Hacker News Reader

Comment by radu_floricica | Hacker News Reader

Just to spell out something that's been implicit in other comments: the entropy level of your energy matters a lot. Electricity > movement > heat. To go down the chain is almost free, to go up the chain you have to spend quite a lot.

You can do a lot of things with electricity: you can heat things, but also move them around and run your TV, all without any loss. With heat you can just... heat things. So you can't call this an "iron battery", because you don't get electricity out of it, just heat. Maybe call it a "heat battery" or "high performance heat pad".

Also note the efficiency numbers: "High-efficiency electrolysis of iron oxide can store as much as 80 percent of your input energy in the iron fuel" is the efficiency of the process itself. "Using this kind of cyclical process to generate electricity could approach a theoretical efficiency around 40 percent" is because you need to climb the ladder to low entropy again (probably by using the equivalent of a steam engine to run a generator).

View on HN

The most important take away from this is that electrical energy can be easily "multiplied" into thermal energy by using it to move around the almost unlimited thermal energy contained in the environment around us.

For example, a water-water heat pump can easily boost the temperature of your reactor to 80 deg C by using only 1 KWh of electricity for every 4KWh of heat delivered, the rest coming from a nearby large body of water and slowly recovered from the environment throughout the year. So we are looking at on order of magnitude more heat for the same electrical consumption 40% vs 400% efficiency. Instantanous renewable energy prices are low, but rarely that low.

The limit of efficiency is dictated by the boost in temperature needed, so you can't use heat pumps to get the 1000C generated by iron oxidation. There is certainly a niche for industrial processes requiring very high temperatures, but I can't imagine why on earth would a brewery need such temperatures.

But this is not a takeaway that’s specific to this process, that’s how ubiquitous heat pumps work in residential HVAC.

The article takeaway is that iron powder is a surprising electricity—>heat battery.

The other way to put it is that your efficiency is proportional to 1/(delta T). In other words if your heat pump is working between two environments with very close temperatures then the heat transfer is very efficient (for example 19 degrees C outside, 21 inside). But if you are trying to heat your house to a livable temperature while it’s -20 outside, your efficiency is terrible.

This is why for example you see heat pumps installed all over the place in homes in the southern parts of the US where you also need little heating all around, while places up north where you do spend quite a bit on heating cannot really take advantage of heat pumps due to winters being too cold.

> But if you are trying to heat your house to a livable temperature while it’s -20 outside, your efficiency is terrible.

Amazingly (to me), there are now air source heat pumps that stay > 100% efficient down to -20C before regular electric heat takes over:

https://www.nordicghp.com/2017/01/heat-pump-effective-temper...

Though apparently you'd be better off with a ground source heat pump if your average temperature during heating season was below freezing or so.

[Edit: I mention this because I've heard that "you don't see heat pumps in the north" was true in the past but is now outdated -- they've made sense in warmer climates for a while but have only recently crossed over for colder climates, so it will take a while until they're common.]

> heat pumps that stay > 100% efficient down

Note: there are two popular "efficiency" measures, and its important to realize that they're incompatible.

Air conditioners are commonly "energy moved / energy used", which can reach greater than 100%. If you move 150W of heat using only 100W of electricity, you have 150% "efficiency".

---------

I don't know the term for the other kind of efficiency (I'm not a physicist), but lets call it "inverse engine efficiency". This is "energy out / energy in", which ends up being pretty close to "energy moved / (energy moved+electricity used)"

Under this measurement of efficiency, 150W moved with 100W of electricity is 60% efficient. This follows the more standard physics rule of thermodynamics (you can never go above 100% efficiency: it will always take some number of energy to move heat around).

Note: Car air-conditioners are funny systems. They use the heat from the combustion engine to move heat from inside the cabin to the outside world. So you are literally using heat to move other heat.

Isn't this just the coefficient of performance?

> "Note: Car air-conditioners are funny systems. They use the heat from the combustion engine to move heat from inside the cabin to the outside world. So you are literally using heat to move other heat."

Car air-conditioners are not powered by heat from the combustion engine. They are powered mechanically by a belt ("serpentine belt") connected to the engine.

Or, in some cases such as battery electric vehicles, powered electrically with the compressor turned by an integrated electric motor. The extra heat produced is just a byproduct (thermal inefficiency) of producing and transmitting that mechanical/electrical energy.

