Placeholder Image

Subtitles section Play video

  • This video was made possible by CuriosityStream, sign up for the Holiday Nebula bundle deal

  • for just 11.79 at curiositystream.com/realengineering to get ad free access to our new podcast,

  • Modulus.

  • If I was to ask you today what technology breakthrough the world needed most right now.

  • What would you say? This needs to be a technology that we could realistically develop in the

  • next 10 years. So none of that sci-fi non-sense. This is something I think a lot. Nuclear Fusion

  • or cheaper, safer and cleaner fission energy are good candidates. Human society would be

  • transformed by economic fusion power, but fusion is most certainly not a technology

  • that can be commercialized in the next 10 years. Many people are working on safer and

  • cheaper nuclear fission, but that too has many hurdles to overcome. I think about how

  • the world would change if carbon nanotubes somehow became a viable material. A new stronger

  • and lighter material of carbon nanotubes calibre wouldn't open doors to new design possibilities,

  • it would open portals to new dimensions. But that again is not going to happen any time

  • soon. No, if I was to pick one technology that would have the biggest impact on our

  • society today, that is within grasp, it would be cheap, scalable energy storage for the

  • grid.

  • The electricity grid works nearly entirely on a just in time manufacturing method. We

  • generate the electricity just as it's needed. There is no warehouse of electricity that

  • we can dip into.

  • Pumped hydroelectricity does provide some storage and is a nearly century old technology,

  • but is not scalable to our current needs.

  • Lithium ion batteries are our best option right now. They have proven their worth in

  • the Hornsdale Power Reserve in Australia. It was commissioned primarily as a fast frequency

  • response service. [1] This means it can act as both a load when the frequency of the grid

  • gets too high, or as a power source when the frequency of the grid gets too low. Kind of

  • how a flywheel maintains the rotational speed of an engine.

  • You see our grids are designed to operate a particularly alternating current frequency.

  • If the grid deviates from that frequency, it can cause all kinds of issues that generally

  • just result in protective measures being activated to protect the infrastructure, and ultimately

  • cuts power to it's users. A blackout.

  • South Australia was struggling with these blackouts.

  • In 2016, tornadoes ripped through South Australia and damaged some power lines. This caused

  • the voltage and frequency of the grid to deviate from its baseline. [2]

  • This caused the wind turbines to trip their protective measures and lower output. Now

  • this was a massive problem because this is what South

  • Australia's power generation looked like on that day, with nearly 50% of their power

  • coming from wind. [3] To deal with the sudden decrease in output in wind the interconnector

  • to Victoria attempted to increase its power transfer, but rather quickly shut itself down

  • to prevent the line frying itself.

  • The grid basically did the technological equivalent of a human passing out when seeing a drop

  • of blood. A chain reaction of panic. Leaving 850,000 people without power.

  • In response, the Australian Energy Regulator is trying to sue the wind companies it had

  • approved for not doing a job they were not capable of doing in frequency regulation.

  • [4] I feel like they only have themselves to blame. South Australia had built far more

  • wind power than it's grid could reliably handle. It lacked the necessary interconnections

  • to neighbouring grids and energy storage facilities like pumped hydro, batteries, or simply reserve

  • power natural gas plants. It was a poorly planned grid. Instability was inevitable.

  • We are thankfully learning from these mistakes, but as renewables grow the challenge of preventing

  • blackouts like this is only going to grow. We won't just need fast frequency response,

  • but we will also need load shifting. Where we have enough storage to charge batteries

  • when renewables are available and discharge them when it isn't. This is going to be

  • expensive.

  • Lithium ion batteries are the cheapest we have right now, but when it comes down to

  • it, they weren't designed for this job.

  • They are designed to be light and energy dense for portable electronics, but for a stationary

  • battery that's a pretty useless trait to have. It's like having an underwater hair dryer.

  • Just doesn't make sense.

