There are lots of cheap ways of storing heat... The capacity cost of thermal storage has not been the bottleneck for commercialization, but rather the round trip efficiency. This seems academically interesting, but I doubt it will be much more than that.
To a first order approximation, the cost of power from a storage system is going to be $/kWh_source_power/efficiency+ammortized cost of capital. No matter how cheap the second term is (and it can't be that cheap unless someone invents a 10x cheaper turbine), the first term tends to make storage uneconomical unless your RTE is above ~80%. This will get ~60% at best.
I think you're right to the losses, but didn't consider the surplus energy side of things. We now have an oversupply/underdemand problem for sufficiently periods of the day that time shifting motives by the energy producer outweigh the otherwise brutal economics: they can't send the energy anywhere else because of transmission system constraints. Then need cheap ways to store the energy and loss is probably acceptable for the right outcome: better if you can avoid it, but acceptable.
Arguably batteries are ideal for the problem. However thermal storage systems like this one have good scaling to storage side of things: insulation is cheap, adding tanks is cheap. The basic turbine cost can supply wallplate stated power, the tankage gives duration and you can add as many tanks as you have room for.
Alongside simply supplying the energy, batteries are great for maintenance of supply frequency but I don't understand how they do reactive power. Turbine based systems do reactive power really well because of the rotational torque (I think) so represent a distinct role in electricity supply stability. (Its entirely possible batteries can do this too)
Finally, repurposing of steam turbine based generation sites means they have transmission systems, turbines on site and to hand: the capex component can be less.
You can do reactive power with some batteries since they are all behind inverters that can be designed to do that. The Tesla grid batteries (the big ones, don't know about power walls) can certainly do it via API commands (you can specify real and reactive power components, and frequency control, I'm not sure if they can all 3 be run at once though). I'm not sure if all battery companies have this feature though. I used to control several dozen sites with Tesla batteries via their battery control API, though we never used reactive power or frequency control.
We still need storage solutions like this for long term storage, though, as batteries are mainly used for load shifting and short term arbitrage right now, at 6 hours of discharge and less.
That's good. Condensers and like appeal to my steampunk sensibility but they aren't magic or instant. If batteries can do reactive quickly and as well as they do FCAS then it's a huge plus. Getting more than 6 hours is a good point too. I'm sure battery tech is going to include scaling to meet that, this kind of heat storage might just be about time to market.
Inverters are far better than anything rotating-mass-based for control over waveform. At least if you want to, that is. You just take a (possibly GPS-disciplined) quartz oscillator and have no issues hitting 50ppm frequency stability at lower phase noise than can get through the system at 50/60Hz.
They can essentially do this instantly, i.e., substantially sub-wavelength of your line frequency (a PWM of 10 kHz isn't even high for an inverter, and 100 kHz is well within what's feasible for large systems operating at a couple hundred volts; 10 kHz would equate a light-speed delay of 30km, 100 kHz of just 3km, which gets into the territory of relativistic w.r.t. medium voltage rings in cities).
Generally solved by mixing sources, wind, transmission and storage. People freak at the overcapacity build cost a bit.
Hydro systems and pumped storage look useful in this but the extent of summer winter variance is very specific to latitude and much of the planet isn't as badly affected. Wear transmission losses, buy power from lower lattitudes
> batteries are great for maintenance of supply frequency but I don't understand how they do reactive power.
Batteries output DC. They need an inverter to be connected to the grid. The inverter needs to choose a phase angle at which to output current and voltage, through something that looks suspiciously like a boost converter. This doesn't have to be quite the same phase as the line, it can "lead" or "lag" in an attempt to push the line phase angle back to what it's supposed to be.
> The capacity cost of thermal storage has not been the bottleneck for commercialization, but rather the round trip efficiency.
Has there been any effort to do this directly at power plants, so there isn't a round trip?
You have a heat source that generates the same heat 24/7 (e.g. nuclear reactor). You have solar panels that generate power only during the day. So you put the heat from the heat source into storage while solar is generating, then take it back out after the sun goes down.
You're not going from electricity to heat back to electricity, you're just not converting the original heat into electricity until later in the day.
Curtailed renewables might make this economical in the short term, but also provides the same benefit for competitive storage technologies (e.g. LiFePO4 batteries). You'd have to compare the LCOE for each over the lifetime of the installation, and I don't think thermal would stack up well.