> Car air-conditioners are not powered by heat from the combustion engine. They are powered mechanically by a belt ("serpentine belt") connected to the engine.

But that belt is powered by the expansion of gas that takes place inside of a piston, due largely to the increase in heat from combusting gasoline. Ultimately, an ICE engine is a heat-engine (like a steam engine or sterling engine, but different).

Yes, but the useful energy in a combustion engine comes from that expansion of gas driving the pistons. The heat is, for the most part, just a wasted byproduct. (It can be used for ancillary purposes like heating the cabin).

> Yes, but the useful energy in a combustion engine comes from that expansion of gas driving the pistons.

Isn't that expansion of gas largely driven by the increase of heat?

https://en.wikipedia.org/wiki/Internal_combustion_engine

Wikipedia lists the ICE as a heat-engine.

------

With that being said: the octane combustion formula has 25 O2 as input and 16 CO2 + 18 H2O as output (and I assume the H2O is mostly water vapor). So that's 25 molecules of gas input -> 34 molecules of gas output.

So it seems like more "CO2 + vapor" is created than the number of input O2 molecules. But that only accounts for 34% expansion of the volume of the stroke. (25 mols input -> 34 mols output).

The rest of the stroke's power comes from the ideal-gas law: higher temperature means higher pressure and larger volume. Literally the heat generated by the chemical reaction.

SEER (seasonal energy efficiency ratio) and COP (coefficient of performance), measuring energy-moved/energy-used.

https://en.wikipedia.org/wiki/Seasonal_energy_efficiency_rat...

Are you talking about air source or ground source heat pumps?

Either. I am talking about heat pumps. The efficiency formula is exactly the same. The difference is that ground water stays at a consistent 12-15 degrees C IIRC. That is close to what we as humans like to live in, so the efficiency is relatively high heat round. In the meantime air temperature fluctuates by quite a bit. The main drawback of using ground water as your heat source/sink is that it is expensive and costly to maintain relative to a resistive coil with electricity running through it.

> "But if you are trying to heat your house to a livable temperature while it’s -20 outside, your efficiency is terrible."

Heat pumps are designed so that their "efficiency" (performance) is never less than 100% (COP >= 1). In very cold conditions, they may not be more efficient than resistive electric heating, but they will never be worse.

> "places up north where you do spend quite a bit on heating cannot really take advantage of heat pumps due to winters being too cold."

You're talking about air-source heat pumps here. In colder regions, it's better to use ground source/geothermal heat pumps. Ground temperature is consistent year-round even in very cold climates.

True. Geothermal heat pumps will always be better but given that you can get an air source heat pump for 10^3 to 10^4 dollars while a geothermal setup will cost you 10^4 to 10^5, geothermal becomes prohibitive for most residential setups.

Depends on your timeframe of investment. Your air source heat pump will need to be replaced in 7-10 years. The really expensive parts of a geothermo system will last basically forever, the other parts may need to be replaced every 7-10 years, but the costs is about the same as replacing a air source heat pump, and geothemo is more efficient in general. Thus if you live in the same house for 30 years geothermo may pay off.

Your upper limit for geothermal seems a bit high, typical installation is 20-30k. But commercial viability depends on typical outside temperature and time frame for payback

Over here in northern Europe, for a typical single-family house, replacing an oil-fired boiler with an air-water heat pump (that is, the heat pump uses outside air to heat water that can be run through the existing radiators) costs around 10k-15kEUR, replacing the boiler with a ground-source heat pump (incl. borehole drilling, the whole shebang) is around 15-25kEUR.

So the price difference is not huge. In southern parts of the country, and for older houses with less expected life left in them, air-source pumps have proved more popular, whereas further up north and for newer buildings ground-source pumps are more often the choice.

Water heater heat pumps (other than the complexity) do an excellent job of maintaining 40C temperature differentials (eg Vermont or Maine) with better than 130% wall plug efficiency. I think ideal efficiency is proportional to Tcold/dT (T in Kelvin) so it could go as high as 400% if there weren’t fluid, pump, and insulation losses.

The real challenge is reliability and cost of electricity. If your power goes out and you have heating oil you can still live for a week.

Brewing water temperatures peak at ~100C (pressurized steam is dangerous) and even with 20C water that’s an 80K difference so you’re down to 350% ideal, which probably puts you close to breakeven with pump losses. It’s the complexity that kills it.

> The real challenge is reliability and cost of electricity. If your power goes out and you have heating oil you can still live for a week.

