphite has some interesting properties when heated, like it gets stronger 1000-2500C and doesn't expand much when heated [1]. Perhaps those properties help it to store heat effectively at those extreme temps.
> I don't know how they mitigate the radiation losses (huge amounts of insulation?)
My guess is a vacuum flask made of tungsten. Tungsten sheet metal is a thing and it has the second highest melting point of currently known elements, which is 3,422C.
Then for the 'battery' you will want to find a material that has a phase change around the temperature required for this heat engine to operate. The energy required for water to go from 100c water to 100c steam is considerably more significant than the energy required to go from 0c water to 100c water.
Graphite will likely come into play because you need to have electrodes to heat the material up to storage temperature. Graphite is a good for this sort of thing. Although gradual erosion will likely limit the life of the battery.
If that's the case, would the tungsten flask heat up to 2000C and give off the photons that are absorbed by the TPV? I still don't understand how they will control transmission of the heat at those temperatures.
Maybe insulate with reinforced carbon-carbon? I'm not an expert on this, but it doesn't seem outright impossible on the surface, but the details are always what get you.
The insulation I sort of understand. It's how they temporarily breach the insulation to let out the photons to strike the TPV that I don't understand. With normal heat->electricity conversion, you transfer heat to a turbine via a fluid (i.e. water) that is allowed to contact the exterior of a heated vessel, but in this case you need to somehow open a slot to let the photons radiate out.
Install a hinge on the side of the insulation box, swing it open then swing closed a different section that includes the thermocouples?
Or just slide the insulation down in front of the thermocouples when they aren't needed. I have to assume these things mostly work on radiant heat at those temperatures.
You need vacuum super insulation.
Basically take a thermos, and fill the vacuum with (mostly) non-touching reflective foil. It's the golden stuff satellites are wrapped in.
I'm curious what the implications would be for solar panels, or for any device that outputs excess thermal energy (like our computers). Would a solid device like this allow recapturing some of the energy that would otherwise be lost?
TFA says turbines can't operate for heat sources around 2000C. I'd believe they can't operate with steam at 2000C, but can't you decrease the temperature of the steam relative to the heat source by increasing the flow rate?
But you can get 36-40% efficiency with a steam turbine in a fossil-fuel plant, and IIRC coal flame temperatures can exceed 1000C (presumably the steam is much cooler; 500C sounds like a reasonable guess; I could only find that nuclear plants have a relatively low steam temperature of around 280C).
tldr: thermoelectric generators don't have great efficiencies, but by cogenerating heat and electricity they can get viable (if you need heat and electricity, make a bit more heat than you need and put a thermoelectric generator, the non-converted heat will just end up as useful heat). This would be adapted to households, which typically need a lot of heating.
Is there some sort of physical property when you are that hot (2400C) that you begin glowing? Then just open a slot in the insulation so the photons cell whenever you need electricity.
Everything above absolute zero emits infrared photons, as something gets hotter it will also start to emit in the visible range because some of the photons are more energetic resulting in a shorter wavelength. That's why you see the progression from barely visible red to red to orange, yellow and eventually white when you make something hot enough. But it will always continue to output photons at lower energy levels as well, though not in equal proportion (and that's why a hot object is white and not blue, and why a hot gas flame can be blue).
“The heat engine is a thermophotovoltaic (TPV) cell, similar to a solar panel’s photovoltaic cells, that passively captures high-energy photons from a white-hot heat source and converts them into electricity.”
One detail is that objects radiate at all temperatures. The trick is to choose the temperature so that you get a lot of emission in the band that matches the best performance of the PV component.
I'm struggling to figure out what this is. If you can capture heat from a "white hot" object, why is it not just a good solar cell? The sun is easily in this range.
This is great, but all of this feels like it's coming 15 to 20 years too late. I worked with a guy in Berlin who had previously worked in PV energy. He predicted that unless the costs were reduced by a factor of half the price point of fossil fuels that it would go nowhere. We're addicted to cheap, no hassle sources of energy right now, and it's depressing.