Thoughts about the iron flow batteries from ESS? They're using an iron flow chemistry that doesn't require exotic stuff. Looks like they've been testing for a while at a bunch of sites, and brought their first 'real' non-test customer online this fall.
On paper to a layman, they look perfect for at least some applications. Very cheap and easy to build the "stacks", almost completely nontoxic/safe, and relatively easy to scale (I think they don't scale indefinitely with liquid tank size...that the plate surface area also matters?)
> scale indefinitely with liquid tank size...that the plate surface area also matters?)
For flow batteries, the plate area controls the charge/discharge rate and the tank size controls the capacity. In theory you can scale the capacity as far as you can build tanks; I don't know whether there are practical considerations of "self-discharge" or electrolyte degradation.
Renewable energy has pushed the spot market prices for electricity into the negative last year in germany and UK. Soaking up that energy and reselling it later may be viable at low efficiency.
Why would the RTE of this salt solution be so much lower than the molten salts demonstrated in the DOE Solar Two project? According to the technical report at https://www.osti.gov/biblio/791898 they achieved round-trip energy storage efficiencies of 97%.
That is probably quoting it on a heat-basis. Only 3% of heat might be lost from storage, but from the 97% remaining you can only ever convert ~60%[0] of that thermal energy into useful electricity.
[0]Rough #. Carnot efficiency would define your max theoretical for a heat-engine process.
Power-cycle Carnot efficiencies are far more likely in the 20--40% range.
Efficiencies above 50% are claimed by some dual-cycle plants (gas turbines with steam generation), and with cogeneration utilising waste heat, higher yet, but the electrical generation component remains in the 40--50% range.
Lower Heating Value, which ignores the latent heat of vaporization of the water produced by the combustion when computing the energy content of the fuel. Higher Heating Value, which includes that, would give a lower efficiency.
I was presuming resistive heating - a heat pump does make more sense. That said a practical heat pump with even a 1-way efficiency of 97% seems unlikely.
If you have an ideal Carnot heat pump, it is fully reversible.
For example, I can use 10 watts of electricity to remove 10 watts of heat from a cold place and add 20 watts of heat to a hot place.
Later, I can apply the same in reverse. I can move 20 watts of heat from the same hot place, heating the cold place by 10 watts, while generating 10 watts of electricity.
The use of the phrase "efficiency" while discussing ideal heat pumps is misleading - no energy is ever wasted or lost, the law merely limits the ratio of electrical Vs heat energy you get in/out.
So ... yes, the heat pump can be run in either direction, and is "reversible" in the sense that the heat flow can be reversed through the pump.
The process is not thermodynamically reversible in the sense that there's a net energy expenditure either way. No free lunches.
What a heat pump can do is to move a greater amount of energy than it uses to move it. Measurements of this include COP (coefficient of performance), EER (energy efficiency ratio), and SEER or ESEER ((European) seasonal energy ratio).
The COP is a direct measure of energy moved divided by energy input, and ranges from about 2--4, with typical ground-loop heat pumps achieving scores around 3--3.5. That is they move three times the heat energy that they run under. It's as if a furnace output three times more heat energy than fuel input.
In your example, we'd first want to correct power (watts) to energy (watt-hours or joules). But 10 watt-hours of electricity would move about 30 watt-hours worth of heat, for a typical heat pump. If you're heating, you'd dump the additional 10 watt-hours into the heated space, if you were cooling, you'd want the heat pump's own heat to be directed to the external environment.
Discussion of heat gained/wasted with heat pumps is ... complicated. Note that effectiveness in moving heat decreases as the differential being moved against increases. That is, if you're trying to cool a space in an environment that's already warm, you're pumping heat "uphill". Similarly, if you're trying to warm a space from a cold environment, there's not much external heat to extract.
Ground-loop heat pumps benefit by the fact that ground temperatures tend to be more moderate than outdoor ambient air temperatures, so the temperature gradients are more favourable and predictable.
That is what the report says. I think there must be some weasley definition of roundtrip efficiency happening as that exceeds the theoretical maximum given their hot and cold reservoir temps.
Do people talk about storing heat in houses in cold climate regions where there would be no efficiency losses? I always wondered if you could store solar panel energy in large cubes of firebricks that you could dissipate to heat your house. There seems to be some commercial systems (eg. by Steffes) but they are expensive, offer just an overnight amount of energy and seem to be usually used to smooth grid consumption. Why isn't there a version of this designed to store solar energy for longer periods in the winter when insolation is low?
