MIT’s Liquid Metal Stores Solar Power Until After Sundown | Hacker News Reader

MIT’s Liquid Metal Stores Solar Power Until After Sundown

bloomberg.com

61 pointsadventured26 comments

TFA is pretty thin on info. The "$500/kWh" goal seems to come from Musk's battery marketing rather than something inherent to this technology. And the technology itself isn't even hinted at. The description of "earlier versions" seems to imply a chemical reaction like regular batteries. However, speculating based on "liquid metal", maybe they've identified a metal with a melting point around ambient temp (or ambient temp of a sufficiently insulated vessel)? With high enough specific heat (typical for metals anyway), it could soak up energy by melting, and release it by freezing.

The article doesn't have any specifics, but the group has been publishing in this area for a few years: http://sadoway.mit.edu/

It's an electrochemical cell - I know some people who've worked on the project and IIRC some of the main benefits are lower material cost / kWh and improved lifetime as the liquid metal "electrodes" don't degrade the same ways solid ones do.

It's a bit dated now, but in March 2012, Dr. Sadoway gave a TED talk about how these batteries work:

http://www.ted.com/talks/donald_sadoway_the_missing_link_to_...

Very thin:

    "Sadoway, who is also the John F. Elliott professor 
    of materials chemistry at Massachusetts Institute of 
    Technology, wouldn’t say what goes into liquid-metal 
    batteries."

Here's a description from an earlier article -- same idea with different materials:

    "After hitting upon the idea of the liquid-metal 
    battery, Sadoway searched for the perfect electrodes: 
    he ended up choosing magnesium and antimony because 
    they are cheap and separate naturally when in liquid 
    form, the lighter magnesium rising to the top. A 
    liquid-salt electrolyte rests between the magnesium 
    and antimony electrodes, creating a cell with three 
    layers."

http://www.technologyreview.com/featuredstory/511081/am

So it's a chemical battery where the electrodes are liquids. (FWIW: Magnesium melts at 650 °C).

A little bit more info and clarification here [1].

What I find so clever is that the containment is about as simple as a "hot bucket", hot enough to keep the anode, electrolyte and cathode always molten. The are materials chosen so that their different densities keep them separated physically in operation. No membranes, easy scalability.

[1] http://spectrum.ieee.org/energywise/energy/the-smarter-grid/...

I hope this works out, but it bothers me when inventors use slick lines like "If you want something to be dirt cheap it needs to be made of dirt." It suggests to me they've been spending more time polishing their TED talks and investor presentations than in the lab.

Two sentences later:

> Sadoway and his team tested more than 1,000 cells with dozens of alloys and salts to find one that’s commercially viable.

Looks like they can do their homework after all.

Hm, not really. I see what you mean, but that's the kind of thing you start thinking about in the shower or when you're stuck on the real problem and need a break. Even before that, it's been circulating in your subconscious and may just pop out whole. It's not evidence either way of their technical ability.

A professor these days, even at MIT, needs to have the presentation skills of a entrepreneur - raising grant funds is no different to raising a series A.

Sadoways TED talk is in my personal top three for inspiration and sheer chutzpah. I not only hope it works, I hope he takes his billions and funds engineering schools all over

Practically everything is made of "dirt" just with different amounts of power applied in the transformation in to it's new nature.

They claim under $500/kWh, but classic lead-acid costs only $170/kWh.

Source: http://www.allaboutbatteries.com/Battery-Energy.html

I would guess those aren't deep-cycle batteries, and can be damaged by draining them even a few times.

Nothing a good controller couldn't solve.

Are you being serious here? (I cannot tell.) A controller cannot prevent structural damage to the electrodes. I once did research on using lead acid for grid applications. Shallow cycling doesn't save you much. You may double your cycles but if you get just half the energy from each cycle, then the battery's lifetime energy return is not all that dramatically improved.

In 2012, McKinsey & Co. reported on hybrid-electric, plug-in hybrid, and battery-electric vehicle (EV) competitiveness [1]. The focus was the impact of rising petrol and battery prices (measured in dollars per kilowatt hour).

U.S. gas prices are, as of 3 March 2014, around $3.48 per gallon [2]. This means EVs need battery prices south of $300/kWh. Even hybrid-electrics want no more than $400/kWh. The $500/kWh mentioned in the article, around Tesla Motors's costs, is not game-changing for automobiles.

[1] http://www.mckinsey.com/insights/energy_resources_materials/...

[2] http://www.eia.gov/petroleum/gasdiesel/

Just as a side note, Tesla Motors's battery costs are probably lower than this. (In the region of $200-$400/kWh)

http://www.teslamotorsclub.com/showthread.php/17590-Model-S-...

It's cool that this new storage system is 1/3 the cost of current battery technology, but I wonder what the environmental impact of "liquid metal" batteries would be. The article says one of their prototypes used antimony, which is toxic, with effects "that are similar to arsenic poisoning"[1].

[1]: http://en.wikipedia.org/wiki/Antimony#Precautions

Lithium-ion batteries, at least, "contain no toxic metals"[2].

[2]: http://en.wikipedia.org/wiki/Lithium-ion_battery#Environment...

"Earlier versions used molten magnesium and antimony, separated by a layer of salt, to store and release electricity"

This sentence implies that they changed type of medium. I scanned article, but have not seen claims of any other materials.

Lithium and lithium compounds can be toxic:

https://en.wikipedia.org/wiki/Lithium#Precautions

The batteries envisioned by Ambri would be enormous, shipping-container-sized things sitting (stationary) in the same place (e.g. electrical substations) for decades. If one cracked open, it would be a bad day at the plant, but nothing like a nuclear reactor meltdown. It could probably be back to normal in days or weeks.

The fact that the Ambri batteries are made using abundant materials means that those very materials are already present, in vast quantities, in the environment.

> I wonder what the environmental impact of "liquid metal" batteries would be

Keep in mind that in the use case these batteries are designed for, they would largely be replacing coal, oil, and natural gas.

Probably not. It's commonly assumed that the development of grid-scale energy storage will enable a large-scale transition to intermittent renewables like solar and wind. In reality, it will just entrench the lowest cost energy source. If you can store large amounts of energy, why not use it with a natural gas plant that produces electricity at much lower costs than solar or wind?

>> why not use it with a natural gas plant that produces electricity at much lower costs than solar or wind?

I was under the impression that natural gas plants can operate continually and that therefore energy storage isn't needed. Could you explain how cheap energy storage would enhance natural gas plants?

How much energy has to be used to keep the metal liquid? Can you switch the heating system off and let everything solidify without damage if you don't plan on charging or discharging in the near future? (Likewise with unexpected power outages.)

> Earlier versions used molten magnesium and antimony

I once calculated about how much magnesium and antimony would be needed to run the world on solar, using these batteries for storage. Based on the numbers I found in articles like this, and a little wikipedia, it came out to a thousand times our annual production of one of these metals, and ten thousand times our annual production of the other.

For a similar calculation using lead-acid batteries, see this post at Tom Murphy's Do the Math blog: https://physics.ucsd.edu/do-the-math/2011/08/nation-sized-ba...

Seems like this is a similar setup but different chemistry along the lines sodium-sulfur batteries (which are a molten salt). I had thought sodium-sulfur batteries were already at or better than $500/kWh; pretty sure new generations will be. There's already several hundred MW of NaS installed already.

I'm pretty sure one of my high school friends is an engineer at that compnay. He lives in boston and told me he's working on batteries for a company backed by Bill Gates. That's pretty awesome

Reply on news.ycombinator.com