You have to wonder what PV would cost if even a tenth of the resources invested in developing nuclear energy were instead invested in solar energy development. I expect by 1980, solar energy would have been the cheapest way to generate electricity, President Carter would have been reelected, no Iran-Contra affair, and no deregulation of banks leading to the 2008 recession, and likely would have avoided 911, both Iraq wars and the one in Afghanistan. We could have saved all kinds of money, had cheap energy, still had plenty of nuclear power, and it wouldn't be so damn hot all the time.
> You have to wonder what PV would cost if even a tenth of the resources invested in developing nuclear energy were instead invested in solar energy development.
Not enough was invested in nuclear. Fossil fuels received massive funding (and subsidies). Had we deployed more nuclear power for baseline and industry, we would have been in a far better place. The Cold War also messed up things, and caused the reactors that were deployed to be the ones better suited for weapons first, energy second.
Note that our solar panels are similar to computer chips. Investing more money would have sped up the development but not in time for it to be viable in the 80s.
> The Cold War also messed up things, and caused the reactors that were deployed to be the ones better suited for weapons first, energy second.
Sure, but those were Gen 1 reactors. We're overwhelmingly using (pressurized water) Gen 2 reactors these days, which, while being downstream of the same processing as the one required for nuclear weapons, and also theoretically capable to be used to make nuclear weapons themselves, are definitely NOT better suited for weapons first !
The Manhattan Project cost, adjusted for inflation, $22B. That was just to blow something up. Had the United States and Great Britain not subsequently developed nuclear energy, it is likely it never would have been developed because it would have been impossible due to cost. Only a nation can develop nuclear energy, it is not something that could have been developed privately, again, due to cost.
When factoring the cost of the energy produced by nuclear fission, the cost of that electricity, the cost of the development of nuclear energy is never factored in. If it were, it would be clear there has never been a more expensive way to generate electricity than nuclear fission. Nuclear energy development was a freebee, and the biggest freebee in the history of civilization: nuclear energy development, paid for by tax payers, was given away and the tax payers' investment never had one penny of return. "Electricity too cheap to meter," never materialized, and not even close. The tax payers were bamboozled.
When all is said and done, the treehuggers have the weaker argument. Sure, nuclear energy has some environmental concerns, but these kinds of arguments pale in comparison to the economic argument: nuclear energy has never been and will never be economically viable. There are reasons, but they can be ignored, because we can see the result, no investor will touch nuclear. And complaining about the type of nuclear plants serves no purpose because the fission plants we built are the cheapest designs there are. Seeder reactors would be cool, but, you see, they're even less economically viable than the fission steam turbine plants.
We could invest everything, every dollar earned, every possible value society could produce, into nuclear energy development, but even if we did, nuclear energy would remain economically unviable.
Again, if we could just invest a small portion of what we wasted on nuclear energy development into solar energy development... well jusT look at how cheap solar has gotten in 20 years with private development. Imagine if it was 80 years instead of 20 and included massive, mind-boggling, government subsidies. Forget any government investment in solar, if solar subsidies could merely match nuclear subsidies, dollar for dollar, no one would be talking about nuclear power anymore.
Agreed ... but I'll posit the existence of technical 'treehuggers' who knew it early on (and kept quiet after what happened to Oppenheimer.)
Further posit: they were aware that nukes were more about making Pu than heat. No insurance company back in the day would accept the risk either (thus the 1957 'Price-Anderson Nuclear Industries Indemnity Act').
Anyway, I'll quote a 1951 expert:
"t is safe to say ... that atomic power is not the means by which man will for the first time emancipate himself economically.… At present, atomic power presents an exceptionally costly and inconvenient means of obtaining energy which can be extracted much more economically from conventional fuels.… This is expensive power, not cheap power as the public has been led to believe." — C. G. Suits, Director of Research, General Electric, who who was operating the Hanford reactors. [https://www.ieer.org/pubs/atomicmyths.html]
0 comments after 20 minutes on an energy post? How am I supposed to know why this won't work, won't be useful, won't be cost effective, won't scale, and that it's just a fad?