  • Lithium ion batteries are the cheapest form of energy storage available because their

  • mass market adoption has allowed for the economics of scale to reduce their price, but what if

  • we designed a new type of battery. A battery that was designed from the ground up specifically

  • for the grid. To learn more about this, I spoke with Donald Sadoway, a renowned professor

  • of materials chemistry at MIT and founder of liquid metal battery company Ambri.

  • Cut to Prof Sadoway interview:

  • The last thing I do is seek the advice of the incumbents. The incumbents are threatened

  • by radical innovation. You realize that the lithium ion battery did not come from the

  • battery industry. The battery industry refused to even manufacture the lithium ion battery.

  • So Sony, Sony wanted a better battery to power their handheld devices

  • And this is 1990. And Sony goes to all of the big battery producers in Japan , And they

  • go, and they say, here's the here's the formulation, build this. And here's a purchase order for,

  • pick a number, some 10s of millions of dollars. And each and every Japanese battery manufacturer

  • said, “No, I'm not building that.” “We have all this capital investment in the manufacture

  • of nickel metal hydride batteries. We can't build this battery in that plant.” And so

  • they said no.

  • And somebody at some point said, you know, if we want to have lithium ion batteries for

  • appliances. There is only one way we are going to have them. We're going to build them ourselves.

  • And Sony saysWe aren't a battery companysays, “We need batteries. And there's only

  • one we're going to get them we're going to build them ourselves. And so Sony built the

  • first lithium ion battery manufacturing facility.”

  • And very soon thereafter, they were getting inquiries from people who are building mobile

  • phones, saying can we have those? And then people who are building mobile computers,

  • laptop computers, can we have those? And by 1995, nickel metal hydride was pretty much

  • displaced.“

  • So what battery chemistry is Prof. Sadoway is trying to build and can it have the same

  • revolutionising disruptive effect on grid storage that lithium had for consumer electronics?

  • The idea started simply. Professor Sadoway had decades of experience in electrolysis

  • refining for metals like iron and aluminium. That process takes a lot of energy to refine

  • the metal. Why not try to make that process reversible and allow the reverse reaction

  • to give electricity back.

  • This is the basic concept of liquid metal batteries. We alloy and de-alloy metals in

  • a perfectly reversible reaction. They don't need to be light. They need to be cheap. And

  • as Prof. Sadoway says

  • “I say, if you want something to be dirt cheap, make it out of dirt

  • So how do we go about choosing materials for a battery like this? What does the design

  • ideation phase look like? Professor Sadoway is a professor of Materials Chemistry at MIT.

  • Looking at a periodic table is a different experience for him.

  • This is what he sees when picking materials for a technology like this. For the liquid

  • metal battery, we first need to refine our search down to metals and metalloids, which

  • are these elements.

  • Next, we need to maximise the difference in electronegativity to maximise our voltage.In

  • general, electronegativity is highest on top right of the periodic table and lowest on

  • the lower left. So, our electrode materials can be further narrowed down to elements in

  • these two groups. [5]

  • Next, as Prof. Sadoway said, if we want our battery to be dirt cheap, we have to make

  • it out of dirt. So let's plot our relative abundance of elements. [6]

  • Of the candidate elements for our negative electrode, Calcium is by far the most common.

  • Which is the negative electrode for the Ambri Liquid metal battery. However, they didn't

  • arrive at their current electrode materials just by analyzing the periodic table. Experimentation

  • was vital as this is a complex and dynamic system. They have tested several combinations

  • of different electrode materials from these two groups, and there are a lot of complicated

  • interactions to consider. [7]

  • Ambri has landed on a Calcium Antimony cell chemistry. So how does it work?

  • These materials are placed into a ceramic insulated cell together. When a current is

  • applied the materials begin to heat up. Eventually they will turn liquid and the metals will

  • separate naturally as a result of their density differences. The heavier positive electrode

  • sinks to the bottom with a neutral density electrolyte separating the lower density negative

  • electrode on top. This makes building the cell very simple. Lithium ion batteries use

  • complicated coating processes to build their electrodes.