A quick back of the envelope calculation:
A 6.5 feet by 6.5 feet cube (2mx2m) could store about 4500kWh at 1000C, this seems close to energy need for heating for more than a month ( Energy density of fireclay from here: https://www.engineeringtoolbox.com/sensible-heat-storage-d_1... ).
That's a very large amount of bricks and a potentially dangerously high temperature but still...
Pottery kiln manufacturers might be able to design this? Kilns go to higher temp than this and are used indoors.
Right. The ordinary firebrick system with separate heaters would be limited to about 1000 C, although I've seen a concept using sand flowing through a heater that would reach 1200 C.
In cold climate regions solar power can't heat houses.
If talking about individual houses, not condos, most economic solution is a two-stage water heating powered by a wood pellet or gas burner, where the second stage has a 3-5 ton water tank as a heat storage, so that the burner doesn't need to run continiously. This is a proven, simple, maintainable and safe solution.
We're talking ~50-150kW thermal power here. Solar can not reach that.
Several cubic meters of almost-boiling water is hazardous enough.
Keeping a highly corrosive liquid at 700C in your own home.. no, thanks.
Passive solar heating requires loads of storage, and often active systems to transfer heat.
But there’s a much easier, more economical way to use solar power for heating in cold climates: heat pumps! PV panels give you electricity and then you get 2-3X the power for heating. There are economies of scale in the supply chain, low CapEx, and not a lot of plumbing.
yeah, maybe solar plus geothermal heat pump would be the most advantageous system right now. Although you can't store and there isn't much solar in December and January here.
They pump heat into the ground in the summer for air conditioning and pump it back out in the winter for heating.
You don't need to be "much", you need to ideally match the TCO of other fuels (or come close, if you care more about reducing fossil fuel use and CO2, which is arguably more important than TCO. People obsess about heating costs but then run out and buy $60,000 luxury cars to do an hour+ commute each day in.)
Not all cold climates lack sun. Here at 6-10,000' in northern New Mexico, we still get lots of sun in the winter, but also very cold (overnight) temperatures. Air source heat pumps make more sense in almost every way, and will operate down to -13F/-25C.
Our 6.6kW array can provide roughly 1/3 the electricity we need to heat all of our 1875 adobe home, which is poorly insulated at best. If this was closer to passivhaus standards, I imagine it could get close to 100%.
Yeah, as I just mentioned adjacent to this, our 6.6kW array produces 1/3 of what we need during the winter, but 3x what we need in the summer. Last year (the first full year), we generated 93% of our electricity use over the full year. We heat solely with air source heat pumps.
What is the benefit of storing the heat of wood pellet or gas? Can't you just use those on demand?
Also hot firebricks was my proposed solution. Water has comparatively low energy density and I wouldn't want the corrosive liquid either.
I understand that there isn't a ton of sun in winter months but maybe you could over provision your PV, store summer or fall solar energy and top up with whatever you can get from the sun in the winter?
> What is the benefit of storing the heat of wood pellet or gas? Can't you just use those on demand?
Burners scale badly to lower power. Also, automatic pellet feeders are prone to breaking down, which you absolutely do not want to happen when it's -15C on the outside.
Thus a thermal buffer is needed.
> Also hot firebricks was my proposed solution. Water has comparatively low energy density
Water actually has about four times more energy density than bricks.
> I understand that there isn't a ton of sun in winter months but maybe you could over provision your PV
Prohibitevily expensive in both money and area needed
I think cases where you use the round-trip losses to heat things could be an exception. That’s also what the article says in its first paragraph: “If you want both heat and power — for a low price — there have been few good options”
Say you’re living up north in Norway. Plenty of sun in summer, very little of it in winter plus guaranteed heat and electricity demand in winter.
Say a tank survives for XXX years before corrosion demands the entire thing needs to be scrapped. Wouldn't that still be an economically attractive option for storing surplus solar energy during the day? Or is the daily upkeep of the installation such that there are significant recurring costs that dwarf the initial capital investment?
What do you think of TCES, in particular for low-grade heat and cooling at the household level, where its lack of reservoir insulation matters in a way that it doesn't matter at utility scale? Is there a good review article you'd recommend reading?
NREL has some decent material on LCOE calculations.
Amortized capital cost would be a function of upfront cost, lifetime (years), cycles (and per cycle efficiency over time), capacity factor (e.g. average drawdown of the batteries per cycle), and discount rate.
