Will We Run Out Of Lithium

Tech Boss
138 followers
17 Video Views·Oct 29, 2022
Will We Run Out Of Lithium
Lithium is the atomic element at the centre of the battery revolution that we've witnessed over
the last decade, with its demand doubling from 2009 to 2019. And now, with electric vehicles
and solar storage rapidly rising in popularity, we have to wonder, could we run out of lithium?
Energy usage propels our industry, our transportation, and our economy. Human
society has perpetuated a quasi-exponential growth in energy usage since the industrial
revolution, and arguably even before. These trends are fascinating in their own right,
implying that our distant descendants will one day need to harness the energy of entire stars,
or even galaxies, to meet their needs, known as Kardashev civilizations,
as we've discussed on this channel before. In the nearer term, though, our growing thirst
for energy might not affect our star or our galaxy, but it is affecting our planet. As someone
who studies other planets and looks for life elsewhere in the universe, projecting our own
future out has always fascinated me as a way of providing some kind of guidance as to what else
might be out there. So, full disclosure, I am an astrophysicist, not an economist,
but I think with some simple calculations here, we can gain some important insights
about lithium supply. When it comes to energy, we live in a time of transition. The combustion
of fossil fuels has historically been the engine of our modern industrial society,
powering everything from our homes to our factories, our cars to our airplanes. But more
than a century of feverish burning has altered the chemical composition of our planetary atmosphere,
in particular increasing the concentration of the heat-trapping gas carbon dioxide by 50%.
As we increasingly feel the climatic consequences of these changes, there is a growing sense of
urgency to transition away from fossil fuels to cleaner alternatives. Lithium-ion batteries have
emerged as one of the keystone technologies facilitating this transition. Invented by the
Anglo-American chemist Whittingham in the 1970s, early versions were toxic, non-rechargeable,
and often flammable. But decades of research and refinement has led to a far safer and fully
rechargeable device, capable of thousands of charging cycles. Lithium-ion batteries have
already had an enormous impact on the world of portable electronics, being found inside your
smartphone, your tablet, your gaming devices, but all of that may simply be a warmer pad for
what is to come. Because of electric cars like this one and solar storage systems,
our thirst for lithium may be about to elevate to a whole new level. There's three basic reasons
why lithium in particular reigns supreme when it comes to batteries. First, it has one more
electron than it really wants, defining the so-called alkali metals on the periodic table.
Such metals try to shed this extra electron to leave it with a full valence shell, which is their
most stable electron configuration. This tendency to easily shed electrons makes lithium well-suited
for batteries, but the second advantage is that it's the lightest of the alkali metals,
crucial for portable devices with just three protons and four neutrons typically. And third,
lithium is a fairly abundant and easily extractable metal, making up 20 parts per
million of the lithosphere and naturally found in concentrated pockets across the planet.
And I'd be remiss as an astronomer if I didn't mention that about one-fifth of the lithium
on our planet was likely forged in the Big Bang itself, 13.8 billion years ago. I mean,
just think about that next time you're holding a piece of lithium in your hand.
In brief, lithium-ion batteries start with a positive cathode and a negative anode,
separated by electrolyte. When discharging, the lithium atoms in the anode become ionized
and released to the cathode, thus generating a flow of free electrons to equalize the charge.
Charging is really just the opposite, where the lithium ions are released by the cathode and
received by the anode. The choice of materials for the electrolyte, the anode, the cathode all
affect the properties of the battery, and that's where a lot of research is currently focused.
Several different battery chemistries are in use and under development,
such as Lithium Nickel Manganese Cobalt, or NMC, and Lithium Iron Phosphate, LFP,
but the considerable advantages of lithium as the lightest alkali ion make it difficult to replace.
And even more advanced battery concepts like solid-state batteries
similarly depend on lithium, arguably even more so.
Now, whilst there is an increasing demand for many of the metals that go into these batteries,
for most of them one can find alternatives without huge trade-offs in performance,
whereas for lithium, it really is the linchpin to this entire battery revolution.
