12th May 2021 JANUS ANALYSIS

Lithium: Demand to Double Inside a Decade

• 71% of current Lithium demand is allocated toward batteries

• Forecast global Li Demand to grow 100% over the next decade
• By 2030, EVs will consume 61% of all Li-ion battery demand
• 2020 BEV sales in EU, US and China totalled 2.31m
• Forecast global 2021 BEV sales at ~3m
• Continued corporate consolidation reinforcing existing producer Oligopoly
• BUY recommendation large producers such as ORE/GXY & SQM
• Speculative BUY recommendation on ETCM, VUL and Cornish Lithium

After several years of stagnant production and depressed pricing, lithium

demand is again growing strongly, its future fundamentally tied to the growth of
EVs and hybrids. In deriving our forecast that demand would double inside the next
decade, we used the unrealistic (but extremely conservative) assumption that lithium
demand will be zero in every other sector other than Li-ion battery growth for EVs.
The exact quantum of growth is extremely difficult to estimate, for instance, there
is up to a 650% variance in Lithium contained within a BEV, plug in hybrid, or a vanilla
hybrid; moreover, given the growth in global Carpooling, which we estimate to be 59m
members, there is debate as to the future number of vehicles demanded. Despite this
uncertainty, the only headwind we see impacting Lithium demand in the short-term, is
the ongoing global semi-conductor shortages; forecast to continue to impact global
vehicle production until late 2022 or early 2023.
Because of its oligopolic structure, we believe that the vast majority of
additional capacity increases will come from existing producers, especially
those who own salar resources. We note that in the previous expansion phase (2010-
2015), 78% of all new and increased capacity came from existing producers, with the
average time to enter production from the point of FID between 2.5 and 4 years, which
in mining terms is very rapid. This raises questions as to how quickly smaller
Greenfields explorers will react and take advantage of rapidly changing market
conditions?
Predicated on first mover advantage, the ability to quickly increase name plate
capacity, and take advantage of improving market conditions, we recommend
ORE/GXY and SQM as strong BUYs. Resulting from their recent merger, ORE/GXY
intend to increase nameplate capacity by 225% to 130ktpa LCE. SQM (the world’s
second largest producer) is in the midst of increasing its lithium carbonate production
by 50%, and its lithium hydroxide output by 40%, inside the next 18 months; whilst
committing to a FID to add a 116% (attributable) lithium hydroxide development at its
Mt Holland JV.
There are also a number of unconventional suppliers that have captured our
imagination. They either have a capex advantage or a dual income model, and are
able to deliver a superior investment case than a stand-alone Greenfields explorer.
Speculative Buy recommendations include e3 Metals Corp (CVE: ETMC):
utilising existing O&G holes in Alberta, targeting the Leduc Reservoir hosting lithium
enriched brines; Vulcan Energy (ASX: VUL): plans to harness geothermal energy in
the Upper Rhine Valley, Germany to produce Lithium; and the unlisted Cornish
Gaius L.L. King Lithium, with some of the most prospective, untapped ground globally, hosting
lithium, tin, copper and geothermal (deep and industrial) prospects.
 Lithium Demand & Supply

Table of Contents

Page

Lithium Demand? Let’s Talk Batteries! 3

Lithium Shortages? Not in Our Lifetime! 7

Bow-wave Effect? Production is always Lower than Forecast! 10

Global BEV Sales Trends 13

EV Adoption Increasingly a Political, Not an Economic Narrative 16

Effect of Hybrid vs BEV Uptake on Lithium Demand 19

Near-term Recycling as a Secondary Source? Uneconomic! 22

Effect of Carpooling on Global Vehicle Growth? Substantial! 24

Rise of Graphene/Solid-State Batteries? Not any Time Soon! 24

Appendix A - Lithium Geology and Extraction 29

Disclosures 31

2
 Lithium Demand & Supply

Future Lithium Demand? Let’s talk about Batteries!