At least over here, all oil boilers I've seen use mains electricity driven pumps, injectors etc.

>But if you are trying to heat your house to a livable temperature while it’s -20 outside, your efficiency is terrible.

For air source, in the north (at least in canada) they use ground source ones put below the frost line. So the ambient temperature down there is >0C .

Brewery’s need high temperatures simply as a way to quickly heat a large volume of liquid. Most recipes involve boiling the mixture for around an hour, and you want to reach boiling quickly. If your heat pump is only at ~120C you would need a lot of surface area which makes cleaning difficult. On top of this reaching 120C via heat pump is already inefficient and the heat pump ends up being expensive compared to resistive heating.

Breweries need multiple temperatures, and that's an area where heat pumps could still be applied. I think the common mistake is either/or solutions and you're making it here. Water has a high thermal mass. Starting with ground temperature water and heating it up greatly increases the number of BTUs.

And then after you've cooked it, you need to chill it down to around room temperature. Where you dump that heat could make a huge difference.

It takes vastly more energy to start and maintain a rolling boil for an hour than to reach 50c. If we where talking a giant scale industrial process then optimizing this stuff via heat exchangers would be a major consideration. However for a Dutch brewery located in such a cold area it’s much harder to justify such specialized investments with associated maintenance costs.

Remember this inherently allows them to load shift and thus get much lower electricity prices as well as reducing peak demand.

Yes but in a renewables world Electricity will not necessarily be ready on demand. In that case the multiplier is irrelevant. Nor may it be available locally.

With iron burning you could theoretically create iron using solar power in Sahara and then ship that to the Netherlands for use in breweries.

Thus you can decouple where and when power is generated from usage.

Wouldn't any heat energy gains be negated by the energy needed to ship the heavy iron from the Sahara to the Netherlands?

The energy cost of sea transport is like 50 less per distance traveled than on land and it has being like that for the last 3000 years. So the cost of shipment to Netherlands is just a cost of transporting to the nearest sea port from Sahara.

you need to run your ship on iron fuel as well. I guess that is the test though, can it ship itself?

I think an iron-powered ship would have to be a steamship.

There are still a small number of steam-turbine-powered merchant ships being built- specifically LNG carriers, where gas that boils off from the cargo can be used as fuel.

In Lensmen, they build iron-powered spaceships

They are far more powerful than other kinds of spaceships

Iron is heavy, dense as a material, but the cost-efficient of moving it around as an energy store depends on it's energy density (and it's initial cost), not it's basic density as a material.

> So we are looking at on order of magnitude more heat for the same electrical consumption 40% vs 400% efficiency. Instantaneous renewable energy prices are low, but rarely that low.

Wouldn't this imply that when heat is the goal we should always use these pumps? What is the limit of usefulness of these pumps? (eg, what if a whole city did it for all its heat needs?)

For one thing, heat pumps are more expensive than heaters. A heat pump is basically an AC unit that runs in reverse - way more expensive and complicated than a fire or a long thin wire.

It depends how much of a temperature delta you need. In cold climates often the efficiency is comparable or worse than resistive heaters

The big limit is temperature differential matters. The smaller the better!. 40C differential is about the max modern equipment can do with enough efficiency to be worth bothering with. Your home freezer might do more than 40C in extreme cases, but it isn't efficient (We also don't have many other options for this mode, for heating we have other options that work)

I can see that heat pumps can only move a certain amount of energy and to a similar extent electricity.

I remember looking at on-demand hot water heating. It's easier to do with natural gas.

But for electricity it would require something like a 100+amp electrical circuit and would have limits on flow rate and/or temperature increase.

for a heat pump, it just wouldn't work.

So for hot water the most common solution is to trickle charge a thermal battery - a hot water heater - and release it fairly quickly during a hot shower.

I can see the same sort of thing happening here.

Some industrial processes require very high temperatures, for which electrical generation is ill suited. Glass furnaces are one such case. Having an efficient local energy delivery mechanism that may be "recharged" with electricity is quite interesting.

I don't imagine beer production requiring such high temperatures, but honestly I don't know anything about industrial beer production. Anyhow the tech presented in the article is very cool.

On the other hand, electricity can create almost "infinite temperature" with relative ease. Temperature is, roughly speaking, the average kinetic energy of a material's molecules, and you can accelerate particles to extremely kigh kinetic energy with electromagnetic fields. So if you need that kind of "temperature", electricity is very suitable, even though it's expensive per unit of energy.