There are two main ways that lithium is currently extracted. The first is through
hard rock mining of spodumene ore, for which the largest deposits are found in Australia,
and hence why Oz currently produces almost half of the world's lithium supply.
After mining, the spodumene is processed into high-grade lithium hydroxide,
which can then be used in manufacturing. The other source is from lithium brine,
accumulations of saline groundwater that are enriched in dissolved lithium.
Brines in closed basins within arid regions are particularly easy to exploit,
and South America has enjoyed a booming industry in this regard.
The brine is pumped to the surface and then left to evaporate in a series of ponds under sunlight,
concentrating the brine over 18 to 24 months until it can be chemically extracted as lithium
carbonate. The lithium carbonate can then be converted to lithium hydroxide if needed
through an additional treatment. Besides from these sources, it's also worth mentioning that
lithium clays and even seawater contain plenty of lithium out there. Whilst that lithium is
extractable in principle, the problem is that it's currently much more expensive to use these,
and thus they haven't been a meaningful source thus far.
Okay, so we know why we need it and we know how to get it, so now we are ready to tackle the
question at the heart of this video. Really, there are two ways of asking that question though.
The first is to ask, will we run out of lithium in a sort of an absolute sense, as in the earth
will have no usable lithium at some point in the future. But another, and I think a more poignant
way of asking that question is, will production capacity meet demand in the coming years? So to
get started, let's have a look at what's been happening over the last few decades.
One of the best and open access resources for tracking this comes from the United States
Geological Survey. I'll link down below where you can find that data for yourself. Looking back to
1994, just three years after the first commercially available lithium ion battery was sold, global
lithium production was 6,100 metric tons. Now that's the weight as measured in pure lithium,
but industry standards typically convert this weight into lithium carbonate equivalent, or
LCE for short. Because lithium carbonate is heavier though, we have to make a conversion
here, which is about 5.3, thus giving us 32,500 tons LCE of production in 1994.
Now I highlight this because if you look into this for yourselves, you'll often come across
wildly different numbers, but often that is because they are quoting different lithium
equivalents rather than a genuine inconsistency. Okay, so working in the LCE industry standard,
this is how lithium production has evolved in the years since. There's certainly ups and downs here,
like the 2009 dip coming after the Great Recession and supply downturns during COVID,
but the overall trend is pretty clear and not that surprising. Indeed,
lithium production has increased by a factor of 16 over 27 years. Now this is supply and not demand,
but of course because of market forces, the two are generally tracking fairly close together
during this time. For example, supply really takes off in response to demand just after 2016
as the first truly mass-produced EV, the Tesla Model 3, enters the marketplace.
Looking at this trend, it's clearly not well described by a linear slope or even a quadratic
curve. So is there a simple mathematical fit? Technology development often starts with an
initial phase of exponential growth, so let's try that here. So here's a little mathematical
trick. Exponentials look like linear straight lines if we just change our y-axis here to a
logarithmic scale. And doing so, we indeed see something that looks more linear, and
fitting our exponential function through it, it does a great job. So switching back to the
linear scaling, the exponential behaviour is clearer now and corresponds to lithium
supply doubling every 7.9 years. I mean, it's really not a surprise that supply has been
growing like this, as mining operations are generally scaled to match demand,
enjoy greater investments, and benefit from ever-improving mining extraction technologies.
Looking ahead, extrapolation is always somewhat hazardous, but if we assume that this same rate
of improvement can be sustained for the next 6 years, then we would predict lithium production
to increase something like this, reaching 816,000 metric tons LCE. But the problem with this is
that's almost certainly not going to be enough. Because the demand for batteries in our homes,
in our cars, it's not part of some slow three-decade continuous growth curve, but rather
it is a paradigm technological shift that is occurring under our feet.
So let's look at that chart again, but now also plot the projected demand for lithium in red.
