Open source all-iron battery for renewable energy storage | Hacker News Reader

Open source all-iron battery for renewable energy storage

sciencedirect.com

137 pointsstevewilhelm56 comments

A fundraising appeal for this project was posted to HN in 2017:

https://news.ycombinator.com/item?id=15618494

I contributed. It's great to see the results published.

Actual paper.[1] This looks like someone's science fair project. This chemistry has lower energy density than nickel-iron (Edison, 1901). Iron batteries are commercially available.[2] Not much of interest here.

[1] https://reader.elsevier.com/reader/sd/pii/S2468067219300318 [2] https://www.essinc.com/energy-storage-products/

Lithium ion battery prices are currently below $0.17 per watt-hour and expected to drop below $0.10 by 2024. Does the all iron battery have any advantages aside from being more environmentally friendly and not requiring lithium? I must admit, the "safe for DIY" description makes me want to give it a try.

I'm also curious about it's potential to be built from recycled/junkyard resources, instead of needing access to rare earth metals. (Perhaps post-apocalyptically...)

Just go with lead-acid in the apocalypse. Embrace the aesthetic.

Iron is extremely abundant. It's too heavy for most mobile applications, but might serve stationary needs.

Lithium batteries from EOL cars are also going to become extremely abundant.

Maybe in 20 years. As usage of electric cars is still ramping up, demand for new batteries is going to outstrip supply of old batteries by far. And that's before we ask how difficult it is to extract raw materials from old batteries.

FWIW, the majority of the EV fleet with air-cooled batteries in hotter US states (like Arizona) are already needing battery replacements after ~5 years. See LEAF and eGolf.

FWIW, my city (Lünen, Germany) has 2 Multi-MWh storages built from them (1 from batteries from Smart cars, 1 possibly from i-MiEV batteries). Without any material-level recycling, just simple reuse, as energy density is a lot less important than for cars.

Maybe having 2 of those (admittedly pilot) projects in my backyard biases my estimations of their availability.

My understanding is that Lithium is not a big problem. Cobalt is more rare.

Last time I looked that appeared to be the case. Lithium is extracted[1] from primary sources. The rub then is that supply is limited by low historical demand and can be fixed by tapping other sources. The supply demand curve is thus just steep in the short term.

Cobalt is mostly a by-product of copper mining. Being a by-product means it's highly supply constrained. So the supply demand curve is very steep over the short, medium and long term.

[1] Leached from salt deposits or extracted from saline water sources.

Certainly iron is the metal to look at if you're going for the lowest cost, but the results so far are not inspiring.

There are about nine orders of magnitude in between these results and utility-scale energy storage, but it doesn't seem like that's what they're aiming for—the existing flow batteries they mention in the paper would be a better fit. They seem to be aiming at home applications, but they are seven orders of magnitude away from even the “tens of kilowatts” they cite in their abstract as typical for such applications.

However, both the negative results they mention in passing and the positive ones are major contributions to the progress of such low-cost battery designs.

Iron costs around US$0.03 per kg, while lithium is more like US$300 and nickel is US$30. In the intervening 3 orders of magnitude, cheaper than nickel but not as cheap as iron, are aluminum, antimony, arsenic, cadmium, carbon, cerium, chromium, copper, lead, manganese, samarium, silicon, tin, titanium, vanadium, and zinc.

Some of these are unpromising precisely because they are too electronegative; zinc, as they said, cannot be reduced electrolytically in water; the same problem applies to titanium, aluminum, and carbon—but maybe a different electrolyte could solve that problem. Others, like lead, are already in common use for batteries.

Chromium, titanium dioxide, vanadium, tin and tin oxide, copper, manganese and its oxides, cerium, samarium, and of course lead seem like plausible candidates. Indeed at least zinc-cerium batteries exist.

I'm not sure their materials even sound all that cheap. From the article

"The total cost of materials is $0.1 per watt-hour"

Or presumably $100/kWh

On the other hand with Zinc Air "our fundamental raw material cost of zinc is just $2-$3/KWh." from https://www.altenergymag.com/article/2019/03/zinc-air-batter... talking about NantEnergy's rechargables.

Yeah, the interesting comparison is the energy storage per dollar, not the cost of the raw materials.