Originally used as a high-temperature lubricant (still accounts for 4% of total

consumption - see Figure 1) in applications such as aircraft engines during WWII,
lithium demand soared dramatically as a direct consequence of the Castle Bravo
hydrogen bomb test. The resultant yield was 150-200% greater than was
thought theoretically possible, it quickly became evident that the discrepancy was
the result of both δ6Li and δ7Li producing tritium in the midst of a thermonuclear
reaction, greatly enhancing the device’s output. Lithium oxide is also widely used
for processing silica, accelerating the melting of glass, reducing both viscosity
and melting point, and decreasing energy requirements; improving melting
behaviour of aluminium oxide during the smelting process.

Figure 1: Global Lithium Production (metric tonnes of contained lithium). Capacity increases
averaged 2.5 - 4 years from FID (final investment decision) to production, which in developmental
terms within the mining industry, is considered rapid.

Source: Ministry of External Affairs (2019), USGS (2021), Roskill (2021), Janus Analysis

Prices were suppressed for a decade following 1995 after excess lithium stocks
were (~41kt LiOH) disposed onto the open market from the US nuclear weapons’
programme. Primary demand began to grow from 2007 onwards as the
manufacture of lithium-ion batteries required a supply response; initially, Li
battery usage was dominated by disposables with voltages of between 1.5 and
3.7V. By 2010, Li-ion growth was dominated by increasing demand for
rechargeable batteries, essential for smartphones, tablets, laptops, power tools,
and other consumer devices.

By 2015, a fundamental change began to occur, global lithium battery demand

was primarily driven by large, dense Li-ion batteries catering to a burgeoning
EV/Hybrid market, measuring a 22.6% CAGR (see Figure 1). A single Tesla S
model 90kWh battery uses 63kg LCE, an equivalent content of ~13k cell phones;
underlying why Li market analysis is increasingly about future EV and Hybrid
demand.

We estimate the actual number of EVs sold in 2020 in the major markets of the
US, EU and China to be ~2.31m (~4.6%) units out of ~50.7m vehicles sold
collectively in those jurisdictions. The relative slack in uptake over the past
3
 Lithium Demand & Supply

decade has been blamed, in some quarters, on low oil prices resulting from US
shale developments. On that point, it is worth noting that oil still forms the basis
of our global transport economy, providing more than 97% of the fuel that powers
the world’s cars, trains, shipping and flight. Moreover, it is at a level little different
than that of four decades ago

Figures 2 & 3: Number of newly registered EVs collectively from the US, EU and China (left); and comparing annualised versus CAGR
growth rates over the same period (right). Collation of data from Figures 18 to 23.
3,500,000 140%

3,000,000 120%
Annualised Growth CAGR
2,500,000 100%

2,000,000 80%
65.8%

1,500,000 60%

1,000,000 40%

33.7%
500,000 20%

- 0%
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021F 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021F

Source: Janus Analysis

According to Chemetall, the cost of lithium is relatively immaterial to the

production costs of Li-ion batteries, only equating to ~1% of the total battery cost,
implying that demand is relatively inelastic to price. A doubling in price,
therefore, would have a negligible effect on underlying production costs, or even
on final demand. The accuracy in forecasting Li demand has become perilous,
being inextricably linked to idealistic proclamations, which in hindsight, more
often than not, have little basis in regards to facts on the ground. Resulting in a
number of high-profile projections being inaccurate by an order of magnitude.

Figure 4: Chilean Li2CO3 annual export prices (USD/tonne).

Source: Bloomberg (2019), Metalary (2019), Lithium Benchmark Intelligence (2018), BP Energy Review (2018), sigumBox (2013),
Servico de Aduanera (2011), Battery University (2019), USGS (2021), Janus Analysis

4
 Lithium Demand & Supply

After peaking in late 2018, prices fell >50% (on an annualised basis), the result of
significant growth in primary supply outstripping underlying demand, which was
compounded by the Covid pandemic where global car sales fell -16.5% (est.) in
2020. Spot prices have strongly rebounded >60%, as oversupply is now
increasingly being replaced by fears of a looming deficit as early as mid 2022.
Pricing, however, is more difficult than for many other commodities, and at one
point, Fastmarkets1 stopped quoting prices temporarily due to market
opaqueness. Typically sold in quarterly contracts, primarily around contracted
volumes, prices are typically set within a band, with LCE pricing within China
often substantially lower (up to 40%) than other international pricing contracts;
although we expect this arbitrage to completely disappear as deficits appear
within the next 12-18 months.