  • This is the charged state, now when a load is applied the opposite electric current is

  • experienced. This causes the calcium electrode to break into a calcium cation and 2 electrons.

  • The cation travels across the electrolyte bridge and combines with the antimony and

  • the electrons that have travelled on the external circuit to form a new alloy. This continues

  • to happen until the calcium electrode is completely consumed. Now we just have the new mixed alloy

  • and the electrolyte. This is the discharged state. To get back to the charged state we

  • simply apply the opposite current and the reverse reaction occurs and creates a fresh

  • battery.

  • Now this brings another advantage. Lithium ion batteries degrade over time. As they are

  • charged and discharged, chemical reactions occur that damage the electrodes and reduce

  • their ability to hold a charge, and many of the ways we need a load shifting battery to

  • operate are the exact ways that accelerate this degradation over time.

  • Taking a lithium ion battery from full to zero charge is particularly damaging. As few

  • as 500 deep cycles, can reduce the capacity of the NCA batteries that Tesla uses by as

  • much as 20%. [8] That represents about a year and 4 months of daily use for our load shifting

  • battery, whose job will be a daily one.However LFP batteries, which Tesla has started using

  • in it's Chinese Model 3s, degrade much slower even under deep cycling and they have stated

  • that they will use LFP batteries for stationary storage in the future. Depending on the temperature

  • they operate at LFP batteries drop to 85 to 95% capacity after three thousand cycles.

  • Higher temperatures result in higher capacity drop.

  • However, Ambri have shown that their capacity fade is minimal even after 5000 cycles [9],

  • thanks to the continual creation and destruction of it's electrodes. Allowing us to fully

  • discharge our batteries on a daily basis for upwards of twenty years.

  • However, as I'm sure you have been wondering, keeping the calcium and antimony electrodes

  • so hot that they melt comes with disadvantages. For one, we are going to lose some of our

  • electricity to heating the materials up to operational temperature. This reduces our

  • round trip efficiency.

  • So, to explain it, if you put 100 units of electricity in, there are some losses because

  • there's some joule heating, and so on and so forth. With liquid metal battery, it's

  • about 80%. Because the difference, the 20% is the energy lost desirably to heat the battery

  • to keep it at temperature. So you say wow, 80%, that's 20% loss. What's up with that?

  • The round trip efficiency of pump hydro is 70%. So we're better than pumped hydro. But

  • the thing is that this is a case of don't don't answer irrelevant questions, because

  • the key question is, what is the cost of electricity.

  • So this is where things get a little complicated, luckily we have an equation to calculate the

  • levelized cost of electricity storage. [10] It's determined by the total costs, which

  • are the sum of the initial capital cost, the continual operations and maintenance cost,

  • the cost of charging and the end of life costs, divided by the total electricity discharged.

  • Based on Ambris calcium-antimony cell chemistry, the cost of electrode materials vastly undercuts

  • current generation lithium ion batteries. With the total cost of the liquid metal battery

  • electrode materials coming in at 17 dollars per kiloWatt hour versus 51.2 dollars per

  • kiloWatt hour for the most common nickel manganese cobalt batteries. [11]

  • If they manage to get the initial capital cost down 66%, that decrease in round trip

  • efficiency is a minor concern.

  • These continual costs are hard to predict. Operations and maintenance costs for lithium

  • ion batteries could include buying more batteries to bring total capacity back in line as the

  • batteries fade. We also have very little data for end of life costs, which will primarily

  • be determined by how easily disposed of or recyclable the batteries are. For both of

  • these metrics, liquid metal batteries will likely have an advantage.

  • However, even with the promise of liquid metal batteries, Lithium ion batteries have a major

  • leg up on any potential competitors. They have had decades to work on the manufacturing

  • process and reduce their price, and they are still getting cheaper.