Jokes aside, this seems impressive, I have no idea what the best applications would be but wikipedia claims that current similar devices have fairly bad efficiency(https://en.wikipedia.org/wiki/Thermoelectric_generator).
This is not similar to anything on that page, as it operates on temperatures of thousands of degrees. The comparison with steam engines is also quite bad, as the Carnot efficiency on that kind of temperature difference is way above 90%.
So, it's just an overrating research PR piece, like the ones people like to complain. This thing probably scales just fine, and may be quite useful. The entire problem is that science gets divulged on those insane PR pieces where it's compared to completely different things, or promise completely impossible results.
At those temperatures, Carnot efficiency is between 80 and 90 percent.
The comparison to steam engines is misleading, but there’s an important distinction. Steam or gas turbines would reach very high efficiencies at these temperatures too, but won’t because of material properties and limitations thereof.
These limitations don’t seem to exist for this new technology. Hence, reaching very high efficiencies becomes possible. In theory… In practice, I don’t see how heat sources with temperatures that high are feasible or could stem from renewable sources. (something with the thermal battery? Wasn’t explained much in the article)
In any case, in comparison to steam turbines, the technology presented here does absolutely nothing in terms of decarbonising the grid, as claimed. It’s just potentially more efficient. But what’s the source for the primary energy?
I think the (unstated) idea is to use an arc furnace, during peak solar/wind output to heat graphite and recover the energy later using their fancy new TPV cell. That's going to require some really good insulation, since your heat source is intermittent and your temperature difference is huge.
My first thought was let's use it in fission (and later fusion) reactors.
Basically, in their words, "a large insulated shoebox full of brick". And I could be wrong, but I think you should be able to scale the amount of "brick" up to whatever size and keep the insulation the same thickness, so the storage capacity would increase by the cube of the scale and the amount of insulation would only increase by the square of the scale.
That would allow you to minimize the fluctuation in temperature - i.e. if it takes 10 days to get up to temperature, because it's big, you don't have to cool it all the way back to room temperature when you take an afternoon worth of energy back out.
With unmoderated fast neutrons, and critical geometry your startup will be always exceptionally close to booming and taking over a huge flank of the market.
For non-mobile storage, it seem that the waste heat (from cooling the TPV) would still be at such a high temperature that it can be used for co-generation to improve total system efficiency. Do existing technologies exists to make optimal use of this "temperature bandgap"? Would direct to steam work efficiently?
The actual generator has no moving parts. The "tanks" for storing the heat can be made from graphite, but the thermal battery made by combining the storage tank with the generator that they propose has to pump heat around using liquid tin (or perhaps liquid silicon) as the working fluid, at temperatures up to 2400C. That requires not just moving parts, but some pretty far out engineering. All of the metals we commonly build things like pumps out of are liquid at those temperatures, after all. And of course, you want pumps that run reliably for years in that hostile environment.
Of all things keeping the energy transition back, steam turbine manufacturing is probably very low on the list. I’m not aware it’s an issue. It’s an old and proven technology.
> the Carnot efficiency on that kind of temperature difference is way above 90%.
This is definitely not my area, but is Carnot efficiency directly comparable to the efficiency numbers cited in the article? Or is the "work" in Carnot efficiency the mechanical work, prior to being converted to electricity?
Yes, it's directly comparable. You can interconvert mechanical work and electricity freely; electricity isn't like heat. Everyday machinery does it with 95% efficiency, but there's no fundamental limit.
This device doesn't really change the energy landscape. Let's rephrase the title: "A new heat engine is as efficient as a steam engine but needs a thermal source 1,800 degrees celsius hotter to work". The device described in the article is interesting in that it has no moving parts and might have an application one something like a nuclear powered spacecraft. Actually trying to harvest energy from TEGs is exceptionally difficult, since renewable energy sources aren't nearly as energy dense as thermal fuels like hydrocarbons or fission. The thermal gradients produced through renewable sources like solar are tiny [1]. It could be used for something like geothermal power, but again it needs temperatures way hotter than conventional steam engines which already work fine for geothermal energy production.