Figure 5: Growth in Li-ion battery demand has risen from 18% in 2010 to 71% in 2020, with a 2022
forecast using current segmental growth trends.

Batteries Glass Ceramics Lubricants Casting/AC Other

100%
10% 9% 9%
90% 21% 16% 2%
5% 4% 4%
30% 6%
80% 5% 6% 6%
6% 7% 10% 8%
70% 9%
7% 12%
60% 13%
9% 12%
50% 10% 17%

40% 24% 81%

71%
30% 26%
56%
20% 43%
28%
10% 18%
0%
2010 2014 2016 2018 2020 2022F
Source: USGS (2011-2021), Roskill (2021), TRU Group (2010), Janus Analysis

The future of Lithium demand is now fundamentally tied to the continual growth of
the EV and hybrid transition which is dramatically impacting the demand for
lithium (see Figures 2, 3 & 5). At the end of 2020, there were at least 50 BEVs
(battery electric vehicles) available to buy from various UK suppliers, which is set
to grow considerably with sales of new petrol and diesel cars banned from 2030.
In addition, there are more than 150 different hybrid models from various car
makers due to be released over the next three years. Hybrids increase fuel
economy by >25%, with most newer models allowing a pure EV for short
distances (<20km). Best suited for city driving, they are at their most efficient
when stopping and starting regularly; although, over long-distances, their
electrical systems add little to the efficiency of the engine.

In Europe, the rapid increase in current regulatory restrictions will ensure further
government incentives and penalties to support BEV sales growth. Starting in
2020, new passenger cars sold in Europe are permitted to emit no more than 95g
CO2/km (collectively); this is legislated to fall another 15% by 2025 (80g CO2/km),
decreasing again to 60g CO2/km by 2030. According to data collected by JATO,

1
https://www.fastmarkets.com/commodities/industrial-minerals/lithium-price-spotlight
5
 Lithium Demand & Supply

covering 21 countries across Europe, the volume weighted average CO2

emissions (NEDC) of vehicles registered during 2020 was 106.7 g/km – 12%
lower than the average recorded in the pcp (the largest drop to date). However,
it wasn’t enough to save all of the major manufacturers from incurring fines. The
Volkswagen Group faces a €100 million (£89m) penalty after it “narrowly missed”
its EU CO2 target; while Peugeot have disclosed that to have met its 2020
objective2, seven percent of all its vehicles sold would have had to be BEVs,
three times higher than its current level.

Figures 6 & 7: 2020 EV producer battery market share (left); and incremental commodity demand assuming all vehicles sold globally were
2017 Chevrolet Bolts; implied massive growth in demand for lithium, cobalt, rare-earths and graphite markets, whilst PGMs, primarily used
in auto-catalysts, are impacted negatively. (right).

-100% 400% 900% 1400% 1900% 2400%

Others
PGM -53%
14%
Steel -1%
Samsung LG Chem Silicon 0%
7% 26%
Aluminium 12%

BYD Manganese 13%

6% Copper 21%
Rare Earths 100%
Nickel 118%
Panasonic
21% CATL Graphite 264%
26%
Cobalt 1928%
Lithium 2511%

Source: Global X ETFs, Korea Herald (2021), UBS, Janus Analysis

The only headwind we foresee facing EV demand in the short-term is the global
semiconductor shortage that has forced various automakers (e.g. Toyota,
Volkswagen, Ford, Peugeot, Fiat, Jeep, Honda, Jaguar, Land Rover) to idle
production lines when they temporarily run out of supplies. These Computer
processors are critical for engine management systems, automatic braking,
airbags, automatic parking and the infotainment systems, etc. The main
manufacturers include TSMC in Taiwan (suffering from drought) and Samsung
and SK Hynix in South Korea, with smaller operations in the United States and
Europe.