I'm showing here projections from three different sources, just to cover our bets. And yet,
all of them show that lithium demand will hugely outstrip even the exponential growth
curve that production has historically enjoyed. So of course, lithium miners are fully aware of
these projections and know that even the exponential growth of the past is not going
to be enough. Look, in a market economy, whenever demand rises, supply tries to catch up. And that's
why we are already seeing massive investments in lithium mining to try and anticipate this
increased demand. Combining projected supplies that are operational, probable, and even possible,
one can see that projected supply should be on track to match demand out to 2028. Again,
I'm showing you here three lines for three different sources for these projections.
And that's interesting because it means that lithium supply should ramp up even faster than
the exponential that we've seen in the past and hints at just how transitional the phase we're
entering really is. Now, so far, we've only gone out to 2028. Let's now go out to 2040.
There are few projections out this far, but the basic picture is one of continued growth in
lithium demand that then quickly outstrips production. On the face of it, this is quite
a disturbing picture, implying a huge shortage in lithium in the 2030s. However, let's remember
that these projections only account for planned lithium mining projects. It is quite feasible
that additional and as yet unplanned lithium mining projects will emerge in the coming years
to pick up this shortfall. But regardless, the real point here is that meeting this enormous
demand will be a huge economical, logistical and industrial challenge ahead of us. In the same way
that we might question whether the supply projections could be an underestimate, it's
also worth asking the same about the demand projections. The electric vehicle EV market is
certainly the most important factor here because in these projections, they're expected to dominate
global lithium demand past 2025 and comprise 70% of that demand by 2030. For some extra context here,
a 100 kilowatt hour EV battery has the same energy storage as about 2,000 consumer laptops
or 10,000 smartphones. So the question then is, could EV demand grow even faster than that which
is being projected here? To explore this, let's again take three projections from various outlets
for the growth of the EV market, where the y-axis here is what fraction of new car sales that year
would be some kind of EV, either fully electric or hybrid. Projecting out to 2035, the average of
these models is 19% and out to 2040, that's 31% less than a third. Now remember that's not 31%
of all cars on the road being electric by then, that is 31% of new car sales being electric in
almost two decades from now. Now perhaps if you're like me and you own an EV or thinking
buying an EV, that number might seem a little bit surprising because it implies a rather slow
transition to EVs. Government here plays an important role in the speed of said transition
because of course they can accelerate it by offering tax credits, incentives and even most
powerfully issuing mandates. So let's look at mandates in particular because they provide
a well-defined lower limit for the EV market over time. In California, the Air Resources Board is
presiding over a proposal for 35% of vehicles to be zero emission by 2026 and a complete ban on
new cars with internal combustion engines by 2035. Federally, the government stance is not
quite so aggressive as this with the Biden administration stating that they have a goal
of 50% of new car sales being electric by 2030. Indeed, the US seems to be taking that goal
seriously with the recently passed Inflation Reduction Act representing a $369 billion
commitment to that end. Heading over to the UK, in November last year the government announced
their plan for a ban on internal combustion engine cars by 2030, including hybrids soon
after in 2035. In the EU, a proposal to ban all internal combustion engine cars by 2035
received backing from parliament in June this year and may soon become law. Last July also saw
the announcement of the same date being targeted by the Japanese government but here mandates
eco-friendly cars to allow for hybrids and accommodate the Japanese car manufacturers.
And the big one, China, which of course is the largest car market on the earth,
has a similar eco-friendly mandate by 2035. And of course I have to mention the case of Norway,
shout out to any of our Norwegian viewers out there, because in Norway last year two-thirds
of all car sales were electric, putting them well on track for their 100% goal by 2025.