  • Ambri have proven this cell chemistry works on the bench scale, but actually bringing

  • a product to market is much harder than proving the science works.

  • It's simply the the long journey from lab bench to, to marketplace. We, you know, here

  • at MIT, I with my my team of students and postdocs, we worked on this. I had a concept

  • and then we we reduced it to practice and then got it to the point where we said it's

  • time to start a company.Now, how do you take that and turn it into a marketable product

  • that is able to be manufactured? At the university, you know, you make five cells and one of them

  • works and you get a publication out of it and everybody's high fiving and so on. But,

  • but in manufacturing, you have to have, everything has to work. So, so we had to design the manufacturing

  • process. And there's nobody to turn to there's no there's no model. I can take the most brilliant,

  • the most competent people in the lithium ion battery sector. And almost everything that

  • they know is in applicable because they're the lithium ion chemistry is different, which

  • means that the format of the battery is different. their needs are different. I mean, they have

  • to guard against thermal rise, we have to guard against thermal fall. We want to keep

  • our batteries hot, they're trying to prevent their batteries from getting hot.

  • And there are dielectric hermetic seals that have to survive 500 600 Celsius. So obviously,

  • they're going to have to be ceramics. But ceramics are brittle, fragile, and they don't

  • like thermal excursions, but we have to be able to, to endure thermal excursions, and

  • I can give you a ceramic and you can do it like that. But it's going to cost something

  • around a NASA price point

  • Designing an entirely novel product is not easy. Those dielectric hermetic seals in particular

  • are a tricky bit of engineering. They need to be dielectric, to separate the positive

  • and negative electrodes. They need to form a seal to prevent gases and moisture from

  • entering the battery and causing corrosion and secondary reactions. It needs to be corrosive

  • resistant as those molten salt electrolytes can corrode many materials and to boot it

  • needs to be heat resistant since the battery operates at 500 degrees celsius. Those are

  • 4 very specific combinations of material properties that don't come with an off the shelf rubber

  • o-ring.

  • It's one thing to design a prototype that works, but it's an entirely different beast

  • to design a product that can be manufactured cost effectively and reliably.

  • When lithium ion batteries first came to market in the 90s, their price per kilowatt hour

  • was upwards of three thousand dollars, but over the past 3 decades that price has continually

  • dropped to about 150 dollars per kilowatt hour. [12]

  • There is no scenario where Ambri comes out of the gates at this price point, no matter

  • how cheap their electrode materials are, the price of a novel manufacturing method will

  • offset any cost savings until economies of scale take over.

  • This difficulty of bringing a new technology to market, despite the obvious potential advantages,

  • is called technological lock-in. And it makes it incredibly difficult for newcomers to enter

  • the market. If they can't compete with cost straight out of the gate, they are going to

  • struggle to find buyers.

  • In order for new products like this to get to market and start their journey to affordability,

  • they often need to find a niche market where their advantages outweigh their cost. So where

  • could liquid metal batteries find this niche market?

  • As we explained lithium Ion batteries are temperature sensitive. Without proper thermal

  • management lithium ion batteries will at best degrad faster, but they can also malfunction

  • or even catch fire. [13]

  • This has already happened, with a large grid scale lithium ion battery in Arizona. Where

  • battery degradation led to a thermal runaway. In other words a rack of batteries failed

  • and caught fire. [14] Leading to the shut of every battery storage facility in the state

  • until the cause of the problem was found.

  • These disadvantages of lithium ion batteries are exactly what is going open the door for

  • liquid metal batteries. The liquid metal battery can work just fine in extreme conditions.

  • After all, the entire product is designed to operate at 500 degrees. No cold or warm

  • environment is going to interfere with its operation. Making the battery better suited

  • for hot weather climates.

  • In an application where the batteries need to operate in a warm climate, while being

  • used daily and under deep cycling, liquid metal batteries may be able to justify their

  • initial high price for the right