This semiconductor shortage is also affecting products such as TVs, mobile

phones, game consoles, and was initially the result of lost production during the
pandemic (a fire at a Japanese chip manufacturer3, Renesas Electronics, didn’t
help either). A new surge in chip demand driven by changing habits fuelled by
the pandemic, however, has exacerbated existing shortages that are not
expected to be resolved until mid 2022, or early 2023.

2
https://www.jato.com/jato-dynamics-analysis-of-eu-co2-emissions-in-2020/
3
https://www.reuters.com/article/us-japan-renesas-taiwan-idUSKBN2BM09M
6
 Lithium Demand & Supply

Lithium Shortages? Not in our Lifetime!

Primary lithium demand began to grow from 2007 onwards as the demand and
manufacture of lithium-ion batteries required an immediate supply reaction.
Exploration success quickly followed, using existing geological models, targeting
salars and pegmatite operations. Displaying a classic supply response, from
2009/10 onwards, additional Lithium Reserves and Resources were being
delineated (displayed cumulatively – see Figure 8)4.

Figure 8: Growth of lithium Reserves and Resources over time.

Source: USGS (2021), Janus Analysis

Figures 9 & 10: Global lithium Reserve distribution from 2013 (left) to that in 2020 (right).

Source: USGS (2014-2021), Janus Analysis

4
Although this Reserve/Resource compilation is not necessarily JORC or 43-101
compliant, it is based on audited cumulations, and is, therefore, conservative and
reasonably accurate.
7
 Lithium Demand & Supply

Taking a snap shot in time, production in 2013 was dominated by Chile and
Australia (see Figure 11), however, substantial nameplate capacity was added
from 2015 onwards, in particular a threefold increase in output from Greenbushes
has meant that global production is increasingly dominated by Australia (see
Figure 12). This position is expected to be challenged in the near future by new
and existing facility expansions, primarily in South America. These include SQM
expansion at its Atacama salt flat operations, Australia’s Wesfarmers (ASX:
WES) FID on the Mount Holland project, Albemarle (NYSE: ALB), readjusting
plans to add about 125kt of processing capacity, while IGO Ltd (ASX: IGO) sold
its 30% stake in Tropicana, to fund its lithium expansion plans.

Figures 11 & 12: Global lithium production, comparing 2013 (left) with 2020 (right) abundances.

Source: USGS (2014-2021), Janus Analysis

Reserve numbers equate to >220 years supply at current production rates (see
Figure 13). If we include Resources, we have an additional ~1,150 years of
supply at current demand5. However, these kinds of numbers show nothing
regarding supply and demand dynamics, or the economics behind extraction.
For example, at the height of the last iron ore boom, BHP and Rio were
spending, on average, US$200m per Mtpa additional capacity, despite hundreds
of years of proven Resources. Despite record prices, only a single company
joined the ranks of the seaborne Oligopoly (i.e. Fortescue), and much of that can
be attributed to a unique individual, an unconventional and (at the time) unproven
mining technique.

The key takeaway in understanding future Lithium supply is that the quantum of
global resources is meaningless. Despite a 200% increase in global lithium
output as a result of various mergers, particularly led by the Chinese, the market
has continued to consolidate the existing oligopolistic market structure (see
Figures 16 & 17). Lithium supply has to be viewed within certain economic
parameters, that the theoretical wave of supply about to inundate demand, will
never eventuate; as explained in the following section titled, the “Bow-wave
Effect”. Not unlike the seaborne iron ore market, the oligopoly market structure
allows producers to cut or expand output quickly according to perceived demand

5
Geologically speaking, salars and pegmatites remain the most important lithium deposit
types in terms of production and undeveloped resources, however, there are some
relatively undocumented clay deposits and unconventional brine concepts that could
easily add substantial tonnages. Lithium remains relatively under explored globally.
8
 Lithium Demand & Supply

or supply, in turn providing a disincentive for other potential participants (i.e.

Greenfields explorers) to secure developmental funding.

Figure 13: Current Lithium Reserves into comparative context, dividing Reserves over current
production (in Years).