If one has a software background, it might be easier to just to simulate 10 years of cycles to figure out how many kWhs come out of the battery bank per year. Adjust by a discount factor (cycles today pay more in real dollars than cycles in 10 years), and divide the upfront cost by that number.
It sounds like if you could heat things to a higher temperature while keeping costs the same or slightly higher it would probably improve things then? That's useful to know.
It seems to be an unfortunate property of nature that some of the most-useful chemical and physical systems are also some of the least-friendly.
NaOH, in large quantities, is an unforgiving material. As I've read my way into carbon-capture technology, I was saddened to find it there, too. HF also comes to mind in this direction (but is in another league). There's a correlation between those properties that make them useful and those that make them dangerous.
The most powerful counterexample to this notion that comes to mind, though, is water. What a spectacular chemical, one that is so useful we have built ourselves from it, yet we can also swim in it for hours without concern for any adverse effect beyond temporarily-wrinkly skin. (While it's less-relevant for thermal solar, it's darn good at heat-capacity, too! :) )
I mean, water is also aggressively hostile. We're just really well adapted to it. Even pure water will turn metal into rust unless you're using special alloys. Salt water is so bad that even alloys won't save you, you have to depend on sacrificial anodes and exotic surface coatings. Steam takes everything bad about water and jams the fast-forward button until the tape catches fire. There's a reason the first hundred years of steam power is basically just a timeline of scientific metallurgy and gasket technology.
NaOH is merely "ordinarily dangerous", i.e. safe enough to be sold to households with a warning sticker on. As long as you don't get it on your skin in large quantities you're fine. You can fairly easily dilute it into harmlessness, although you then have to be careful not to kill a river. If you leave it for long enough it will slowly react with CO2 out of the air via carbonic acid, precipitating out as harmless sodium carbonate.
It doesn't have any enduring toxicity. It won't get into your bones or your genetics. It's not bioaccumulative.
It’s dangerous because you’re also partially made of significant amounts of fats (and other things), which gets very very painful and can make it somewhat difficult to continue living when they get turned to soap while still part of you.
NaOH is mostly dangerous because it is so alkaline. When reacted with fat to create soap the alkaline OH- is reacts with the fat and is 'used up', leaving a mostly neutral product
This link here https://aris.iaea.org/PDF/MSTW.pdf seems to provide a little bit more info on how they manage to keep the corrosion within acceptable parameters.
It involves the addition of an unnamed chemical compound functioning as a reducing anode straight into the fuel, which is a fairly awesome thing I wasn't aware was possible.
Isn't the point that a patented process is documented as part of the patent application? Industrial processes and materials seem like a great use of patents from my POV. Psychology and biology I'm more concerned with.
Sounds promising. I work in metal cladding, and there is significant interest in molten salt resistant coatings. It doesn't sound like this solution would work for molten salt nuclear reactors, which is a shame because they could be revolutionary if perfected.
IIRC the seaborg reactor design uses a "traditional" flouride based salt for the fuel salt. The NaOH is a separate circuit used as the moderator, instead of a graphite block like other thermal spectrum molten salt reactors. Being a liquid, this sidesteps the problem of radiation damage to the graphite with associated swelling, cracking etc. etc.
Well, then the question is, what do you have in between the tanks? Is it red hot too or not? If it's not red hot, you're losing a bunch of heat through the extra surface area. But if it is, now you have a bunch more material that's annoyingly reactive (because it's red hot) and may decide to melt down into a pool of red-hot material on the ground, letting the lye tanks cool down and possibly spill.
What does this add over existing [1] solar thermal plants? The approach seems to be roughly the same: heat a material with solar power and then operate a heat engine with it. Is storing heat generated by photovoltaics better than using solar energy to heat material directly?
When you're dealing with solar (or any renewable really), you have to think about the generation of electricity and storage of that electricity.
Solar thermal plants can be used to generate energy, but they're not great at storing it for dispatch at later times. Even the link you provided has thermal storage as a component of the system-- you need to figure out how to dispatch electricity at night.
What's proposed is a cheaper storage solution, and they state it can still provide power after 10 hours.
Solar thermal inherently has built in storage: the molten salt heated during the day continues to provide energy through the night.
> Even the link you provided has thermal storage as a component of the system-- you need to figure out how to dispatch electricity at night.
This is referring to the thermal storage of the molten salt tower.