Another example for how electricity is a very valuable type of energy.

An asynchronous high intensity heat source, asynchronous in that it can be "charged" from electricity ahead of utilization time, whenever intermittent energy sources are high. According to the article at 80% efficiency, that's pretty good for a storage method that is cheap in capacity (think seasonal storage or beyond).

It's not the universal storage solution to end all search for other storage solutions because it's only really applicable in a niche (heat consumption), but that niche might not be as small as people think. I think that besides applications that require high temperatures it could also be a very good match where a particularly high maximum temperature isn't required (so heat pumps and the like could be applicable) but peak power demand is of a rarely occurring, intermittent nature. It might even be applicable to once-in-a-lifetime heat applications if they can be designed for easy oxide recovery.

What remains open is how cheap "charging" throughput can be built, as this is the weak spot of all other power-to-fuel concepts: the conversion facilities tend to be too expensive to remain idle outside of electricity oversupply times. And who knows, theoretically conversion throughput capacity might even turn out to be so exceptionally cheap (I certainly wouldn't expect that, but I'd be glad to be surprised) that even the inefficient electricity->fuel->heat->electricity might become competitive, once conversion outgrows direct heat applications. A market for inefficient but cheap long-term storage would definitely exist if there was any supply.

It doesn’t need those high temperatures but it takes a lot of energy to boil 10000 liters of water or whatever they end up using per batch.

This is a much more plausible use for the energy produced by this method than on-site glass bottle manufacturing, alluded to elsewhere among the comments.

>Some industrial processes require very high temperatures, for which electrical generation is ill suited.

Seems like an engineering challenge to me not an actual limitation.

It seems like the iron powder described in the article might meet this engineering challenge.

It's simply due to cost. Heat is typically cheaper than electricity, but only because we use heat to generate electricity with some losses.

And also because externalities aren't priced in, like carbon amounts.

For industrial-scale methane/coal burning, sequestering 90% or maybe even more of the CO2 is still a lot cheaper than using electricity. Don't get me wrong: it costs around 10% of the thermal load in electricity (technically, driveshaft power for turbomachinery).

It is true that it is a less efficient way of storing electric power than a battery as you need to created steam to drive turbines.

Yet this has some huge advantages over batteries:

1. We already know how to create iron at massive scale. Various iron smelters are going green already.

2. This has much higher energy density than lithium-ion batteries. More similar to gasoline.

3. The cost of storage is very low. Capital expenses to build batteries is much higher.

4. You can save a lot more of money by using existing industrial infrastructure. We can convert existing coal plants and any industrial process needing heath such as concrete production to use this. This makes it possible to get a solution in quickly . Scaling up global battery production in contrast is an enormous undertaking and we must prioritize EVs over grid storage at least initially.

My counterpoint would be that efficiency isn't always the most important variable in an energy supply chain, especially when your energy supply is renewable, and thus non-polluting, and as the marginal cost for generation of renewable energy continues to drop.

There are many situations where it would be much less expensive and simpler to replace a coal- or oil-fired boiler with one that runs off of iron filings, while still maintaining all of the rest of the equipment as-is. These locations could be off the grid, or in locations where it's not feasible to get a connection capable of supplying the megawatts of electricity it might take to produce enough steam.

You have to compare the efficiency to other similar options. For instance, (hydrogen-based) fuel cells also only have around 40-60% efficiency…

Though, I would assume that iron-powered (steam?) vehicles would not be competitive due to the weight ? (Even despite the extra weight that hydrogen or electric batteries add.) Maybe iron-burning steam boats ? Steam rail ?

Nevertheless, fixed processes (far from hydropower), especially those requiring heat energy, seem to be a great niche for iron burning ! And I say niche, but this niche might be even bigger than transportation, which IIRC is only like 25-30% of total energy consumption ?

My immediate thought was production of quicklime for concrete.

Yeah, and concrete making releases a lot of concentrated CO2, so there's already a lot of potential for algae-growing to make liquid hydrocarbon fuels there !

With large amounts of heat in your home, you can do cogeneration (combined heat and power). Heat some water via oxidation, produce electricity with the steam, collect the excess heat to keep the house warm. And you can do that all winter long as long as your basement is full of iron powder. And in the summer, you reduce the iron oxide again with excess solar power. (That's not what they want to do here but that would be the idea)

> "So you can't call this an "iron battery", because you don't get electricity out of it, just heat."