250

200

150

221
100

50
81
68
44 50 33
20 19
0
Lithium Aluminium Copper Cobalt Iron ore Lead Nickel Zinc
Source: USGS (2021), Janus Analysis

Recent production capacity increases have been driven more recently by Japan
and Korea, propelled by EV Li-ion battery demand, accounting for >71% of global
sales (see Figure 5). Historically, China’s dominance in the lithium market has, in
large part, been the result of it being the world’s largest EV market in 2020
making up >56% of vehicles sold globally, and >95% of the commercial vehicles
in operation; collectively, substantially larger than the rest of the world combined.

Figures 14 & 15: Global segmental lithium consumers in 2018 (left) versus those in 2019 (right), illustrating very recent capacity increases
by Korea and Japan.

Source: USGS (2019), Statista (2021), Janus Analysis

9
 Lithium Demand & Supply

“Bow-wave Effect”? Production always less than Forecast!

The phrase “Bow-wave Effect” describes the pattern behind the addition of
productive capacity for any commodity (with the exception of gold), which bears
little or no relation to the underlying quantity in Resource base. As the analogy
implies, the boat that follows behind never quite catches the “bow-wave”. Not
unlike many other commodities, there is always the appearance of an enormous
“wave” of potential Lithium supply about to engulf the market. However, what
inevitably comes to fruition, is merely a fraction of stated/planned capacity, and in
lithium’s case, the vast majority of additional capacity/new projects has been
brought online by existing, not new, producers (see Figures 16 & 17).

Figures 16 & 17: Hydroxide-lithium forecast vs 23% actual delivery of plant/project expansions (left); and Carbonate-lithium forecast vs
22% actual plant/project expansions (right).

Source: Source: ORE (2019), Janus Analysis

This “bow-wave” analogy has several important market implications, namely:

• Only existing (or very few new) lithium entrants will add to primary
supply.
• Until recently, there was excess capacity among existing brine producers
(rumours that some salar operations were operating at 60% nameplate
capacity) will dissuade the debt markets from providing developmental
funding to new entrants. This will suit existing market participants, who
will continue to actively manage supply to align with perceived demand
growth, for at least the next several decades.
• The large existing Resource base (see Figure 8) and surplus capacity
from recent developments, coupled with the industry’s ability to increase
production rapidly, will continue to deter Greenfields investments; and is
the primary reason why the lithium market has retained its oligopolistic
structure.

10
 Lithium Demand & Supply

The above oligopoly market structure is not dissimilar to what we observe in the
iron ore sector, that despite seaborne iron ore tonnages increasing >275% over
the past two decades, with the sole exception of Fortescue Metals, virtually all of
the additional global seaborne capacity came from three existing iron ore-based
Mining Houses; the Australian BHP and Rio, and Brazil’s Vale.

In 2012, global lithium production was dominated by four producers (and China).
In the intervening eight years, despite global production increasing 100%,
primary supply is arguably even more concentrated (see Figures 18 & 19). We
believe that this market consolidation will continue to occur despite the fact that
we believe that Lithium demand will double over the next decade (see Figure 30).
As evidenced with Australia's Orocobre buying Galaxy for $1.4Bn; creating the
world’s fifth largest lithium miner, combining hard rock, brine, and chemical
assets across Australia, Argentina, Canada and Japan. Superficially, there are
very few corporate or operational synergies to be gained, other than
strengthening its balance sheet (~$487m cash), improving access to finance and
streamlining product marketing. Critically, it will allow plans for the merged entity
to more easily finance its increased production to >130kt LCE (lithium carbonate
equivalent), up from ~40kt currently. Particularly interesting is Rio Tinto’s
entrance into the sector, following a breakthrough production process at its Boron
mine, recovering Lithium from waste piles accumulated over the past 90 years;
with a FID pending ($50m capex for 5kt pa lithium carbonate). Critically, Rio has
another borate project in Serbia that may also begin to produce lithium as a co-
product. If these projects are successful, expect Rio to make a major acquisition.

Figures 18 & 19: Comparing global lithium producers in 2012 (left) with 2020 (right).