> 54 metres (177 ft) high and 48 metres (157 ft) wide
> more than 2,500 h/year [sunlight]
> peak power of 3200 kW
> Temperatures above 2,500 °C (4,530 °F)
Sounds like it could be useful as a "default load" inside an otherwise inactive solar furnace at least.
You're describing solar thermal energy [1]. Use solar collectors to turn light into heat, then use a heat engine to turn that heat into electricity. This TEG could be used as a heat engine for this task. But again, our heat engines are already capable of this task and don't need such high temperatures. A solar collector array even getting to this TEG's operating temperature might not be feasible.
Photovoltaics just turn solar energy into electricity, and don't need the heat engine. This has made them way cheaper to deploy than solar thermal energy. So unless there's something very important about this new TEG, the solar thermal vs photovoltaic calculus doesn't really change.
Right. But we already have that technology with conventional heat engines which have the advantage of much, much, lower operating temperature requirements. If you have a 3,000 degree vat of thermal storage material this new engine stops working after draining 1000 degrees. Existing heat engines can usually work down to several hundred Celsius - though superheated steam engines need around 700 Celsius. But that's still an extra 1000 degrees you can bring it down, even in the conservative case.
Compared to other TEGs. Not compared to steam turbines. The article is actually being very generous in saying it's "as efficient" as steam turbines. Steam turbines are more efficient with scale, and industrial ones for power generation are over 90% efficient [1]. This new TEG's efficiency is "around 40 percent". Higher than the previous TEGs in the 25-35% range. But not compared to steam engines, that also benefit from much lower operating temperatures.
1. Multistage (moderate to high
pressure ratio) steam turbines have thermodynamic efficiencies that vary from 65 percent for very small
(under 1,000 kW) units to over 90 percent for large industrial and utility sized units.
RTGs do not get anywhere close to 1800 degrees Celsius. Even if they did, it wouldn’t be a game changer, because you can make up for loss of efficiency with a bigger RTG.
>why this won't work, won't be useful, won't be cost effective, won't scale
Not an expert, but reading this a few negatives popped out. Basically they are heating a black body to 2400C and then making electricity from gathering the emitted light in a cell. They get to pick a temperature to match the bandgap of the cell.
The key problem is getting something that hot without using another (lossy) form of power. The Sun's surface is ~5600C so that's enough headroom to get there from solar. That's cool. But are there any fission reactors that get (or could get) that ridiculously hot?
"The team’s design can generate electricity from a heat source of between 1,900 to 2,400 degrees Celsius"
That's way up there. That's well above the melting point of steel.
That's above the highest temperature jet engines made for experimental aircraft.[1] Most jet engines try for exhaust gas temperatures around 600C or so, for a long useful life. Typical nuclear reactors, around 300C.
It's not impossible to operate up at those temperatures. Every steel plant does it. There are ceramic and brick materials that can deal with such temperatures.[2] The storage medium would probably be some molten metal.
This seems way too much trouble just to store energy.
Now if this thing worked at 600C or so, there would be more uses.
I'm a little unclear on the concept. You say storage medium, as if the graphite is expected to remain at very high temperature for an extended period. Preventing a chunk of super hot graphite from cooling is, if anything, an even larger engineering challenge than getting it that hot in the first place!
No, they claim no moving parts for the generator, but refer to engineering designs that use pumped liquid tin to move the heat within the system. Clearly the challenge there is building a pump that can handle liquid tin at 2400C.
I could be wrong, but I don't think you'll find EM pumps in use for any liquid metals at temperatures above about 1000C. The 2400C temperatures required for this thermal battery concept present a significant materials engineering challenge for anything that touches the heat storage medium. There just aren't many mechanically sound, or electromagnetically capable solid phase engineering materials to work with at that temperature.
I'm not saying what they describe can't be done, only that getting the photovoltaic part to work isn't the biggest engineering challenge the concept faces.
I think the sheer size would make it interesting. The heat energy potential of acres full of graphite is enormous and presumably much cheaper to construct than an energy equivalent battery. Now I wonder how it holds up to other methods of storing energy.