But add a fuel cell and the same reaction does give you an "iron-air battery":

https://en.wikipedia.org/wiki/Metal–air_electrochemical_cell...

A really interesting application of this came to mind that seems like an optimal use of this “iron heat battery” from a pure physics perspective.

Each kind of “energy storage application” can be broken down into different properties /requirements that can vary quite a bit.

Real time electricity response for 2-8hr periods so far seems well suited for Li-ion (LFP), given it is electrical energy, speed of response, and that it can be deployed virtually anywhere.

On the opposite end of the spectrum - you have seasonal energy storage (winter in colder climates, for example). When thinking about “energy containers” - Having iron powder doesn’t have the same requirements of alternatives I’ve heard, and seems like it could be the easiest to scale. Physically storing heat requires very large well insulated structures (rocks underground I’ve seen), hydrogen requires compressed gas cylinders, pumped hydro and compressed air needs specific geology (large valleys or large sealed underground caves).

Iron powder? An oxygen free warehouse Amazon sized. Helped that many winter applications ultimately only need heat, this seems like a good candidate that could be rapidly deployed given how on a relative basis - capital light and material accessible this is.

An oxygen free warehouse is overkill. Iron needs both oxygen and water to rust, so you can store it in a warehouse with just some climate control to manage the humidity. You could treat it like any other flammable powder, where you keep dust down as much as possible, and keep flame sources away.

Fine steel wool will burn readily, yet it keeps underneath the sink and only rusts over a period of years.

Electricity is steak, heat is hamburger.

I always thought about it as

Electricity <-> movement

↓

High quality heat (e.g. welding torch, tungsten lamp)

↓

Low quality heat (air conditioning)

You are right, of course, but I wonder if you couldn't use "waste heat" as the energy input of this "heat battery"? This would make the economics much better: instead of going, say, nuclear -> steam engine -> electricity -> iron -> heat (-> electricity), you could skip the first conversion losses.

Not sure how doable it would be from a chemistry point of view, though: you probably need at least one extra reagent or catalyst.

Source temperature of waste heat is an important factor. Heat radiating from a 2000C source is a lot more valuable & useful than that from some warm water at 60C.

You can't use that heat to raise the temperature of something to 61C, for example.

But you could use a heat pump and 60C water to raise the temperature of something to 61C. And would take a lot less energy to run that heat-pump than to actually heat that something to 61C.

What I've become interested in, is how easy it is to 'abuse' heat pumps like this. Could you take a heat-pump and this 60C water to heat a boiler to the point you can run a steam-turbine that produces more energy than it takes to run the heat-pump?

If not at 60C water, does this work at any temperature? There remains free energy in the temperature differential between an ambient 21C and the temperature of warmer water. It seems to me this should be harvest able. There are significant limits here from the Carnot theory, but they don't say this is impossible.

No that explicitly violates the second law of thermodynamics.

A heat pump is like a water pump: it's a lot easier to move water to a tank 61 feet off the ground from another tank 60 feet off the ground than from ground level, but you can't power the pump producing 1 foot of pressure head with a turbine being driven by 1 foot of pressure head.

Well, not without losing some water in the process. If you spill 10 gallons of water through the pump to pump 1 upward, sure that works.

And this is the exact idea behind ram pumps: https://en.wikipedia.org/wiki/Hydraulic_ram

Well in this case you need to pump 10 gallons up to get those 10 gallons to spill down. At some point in the chain, you need to consume energy from the environment to drive the system.

> At some point in the chain, you need to consume energy from the environment to drive the system.

Agreed, that's where spilling water comes into play:

         ┌───┐
    ┃  A ┊┃  │
    ┗━━━━━┛  │
             │   
             │┃┈┈┈┈┈┈┈┈┃
             └┨   B    ┃   
              ┗━━━━━┳━━┛
                    ┇
               ┃  C ┇ ┃ 
               ┗━━━━━━┛

Here, B->A is driven by B->C spillage. The critical point here is that B is a low quality source, but plentiful. rocqua's point was likely similar, if you replace height with thermal energy. You can't use the energy you put in A to move it from B to A, because it costs you at least as much to move it up.

However, we're not bringing it down to B. We're bringing it down to C (ambient 21°C in their example). There will be losses, but shouldn't it be doable, depending on the B-C potential?