Source: SQM (2019), Roskill (2013), RKE (2021), Janus Analysis

Several existing producers with favourable expansion prospects include:

• Sociedad Quimica y Minera de Chile (NYSE: SQM): guidance is 120kt

of lithium carbonate and 21.5kt of lithium hydroxide production rate by
the end of Q421; to be expanded to 180kt of lithium carbonate and 30kt
of lithium hydroxide by the end of 2023 from their Chilean assets. In
addition, SQM have made the FID (50/50 JV with Wesfarmers) for the Mt
Holland project, involving the development of an open-pit and a
processing plant at Mt Holland and a refinery at Kwinana, Western
Australia; targeting 50ktpa of lithium hydroxide.

11
 Lithium Demand & Supply

• Orocobre/Galaxy Limited (ASX: ORE/GXY): ambitious expansion plans

(see above).

Whist we propose that the sector will remain highly concentrated, there are a
number of potential unconventional suppliers that we think are worthy of
investment consideration. The creative and industrious “potential” for all of these
junior projects is that they are either using an existing capex advantage, and/or
an uncorrelated income model, which will allow them to enter a market, whilst not
having any geological barriers per se, certainly has market and financial moats
(i.e. barriers of entry).

• e3 Metals Corp (CVE: ETMC): utilising existing O&G holes in Alberta,

targeting the Leduc Reservoir hosting lithium enriched brines;
• Vulcan Energy (ASX: VUL): plans to harness geothermal energy to
produce lithium in the Upper Rhine Valley, Germany; and
• Cornish Lithium: unlisted, but we feel it could be a future small
company Rockstar, assets include lithium, tin, copper and geothermal
(deep and industrial) prospects.

12
 Lithium Demand & Supply

Global BEV Sales Trends

Approximately six percent of new vehicles sold in Europe is now a BEV (nearly
11% if you include plug-in hybrids); with 745k sold in 2020 (see Figure 20),
consensus suggesting that BEV sales will increase >40% to ~1.05m in 2021.
Unique and unlike China and the US, compound growth has been relatively
stable (see grey line in Figure 21). In a market that sold ~9.9m units in 2020,
plug-in EV market penetration was ~7.5%, helped by significant State subsidies.
Moreover, at its current growth trajectory, there is a chance that the European
BEV market will be larger than China’s inside three years.

Figures 20 & 21: Newly registered EVs (including plug in hybrids) in Europe from 2011 to 2020E (left); and annualised versus CAGR
growth rates over the same period (right).
1,200,000

100%

1,000,000
Annualised Growth CAGR

80%
800,000

56.5%
60%
600,000

40%
400,000
40.3%

200,000 20%

- 0%
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021F 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021F

Source: Forbes (2021), Bellona (2021), Janus Analysis

Figures 22 & 23: Number of newly registered EVs in China from 2011 to 2020E (left); and comparing annualised versus CAGR growth
rates over the same period (right). 2020 growth in Chinese BEV sales was ~10.5%, recovering from a -6.2% decline in 2019.

2,000,000 400%
Anualised Growth CAGR
1,800,000
350%

1,600,000
300%
1,400,000
250%
1,200,000
200%
1,000,000
150%
800,000

600,000 100%
79.4%
400,000 50%
37.4%
200,000 0%

- -50%
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021F 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021F

Source: WSJ (2021), S&P Global (2021), SCMP (2020), Janus Analysis

13
 Lithium Demand & Supply

Historically, China has been the largest EV market globally, accounting for half of
the world’s consumer EV vehicles sold and >95% of the commercial vehicles in
operation6. Despite the official target for BEVs (including plug-in hybrids) and
hydrogen fuel cell vehicles remaining at 20% by 2025, the government will cut
current subsidies by 20% this year for private vehicles and 10% for EV public
transport, including buses and taxis. Given that EV sales have largely plateaued
over the past three years (see Figure 22), coupled with subsidy cuts, our earlier
forecast assumed a drop in EV sales to around 2019 levels. However, for
reasons largely unknown, possibly related to a combination of stable economic
growth, stimulus policies on vehicle consumption, and numerous manufacturer
promotions, Q121 EV sales are up ~37% (roughly in-line with ICE sales which
are returning to pre-pandemic levels), a rate we have assumed will be maintained
until the end of the year.