Now even ignoring the case of Norway, we can see a kind of quasi-convergence towards 2035
as where we'd expect an EV-dominated market. So putting this all together, along with the
relative size of global car sales in each region, it looks like the plans of many governments are
much more ambitious than the EV projections we saw earlier, where recall only a third of cars
were electric by 2040. In fact these five regions alone account for more than 70% of current global
car sales. If we combine these numbers and assume that their global share of car sales will be
approximately unchanged, then we find that EVs should make up at least 25% of new car sales by
2030 and 61% by 2035. And remember that's assuming that the other 30% of the world doesn't buy a
single EV and no one goes above their mandated minimum. Putting these lower limits back onto
our earlier chart of projected sales, we can see a pretty clear and stark tension. Now of course
mandates can change, laws pass and fail, but I think it's clear that the current stated goals
of the major car selling regions of the world imply a much higher demand for EVs and hence
lithium than current projections capture. Now I think this mismatch is probably because of the
recent flurry of announcements of mandates that we've been seeing, but it raises the question,
what does all of this mean for lithium demand going forward? Converting from EV demand to
lithium demand is complicated. It depends on projections about global car sales, the battery
size of each car and the lithium needed per battery. Now I could go ahead and make guesses
for those numbers but they won't necessarily be the same as those used in the various projections,
so that's not really fair. Instead I think it's fairer to simply scale those original projections
up to account for the increased EV demand. Let's call this the enhanced demand scenario.
Okay so recall that by 2030 70% of lithium demand is going into EVs in the current projections.
It's trending up, but let's just assume that by 2035 it's still at 70% just to lowball our
numbers. We can also see that the total lithium demand in 2035 is projected to be 3.7 million
metric tons LCE. Unfortunately we only have one projection going out this far, but reassuringly
the projections were consistent at earlier dates. Breaking apart this number it corresponds to 2.6
million metric tons LCE going to EVs and the remaining 1.1 to everything else. And that 2.6
number corresponds to an EV market share of 29% in the current projections. There's actually
considerable disagreement in the projections here, so I'm using the largest of the three
such that I have to scale things the least. So again I'm trying to keep our estimate as lowball
as possible. So if we need to scale that 29% number to match the government mandates, which
we estimated to be a minimum 61% by this point, that's an increase of a factor of 2.1, more than
double. Okay so let's now scale our 2.6 million metric tons LCE of lithium by a factor of 2.1 and
then finally add it back on to the other 1.1 million metric tons for non-EV products. That
means that we would need to revise the 2035 lithium demand projection up from 3.7 million
metric tons to 6.6 million metric tons, a roughly 80% increase. Even though we tried pretty hard to
lowball this number, pulling this back onto our lithium production chart shows that it's a pretty
dramatic shift. Fitting our previous projected demand curves with another exponential, which
again does a great job, we can now attempt to scale this demand curve for the enhanced
EV demand scenario, which would give us something like this. Now I'm going to be careful not to
project into 2040 here, since in this enhanced demand scenario, 2035 should represent a switch
over to a flatter demand as the EV market saturates, giving us perhaps something more like
an S-like curve after that point. Nevertheless, when we compare this chart with the projected
production rates from before, then clearly even nominal demand projections will greatly outstrip
plant supplies, but the enhanced demand scenario triggered by all of these mandates would greatly
exacerbate our dilemma. In fact, subtracting the two, we can see a significant lithium deficit
emerge in the late 2020s, reaching a staggering 4.6 million metric tons LCE by 2035. Whichever
demand model we use, we can see that lithium mining faces a monumental challenge to ramp up
at historically unprecedented rates. So one way of answering that question is to say, yes,
we likely will run out of lithium, at least in a day-to-day sense, unless we find some additional
and as yet unplanned supplies of it. The EV revolution would then face a real risk of production
falling behind demand, leading to spiraling lithium prices undermining any financial
instruments that were designed to incentivize their widespread adoption. But the good news is
that we have time. We can see this mountain coming and our final discussion might actually provide
a possible answer. At this point, we're pretty well equipped to answer the other way of asking
the primary question in this video. Will we run out of lithium in an absolute sense? According to
the 2022 report of the US Geological Survey, known global reserves of lithium are 89 million metric
tons. That's pure lithium weight, not LCE. However, if you look at the numbers from the previous
reports, which I'm showing you here, then you can see that that number has been constantly going up
over the years. Not because the earth is somehow making more lithium, but just because we keep
finding more of it. And so in the future too, we should expect this reserve number to keep
increasing. Nevertheless, for simplicity and for the sake of this video, let's just round that up
to 100 million metric tons, but with the understanding that that's probably an underestimate.