Wouldn't Peltier modules be 100% efficient in theory for rising B to A (as long as thermal losses go to the hot side, that is: tey keep a side cooler), for instance? If so, the generated energy is the one stored between A and C, minus conversion losses, minus A-B. That is, B-C minus conversions losses from A-C. Probably not very efficient at the end, though worth it if you have plenty of B (and a excellent "C" sink).

Draining B into C is consuming energy from the environment.

You most certainly can drain some of the water from B to C so as to move other water from B to A. But you will always do so at a net loss.

The unit of water moved from B=60 to A=61 gains A-B=1 unit of potential energy, in addition to the B-C=40 units of potential energy it already had for a total of A-C=41 units. However, without a 100% efficient (read impossible) pump and turbine combo, it takes more than 1 unit of energy to raise up that unit of mass. This means you need to spill more than 1/40 units of water from B to C to get that 1 unit from B to A. Now dropping that water from A to C would get you 41 units of energy, 1 more than if you had spilled it straight from B to C, but it cost you more than 1 unit of energy to get into this situation.

A is indeed a higher quality source than B, but you are using that low quality source of B to generate your energy, and the efficiency losses there will always be greater than the gains you make on the higher one. To break even the initial height/temperature differential would have to be infinitely high.

Now if for some reason you could not harvest energy directly from B-C, yes you could use water going from A to C to power the pump from B to A, it would just be less efficient. This would be the equivalent of a siphon.

I think we agree on everything, thank you for expanding your reasoning.

> Now if for some reason you could not harvest energy directly from B-C, yes you could use water going from A to C to power the pump from B to A, it would just be less efficient.

I also think that was the initial point: some higher-efficiency generators ("steam turbine") are only available for high potential differentials. Whether raising the potential offsets the gains likely depends on the specific setup and energy source.

> Not if the total free energy in the system goes down right?

The heat pump requires power to run, the question is whether the power to run the heap pump is smaller than the power required to just generate the same heat differential directly.

No, the question is whether the power to run the heat pump is smaller than the additional power you'd gain by increasing the differential, which is always no.

You would always get more energy out of draining the whole 60 foot water tank directly to ground than you would by lifting the water to 61 feet and then draining it to ground.

So, say you have a tank at 61 feet, one at 60 feet, and a desire to move the water up.

Just let some of the water out of the 60 foot tank through a turbine. Put a shaft on that turbine, and put a pump on the other end. That pump should be able to pump water from the 60 foot tank up to the 61 foot tank. You would get less water in the 61 foot tank than in the 60 foot tank. Quite a bit less actually, but you could still harness the energy in the 60 ft tank to get some water to the 61 ft tank.

That is an external energy input.

No one is saying it's impossible to lift water up 1 foot (that's incredibly easy), the problem is it always takes more energy to lift water 1 foot than you can ever get from lowering it 1 foot. Likewise a heat pump can easily and efficiently create a heat reservoir, but you need to drain some other, larger heat reservoir to do so.

Yes, we are explicitly losing free energy from this.

Thing is, the free energy in 60C wastewater is hard to use, and the free energy in 200C steam is much easier to use (through steam-turbines). Hence, it might be worth it to sacrifice a large part of the free energy in the 60C wastewater to get it to a useable form.

What are the alternatives to use 60C wastewater to generate electricity? I can think of a sterling engine or a thermoelectric material. These would probably be more efficient, but they are new technologies, with practical limitations and development limitations. Hence, if my idea could generate any form of electricity, it might be more feasible than the other ideas.

From what I know about heat engines, I know it might not be possible to generate electricity this way.

For short-term storage at a nuclear plant, you could use molten salt thermal storage. In terms of a 'battery,' decomposing ammonia to H2+N2 is one way to directly convert heat to chemical stored energy: https://eeme.anu.edu.au/research/research-projects/ammonia-b...

There are other more complex processes like the sulfur-iodine cycle that was investigated for producing hydrogen from nuclear heat: https://en.wikipedia.org/wiki/Sulfur%E2%80%93iodine_cycle

As far as I can tell, these processes are probably less economical than generating electricity & then doing electrolysis.

Yes, it is not battery for electricity. As here the point was to get heat where heat is required, it is still very spot on.

Thanks for pointing this out. I was wondering if there was a catch. So this might be a good way to heat things but not to generate electricity.

Still an option for the mix if you have to capture a huge surplus of renewables. And there are no metallurgy smelting plants nearby.

Reply on news.ycombinator.com