Figures 24 & 25: Number of newly registered EVs in US from 2011 to 2020E (left); and comparing annualised versus CAGR growth rates
over the same period. Estimated growth in EV car sales is ~1.4% in 2021, compared with -10.1% in 2020 and a -8.9% decline in 2019. In
a market that sells ~14.45m cars pa, EV market penetration in the US is 2.3% (right).
400,000
250%

350,000 Annualised Growth CAGR

200%
300,000

150%
250,000

200,000 100%

150,000
50%

100,000 29.5%

0% 1.4%
50,000

- -50%
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021F 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021F

Source: S&P (2021), EV volumes (2021), Inside EVs (2020), Janus Analysis

Not unlike China recently, for the moment it appears that BEV sales in the US
have peaked, but this may be an aberration, and we await further environmental
policies from the new Biden administration. Looking behind the rhetoric of
Biden’s recent $2.3Tn stimulus package (over a 15-year period), deducting
amounts allocated toward manufacturing, semiconductors, job creation, research
and utilities, which require little or no commodity input. Only one trillion dollars
has a physical asset development bias. In particular:

• $620Bn on transportation including 20k miles of roadway and 10k new

bridges; including:
o $174Bn dedicated toward EVs, in particular, building 500k
charging stations7;

6
The irony being that China pollutes more than the US and all Developed Countries
combined https://rhg.com/research/chinas-emissions-surpass-developed-countries/
7
This compares favourably with the number of fuel pumps in the US, estimated to be
~920k, via ~115k petrol stations (https://www.marketwatch.com/story/how-many-gas-
stations-are-in-us-how-many-will-there-be-in-10-years-2020-02-16), assuming eight
pumps per station. It will be interesting to see the effect on encouraging future EV
14
 Lithium Demand & Supply

o $80Bn to clear the repair backlog and modernise rail (passenger,

freight trains, and public transit), in particular, the high traffic
Northeast Corridor;
o $25Bn on airports; and
o $17Bn on ports, waterways and ferries;

The unifying demand feature of this stimulus package, therefore, is not metals,
but cement, aggregates, asphalt and rail. Unlike China (<20%), more than 80%
of the US steel production is from recycled sources, their last operating blast
furnace was dismantled a decade ago. Moreover, the marginal increase in steel
demand annually will only be in the vicinity of 20-30Mt pa, which can be covered
by existing capacity. There is some additional demand for copper, but most
stocks in that space are already over-hyped and have poor financial
fundamentals.

As a result, we believe that by just establishing recharging infrastructure is

insufficient to transform US EV sales. It will require either a large increase in fuel
excise, and/or the introduction of a large subsidy programme.

consumers with a rapidly expanded charging network. Although the above budget is
spread out over 15-years, so meaning its effects may be hard to monitor.
15
 Lithium Demand & Supply

EV Adoption Increasingly a Political, Not an Economic Narrative

WSJ reported8 that various auto makers in Europe, including Volkswagen and
Daimler have approached a number of European Governments to introduce the
new taxes on CO2 emissions from petrol and diesel-powered vehicles in order to
make EVs more competitive; primarily targeting highway tolls or higher fuel taxes.
In regards to the auto industry, government/corporate interaction is normal
business practice in that engine technologies are strongly dictated by
government policy. Diesel engines are not allowed in California as a result of
coarse particulate matter generation. Whereas in Europe, as a direct result of
German influence (despite the recent VW emissions scandal) diesel technology
was strongly promoted via tax relief and subsidies throughout Europe. To a point
where Angela Merkel enlisted David Cameron’s support in 2013 to delay certain
emissions targets, because none of the German car makers could meet them in
time.

Table 1: A selection of countries with outstanding proposals to ban the domestic sale of internal
combustion engines, and by what date.