Now using that number, we can use the EV market to get a sense as to when those reserves would
run out, since we know that EVs are projected to dominate lithium demand in the future.
Rather than trying to calculate a date when the lithium would run out,
let's instead ask how many EVs that reserve could be used to manufacture. To do that,
we first need to know the typical amount of lithium in each EV. And to do that,
we first need to choose a typical battery size in each EV. Let's use this car, the Hyundai IONIQ 5
as our test case. No, this video isn't sponsored by Hyundai or anything, but this is the car I
drive. It is a fantastic car and it has a fairly typical battery pack of just under 80 kilowatt
hours. So let's round that up to a hundred kilowatt hours just to account for improved
battery densities and range that we might expect in future generations.
So next we need to know how much lithium each kilowatt hour of battery requires.
This varies somewhat depending on the design and chemistry, but it's around 160 grams per
kilowatt hour for current cars with a theoretical minimum of 70 grams for NCA type batteries and 80
grams for LFPs. Check out the linked article down below for that calculation. So accordingly,
let's take 100 grams per kilowatt hour as a rounded and optimistic value and 100 kilowatt hours as our
typical battery size, giving us a total weight of 10 kilograms of lithium per EV. Now with 100
million metric tons in reserve, which is 100 billion kilograms, that means that we can
manufacture about 10 billion EVs. It's at this point where most videos like this would generally
stop and say, okay, fine, we won't run out of lithium then because there's only 1.4 billion
cars on the road. So we have seven times more lithium than we need, even if every car becomes an
EV. But there's a few reasons why we might want to be just a little bit more careful than that.
First, global car ownership has been steadily rising as other nations approach US living
standards, which by the way, has almost one car per person. So it's not inconceivable that we
could approach several billion cars on the road. On the other hand, it has been argued that this
trend might reverse as self-driving cars and ridesharing increase. So I'm reluctant to increase
this number to several billion. But the second and more pressing concern is lithium recycling.
Virtually none of the lithium in your phone, in your laptop, in your EV comes from a recyclable
source right now. And if we don't figure out a way to efficiently recycle these batteries,
then we're essentially just throwing away this precious metal every time we scrap these devices.
Since lithium ion EV batteries are generally rated at about 10 years before needing replacement,
it would be seven times that or 70 years to effectively deplete the entirety of the
global lithium reserves once EVs dominate the car market. So in the long run, as in the next century,
lithium recycling is clearly essential. However, if we really did refuse to recycle lithium
batteries, there are other sources of lithium out there. I mean, there's 180 billion metric
tons of this stuff in the oceans, dwarfing the supply from mines or brine pools. But
at least for now, that lithium is far more expensive to extract.
So battery recycling could play a vital role in not exhausting our economically viable reserves
of lithium on a timescale of say, a century or so. But even a shorter timescale, a timescale of
decades, recycling could be key. Remember that demand currently looks set to outstrip supply.
At least that's if we define supply in terms of plant production projects. This leads to a lithium
deficit of 4.6 million metric tons by 2035 in the enhanced EV demand scenario. So that means that
we need some new and as yet unplanned supply of lithium to keep the EV revolution alive.
But the good news is that even in the more challenging enhanced EV demand scenario,
it turns out that recycling is theoretically capable of largely meeting that demand.
Let's look at our deficit curve again with the enhanced demand scenario.
That's deficit per year. Let's now treat it cumulatively, so that each year now shows the
total amount of missing lithium in the marketplace thus far, like accruing a debt. With that,
we can see that by 2035, when the market should start to saturate and flatten out,
we're missing about 17 million metric tons LCE of lithium, which is a lot.
Now let's try to see if recycling could pick up this shortfall.