What I'm getting at is why is solar PV -> electricity -> thermal storage preferable to solar mirrors -> thermal storage. Is solar PV less expensive than mirrors?
Solar PV performs better under imperfect illumination conditions. Concentrating solar thermal power can't make use of diffuse light on partly cloudy days, whereas solar PV can. Solar thermal also has to reach a minimum operating temperature before it starts generating steam/electricity. PV actually works better on bright, cold days and will generate full power the instant an array reaches full illumination. These factors, plus the greater mechanical complexity and maintenance requirements for thermal solar, make it very hard for new thermal solar plants to deliver a lower levelized cost of energy than new PV plants.
A US company, Ambri, is also working on a molten salt battery.
"The liquid metal battery is comprised of a liquid calcium alloy anode, a molten salt electrolyte and a cathode comprised of solid particles of antimony"
If you have a renewable energy source in the form of heat, storing it in chemical bonds could be a better alternative in terms of long term energy storage.
molten salt loses approx. 50% of it's energy after 14 days, while hexametaphosphate for example, when kept as salt, can be stored probably for decades.
Hexametaphosphate can be produced from cheap monophosphate salt by heating it in solution to 700c and can be kept in room temperature until it is utilized.
The energy density of hexametaphosphate is around ~ 80kWh , which higher or the same as most molten salt heat storages available.
Disclaimer:
My company (Enzymit) develops enzymes that use polyphosphates, rather than ATP as an energy source to catalyze many types of reactions.
> we cannot let it defocus Seaborg’s current mission to power regions with poor or no access to renewable energy sources through our compact molten salt reactor
Are there a lot of regions where neither solar nor wind is viable but (their new version of) nuclear is?
Solar and wind do need storage, but if their levelized cost of energy is low enough they still beat nuclear. Remember not to use batteries for annual load leveling, which some dishonest analysts try to do.
For seasonal storage and rare event backup: hydrogen burned in a combined cycle plant. Hydrogen can be stored underground for maybe $1/kWh of storage capacity.
High voltage dc interconnects? Importing Danish wind energy is one of the projects supposed to keep the lights on after an exit from nuclear in Belgium.
If you want to use heat for storage liquid metal batteries might be better. (?)(Ambri)
From what I understand they have ~80% round-trip efficiency, functionally unlimited lifespan and no moving parts required. Electrons in/Electrons out.
Not really seeing much here in the way of details, which always makes me worry;
Are they using rare-earth based catalysts to control some chemical reaction?
It's corrosive so will always break down parts, so have they extended the life by 10x or 1,000,000x ?
Storing heat in this medium leads to thermal losses I'm not seeing a mock-up of how it's proposed that this is minimised.
Surely trying to store high-energy corrosive materials at high-temp is dangerous even at this modular scale?
Naively this sounds like their solution to "saving the world" is "boiling acid" which is at least novel...
It's everywhere nowadays and it makes me wince every time I run into it. And for whatever reason it's especially annoying when I can feel it building up to it, like here which is basically the text book example.
Okay, with that off my chest, this technology seems on the surface pretty cool and I certainly hope it works out.
For years, people wrote sentences designed to build tension, create interest and promote the general welfare. Nobody could see a way out of this wince-inducing methodology and scientists estimate that 389 million person-hours of useful waking time are destroyed every year by this irritating "thunder sentence" structure. UNTIL NOW.
NaOH is lye ... relatively nasty stuff (for metal and skin). (MP 323°C) I've read that 'Glauber's salt' (Na2SO4.10H2O<->Na2SO4) (MP 32.4°C) is good for heat storage. Sounds like a better idea for residential applications.
This is nothing new. Molten salt ideas have been around 30+ years and there are projects in Spain and other countries that tested them on small scale. The economics is still poor.
“If we filled up a building the size of the Colosseum in Rome with the salt and heated it to 700 degrees, we would actually” … have a whole lot of pissed off Italians on our hands
There are lots of cheap ways of storing heat... The capacity cost of thermal storage has not been the bottleneck for commercialization, but rather the round trip efficiency. This seems academically interesting, but I doubt it will be much more than that.
To a first order approximation, the cost of power from a storage system is going to be $/kWh_source_power/efficiency+ammortized cost of capital. No matter how cheap the second term is (and it can't be that cheap unless someone invents a 10x cheaper turbine), the first term tends to make storage uneconomical unless your RTE is above ~80%. This will get ~60% at best.
(of course, I could always be wrong)