Indeed. Nant have been talking $100/kWh all in which is cheap but I think that's projected rather than that you can actually buy them for anything like that. Though the material cost is interesting as a potential end point if you got super automated production. A problem with Vanadium flow for example is that the materials are like 60% of the cost of the battery so it's hard to drop much. With LiCo the materials are quite a big % of the battery cost though these things vary with time and may drop. I think Telsa are reducing the amount of Cobalt used dramatically for example https://qnovo.com/82-the-cost-components-of-a-battery/

Sodium ion batteries that don't use much in the way of exotic materials seem feasible. It's the sort of thing where someone has to be successfully selling them to prove that they are reasonably competitive though.

The cost of the raw materials is more interesting with flow batteries, where you can imagine adding big tanks containing the electrolytes without adding more cells.

It still comes down to the cost per unit of energy stored. If the electrolyte is a mediocre storage medium with low costs, it may not be competitive with an excellent storage medium with moderate costs, even after increased manufacturing costs.

Pretty much all of the standard electrode potentials are like a volt or two. You aren't going to find a room-temperature molecular chemistry that gives you 10 or 100 electron volts of energy per atom. Lithium is pretty much the extreme there.

Sulfur-based flow batteries are supposed to be very cheap in principle, with the active chemicals costing around $1/kWh. Of course all these lab technologies face a perilous journey to the market.

https://spectrum.ieee.org/energywise/energy/renewables/new-s...

> zinc, as they said, cannot be reduced electrolytically in water;

Actually this is total bullshit. You can totally electroplate zinc. Electroplating just about started with zinc. That's why zinc plating is called “galvanizing”, even if nowadays hot-dip plating is a more common way to “galvanize” metal. Either I misunderstood the article or it's wrong. (I'd bet on the first, though.)

+1 for discovering that a polymer used in diapers for absorbing moisture is a cheap and effective ion-permeable separator membrane.

Unfortunately only mentioned once. Are we talking about the granules inside the nappy here?

Also doesn't seem to have instructions for processing it beyond the short description.

This looks amazing - and great to see such detailed instructions.

One thing to note is that this probably isn't yet ready for home storage of solar power.

"Our iron battery has sufficient capabilities for practical use in low power devices and projects. The cell’s internal resistance is high, and so the discharge rate is limited."

At the moment it could be useful as a backup for high efficiency lighting. The bill of materials doesn't cover any electronics to monitor and maintain the cells as the charge/discharge.

But all pretty nifty!

Creating compound battery (and other storage) designs to address capacity, cost, and instantaneous peak demand might be fruitful.

Capacitors and flywheels make lousy bulk storage but are excellent for smoothing power demand and spinning reserve.

No, it doesn't provide enough power for high efficiency lighting, or any lighting. Yet.

In the article, it describes keeping an LED lit for 350 continuous hours. Lots of room for improvement too

An LED with 0.5 mA going through it is definitely visually distinguishable from one that's off, but it's not “lighting” either, except maybe for Borrowers in a romantic mood, or ants.

A single LED emits plenty of light. Just not compared to daylight. And this was for many days..

A single LED can emit enormously varying amounts of light. With some quantities of current, it can emit plenty of light. (Unless it's a low-power LED and burns up first.) With other quantities of current, it emits very little light, all the way down to no light when the current is zero. The brightness at a given current varies by the type of LED, but not all that much (unless you're dealing with infrared or ultraviolet LEDs). And at 0.5 mA, all LEDs are very dim, more like “barely visible”.

I suggest you try the experiment. Grab a 7805 and a 10kΩ resistor; you can find these in any discarded electronic device (from 1980 to 2005, anyway) or you can buy them at any electronics store. Connect the resistor between the “output” and “ground” pins of the 7805. Do not connect the “ground” pin to ground. Connect an LED between the “ground” pin (not the “output” pin) and actual ground. (The negative pin of the LED goes to ground.) Connect a positive DC voltage relative to ground to the “input” pin of the 7805; any voltage between 10 volts and 35 volts will work. Now you have a regulated 0.5 mA going through the LED; if you want to confirm this, replace the LED with a 1kΩ resistor and measure the voltage across it with your multimeter. Try different LEDs.

Do they emit “plenty of light”? No, they do not. None of them.

If you want variable current regulation, you can use a variable resistor, but you should probably put it in series with a fixed resistor. You aren’t going to burn up the 7805 without great effort, but you might burn up whatever you're driving with it.

It made 1mA of current, 0.5mW of power. Pretty long bow to draw linking this to renewable energy storage. A kettle uses a kW or 2 just to boil water. You would need a million of them to make a packet of noodles.

"just to boil water" has got to be the biggest understatement that I've heard in a while. There is a reason we use water in radiators: It's because water has a very high heat capacity.