So what is the cumulative amount of lithium out there in manufactured goods? Well, to ballpark
this we can just take the supply curve, shown here, and add up the total amount produced thus
far. We can't go infinitely far back into the past though, so let's only consider cumulative
supply since 2012, since lithium ion batteries have a 10 year lifetime and thus 2012 batteries
are reaching the end of their lives right now. We also can't really count lithium produced but
still in active use, since that's not yet recyclable. So for that, let's again assume a 10
year lifetime, which means that the actual amount of recyclable supply of lithium at a given date
is the total supply of lithium from the year 2012 up to said date minus 10 years. Finally,
you can never recycle 100% of course, so let's assume an 80% efficiency here.
Some emerging companies are actually claiming they can reach 95% efficiency, but we also have
to account for collecting up all of these batteries in the first place. With these numbers,
we can see that recycled lithium could, in principle, fully make up the lithium shortfall
out to 2035, after which remember we expect the market to level out in this scenario.
This is of course an approximate model here but this at least gives us a sense as to what's possible.
Now even if you think I'm being a little bit optimistic with some of the numbers here,
the real point is that recycling is not destined to be some small minor contribution but rather
could become a vital player in the supply chain of lithium in the coming decades.
And so recycling then solves two problems for us. It solves the long term problem of
eventually exhausting the plant's supply of usable lithium, but also the short term
problem facing in the coming years of the day to day supply chain issues.
Alternatively, we could start extracting lithium from clays to make up for the shortfall as Tesla
is currently considering. A process though that has thus far not been economically competitive
with traditional mining. However, I would say the fact that we have a sustainable source to
meet demand is something we should seriously consider before going down that rabbit hole.
Recycling of lithium ion batteries is still in its infancy. Previous approaches essentially
just smelted the battery down to try and recover the metals, which is lossy and expensive. But
recently we're seeing new companies pop up with innovative techniques that promise higher
efficiencies and lower costs. For example, the American Battery Technology Company essentially
runs battery assembly lines in reverse, disassembling batteries on an automated battery line with
some chemical separation techniques at the end. Another company, Lithium Cycle, instead
puts the batteries through a kind of gigantic shredding machine and then uses hydrometallurgy
to separate out the lithium and the other metals. And these companies are attracting
a lot of investment. For example, Lithium Cycle obtained 300 million dollars of investment in the
last year alone. So to finally come to our conclusion, it is possible for us to run out
of lithium. In the near term, that could take the form of a supply deficit that stalls the EV
revolution as well as everything else that depends on lithium ion batteries like energy storage and
portable electronics. In the longer term, a society that depends on huge amounts of lithium but
refuses to reuse any, would eventually exhaust their reserves. For both, recycling could play a
vital role. Efficient lithium ion battery recycling could satiate that lithium deficit that we
discussed earlier, although the techniques being used are still in their infancy and require some
further work to get up to scalability. More realistically, recycling can be supplemented by
tapping new sources like lithium clays during the most explosive part of the demand curve expected
in 2030s. In any case, clearly our plans for a carbon zero economy are going to be quite dependent
upon solving this problem and we need to start planning and investing in solutions for this
now. Look, tax rebates on EVs are great, but it's also time to think about the sourcing of the
lithium too. You know, it's sometimes said that our ability to foresee events, years, even decades
into the future, is what separates humans from the rest of the animal kingdom. We know this
challenge is coming and we still have time to adapt and adjust. Personally, I'm optimistic and
excited about the prospects this lithium ion revolution promises, facilitating widespread
solar microgeneration, zero emission vehicles, and an overall and urgently needed reduction in our
As is often true, a revolution is rarely a spontaneous act. It requires meticulous and
careful planning, often years in the making, and that's why it's important to stay thoughtful,
stay curious. Thank you so much for watching everybody and thanks for sticking to the end.
If you like what you see, be sure to do all the YouTube-y things, like, share, subscribe, you know
the story, and if you really want to help us out, you can click the link up above where you become a
donor to my research team, The Cool World's Lab, just like our latest supporter. So a big thank you
to Brynjolfur Sigur Jonsson. Thank you for your support. So until next time, see you around the galaxy.