Biggest key problem imho is how they expect to store this heat energy. It looks like this cell will, like a PV cell, constantly be absorbing photons. If those photons aren't creating electricity/voltage across a gap then they are being converted into heat. So to keep the medium at temperature you would need to insulate it, to wrap it in a mirror, only exposing the flux to the energy-collecting cell as needed. That means moving parts.
I think most try to keep temperatures under 1000C. I think many FAST reactor designs are looking at 600C operating temps with peak temp reaching maybe 1200C during emergency testing. But my memory might be wrong.
Not mentioned in the article is power density. How quickly can the energy be released? Consider solar panels, you need a table sized cell to get 100W. That can make for a big battery to get grid scale power output if these cells are only as power dense as solar panels. The energy density of a heat based solution can be very high- metals can get very hot and they are dense enough to store a lot of energy. But if you can’t get the energy out of the battery fast enough that limits the applications. By comparison lithium ion batteries can dump power out extremely quickly, which is what makes them great for cars. Hydro is even better.
The article in Nature quotes an energy density of 2.38 w/cm^2. Which means a Gw battery would require 10e5 m^2 of absorber surface, exposed directly and at close range to the radiation from molten metal (which is the heat transfer fluid they propose). It has to be direct, and at close range, because the efficiency they quote relies on the absorber reflecting non-absorbed photons directly back into the emitter, where they are re-absorbed as heat and potentially re-emitted.
That's about 25 acres of absorber, and an implied 25 acre surface area of the liquid metal emitter pool.
There is a basic challenge here to the design - the energy storage density for the thermal battery they envision scales as the cube of the characteristic dimension of the plant, but the power density that can be delivered scales only as the square of dimension. Not saying that can't be dealt with in engineering, but it ain't going to make this easier or cheaper.
Surface area is relevant for solar because the sun is so far away. A local heat source allows you to surround it with 3D shapes not just a flat plain.
As to temperature this thing is for very high temperatures: can generate electricity from a heat source of between 1,900 to 2,400 degrees Celsius. At 40% efficient you need a wide temperature difference which would suggest a high energy density.
This design is photovoltaics, just like solar, but optimized for infrared photons. There is no avoiding the reality that energy storage density will scale as the cube of the facility size, but power density only as the square. And at 2.38 w/cm^2, the scale coefficient is not all that great.
Picture a stack of flat panels with each layer consisting of: (Cold)(Panel)(Hot)(Panel)(Cold) held vertically. So: (Cold)(Panel)(Hot)(Panel)(Cold)(Panel)(Hot)(Panel)(Cold)(Panel)(Hot)
Now you add hot gas at the bottom and have say 4 layers per m. So a 3mx3mx3m cube would be 4 layers * 2 panels per layer * 3m * 3m * 3m = 216m2 of panels taking up a 3mx3m section of floor. At 2.38w/cm2 * 10000cm2/m2 * 216m2 = 5.14 MW of power.
For long term energy banking and if we can get them working, flow batteries seem vastly superior to all alternatives, by scaling storage with regards to tank volume. Instead of some difficult-to-manufacture structure.
I think their application is grid scale and you can scale across hundreds of batteries to provide the throughput you need. I don't know how I feel about having a small molten ball of metal inside the hood of my car. Turns my car into the most dangerous gusher in the case of an accident (for those who aren't familiar, gushers are a gummy like candy with juice inside).
High temperature is the point. The efficiency of heat engines depend on the temperature difference (relative to ambient). The hotter you can go, the better. (granted this thing isn’t really an “engine”, but the trend still applies)
A Carnot heat engine operating between, say, 2600K and 400K can reach almost 85% efficiency.
The higher the temperature, the higher the share of exergy in the heat flux. At high enough temperatures, it’s no longer a feat to convert to electricity at high efficiencies.
It would be pretty hard to create a carnot heat engine that can withstand 2600k. I'm not even sure if for example Tungsten has structurally integrity at that point.
A turbine system would have much better efficiency than 40% if its heat was available at that temperature. For example a closed Brayton cycle gas turbine + steam turbine system. Certainly complex and expensive, but could get much better than 40% efficiency.