I think this battery is meant to be used in small electronics projects. Like powering an arduino or even just an ATTINY.

How did you get that impression from reading what you linked?

From the article:

> The pouch cell as described, can provide ∼1 mA of current. At the typical operating voltage of ∼0.5v, each cell can provide ∼0.5mW. The cell is robust to at least ten cycles of charge and discharge without noticeable loss in capacity. An array of 6 cells is thus sufficient for low current electronics for sensing applications (such as low power microcontrollers like Texas Instruments MSP430).

It's research into making safe, DIY rechargeable batteries, which is pretty damned cool.

The reason the first person commented is because it's clearly touted as a solution intended for large scale batteries for renewable energy.

Restating that part doesn't change that, and it never states that use case as an intended application.

> "just to boil water" has got to be the biggest understatement that I've heard in a while.

It’s also a good illustration of how much we depend on extreme energy densities for everyday use cases. Coming up with a good storage solution for renewable fuels is going to be a defining challenge for our generation.

This is my favorite method so far:

https://qz.com/1355672/stacking-concrete-blocks-is-a-surpris...

Works well, portable, space efficient (could house it in the center of a skyscraper), and easy to disguise as something aesthetically pleasing

I guess abandoned mines with both above and below ground storage areas might also be useful. The deeper the better.

A much larger battery (129 megawatt-hours) can fit into a similar footprint:

https://hornsdalepowerreserve.com.au/

(it takes ~1 hectare, which maybe the tower of blocks won't have quite that footprint, it'll be at least that imposing)

But aren't the drawbacks of a Lithium ion battery going to outweigh the benefits of saving the space?

In a literal sense, apparently not. There's a real actual 100 megawatt-hour lithium-ion grid battery and (concrete?) pie in the sky ideas for using dry mass for gravity storage.

I wouldn't say it's "pie in the sky" when the technology already exists and has been demonstrated. It's a really simple system.

In a sense, you're right that any technology currently in use is better than a technology that isn't, but if every discussion ended there, we'd have no new technology.

My question was assuming someone is choosing between lithium ion and gravity storage. My understanding is that lithium mining is pretty awful for humans and the environment, the batteries can explode easily, and their lifespan is limited.

Besides being a weird shape, what are the drawbacks of storing energy in concrete blocks?

True, but it's also a very non critical example. Lastly, I'm trying to make a solar concentrated kettle. TCO Free kWatts

Methane/propane/butanol all work fine, they just cost more than petroleum derived gasoline.

Hydrogen is also probably feasible, just hard to store.

How many cells would you find in the average EV battery?

The fact that you can take relatively small cells and gang them together is amazing.

Curiously, the article has a link[1] to the sources, located at OSF (Open Science Foundation), but it asks for the login first, not allowing to view without registration. Doesn't look like open science to me.

[1] https://osf.io/4jvr9/

Clicking the "download" link with the pdf icon at the top gave me access to the entire pdf

https://www.sciencedirect.com/science/article/pii/S246806721...

Great detailed publication ! Makes me want to build one.

Curious to know how much efficient this battery can become without using too hard to source material ! And for which kind of cycles.

Even if components are environmentally friendly, electricity production is never footprint free, so if you need to 2x it to compensate...

It makes me think of the Acquion project from Pittsburgh (http://aquionenergy.com/). That was an interesting attempt to commercialize environmentally safe batteries intended for uses where low power density isn't a problem-- eg underground energy storage for windfarms or solar arrays. They went bankrupt a couple years ago, but the concept was really exciting.

The immediate technical challenge with these kind of batteries at scale is the electronics to monitor and maintain the battery banks as well as deploying these things. It would take a truckload to set-up something usable for a house, I think.

To put the costs intro perspective, an 83 Amp Hour car battery (typically a small car might have a 40-60 AH battery and a large car >100 AH) holds 1 kWh (kilo watt hour).

An 80AH battery can be had for about $100 on eBay.

The article states that the cost of materials might be $100/kWh.

So you could get the storage for about the same price using a car battery.

I skimmed through looking for some key numbers but didn't find them. I bet someone who knows how to read the many numbers they did share would be able to hazard a guess:

what would the $/kWh look like?

how many cycles until you hit 90%, 80%, 50% capacity?

100 $/kWh

This reminds me of the cheap water purification thing that got rolled out across poorer bits of Africa. Any cost reduction possibilities for this design?

What are the odds the steel wool could catch fire on these?

Only if the electrolyte dries out, and even then it will be probably covered by salt crust.

Reply on news.ycombinator.com