Source: Janus Analysis

A decade after BEV introduction, however, general uptake has been (to-date)
extremely muted, with adoption rates ranging from 2 to 6% in major markets.
This is driving government policy to sustain existing subsidies, but increase
penalties and legislation to induce consumers choices. Interestingly, there is a
clear sales dichotomy between Southern/Eastern Europe compared with more
the wealthier Northern Europe (e.g. Netherlands, and Nordic countries), with 30%
of the total of all EVs purchases being in Germany where a subsidy of up to €9k
exists per vehicle, while in Norway, 54% of all new cars sold in 2020 were BEV,
aiming to become the first nation to eschew the sale of all new petrol and diesel
cars by 2025 (see Table 1). Some of these regional differences can be attributed
to more investment in recharge infrastructure, and offering greater cash and tax
incentives in order to accelerate adoption. But the largest determinant, we
believe, comes down to wealth and relative affordability.

8
“To Win Electric-Vehicle Wars, Europe’s Auto CEOs Want More Taxes on Gas
Guzzlers “. WSJ https://www.wsj.com/articles/to-win-electric-vehicle-wars-europes-
auto-ceos-want-more-taxes-on-gas-guzzlers-11618570821
16
 Lithium Demand & Supply

Figures 26 & 27: Lithium segmental demand in 2010 (left), compared with 2020 (right) illustrating the growth in battery demand.

Source: USGS (2021), Roskill (2021), USGS (2011), TRU Group (2010), Janus Analysis

Like all things in life, additional factors make this EV purchase divide more
nuanced.

• EV Prestige: various studies examining early EV adopters suggest that

the ideal demographic are younger London based professionals (UK) to
middle-aged men living in rural or suburban multi-person households
(Germany); with ownership levels strongly tied to income levels. In
particular, their willingness to buy an EV was strongly associated with a
perceived higher socio-economic status.
• Premium Price: numerous publications are quoting that batteries make
up approximately a third of an eventual EV vehicle’s cost9, it begs the
question, shouldn’t we have seen EV prices fall by the equivalent
quantum of the battery input overtime? Yet, according to JATO (2019),
EV sales prices in Developed Nations have increased (on average) 42%-
55% over the past eight years. Arguments to explain this discrepancy
cite consumer demand for luxurious interiors and more sustainably
sourced materials, but this rationale fails any simple Empirical test. Does
a Tesla use more leather than it did eight years ago? Have LCD centre
console screens quadrupled in price? Despite Tesla’s unit production
increasing 2,100% over the past eight years, has it suffered production
cost dis-economies of scale? In all cases, the answer is No!
• Corporate Fleets: critically, fleet and businesses, not private consumers
are the largest purchasers of BEVs in Europe (JATO, November 2020);
therefore, implying that the growth in sales the result is of individual
purchase decisions, is possibly misleading. Registrations in Q120 by
businesses and fleets accounted for 59% of the EV total, which is not
dissimilar to Q119, whereby 54% of BEV transactions were also fleet and

9
According to Bloomberg (2020), battery pack prices have recently been cited at less
than $100/kWh, by comparison prices were >$1,100/kWh in 2010, implying a price drop
of ~88% in real terms. Meaning, if the Tesla 75kWh battery that currently costs around
$11,700, would have cost $97.5k a decade ago (if that particular battery existed!). The
quoted average ~$US137/kWh or ~$126/kWh on a volume-weighted average basis,
ignoring chemistry differences. The greatest cost variance being the cathode, because of
the compositional differences, e.g. cobalt, nickel and potentially, manganese.
17
 Lithium Demand & Supply

business sales. To put that into perspective, conventional European fleet

sales as a percentage of total sales typically range from 5% in Belgium to
19.2% in Germany and the UK (Statista, 2021)10.

We caution investors against accepting that these hard targets as expressed in

Table 1, as actuals. They appear to have been imposed without regard to
purchase costs, infrastructure or power network investment, and inherent
operational limitations based on chemistry. As a result, there are a number of
increasingly valid questions that policy administrators need to answer:

• Is total BEV adoption a realistic scenario?

• If not, why not?
• What is a more realistic/sustainable level?
• What type and levels of hybrids will be incorporated into the final sales
mix?

The inherent risk with implementation of political ideals without regard to financial
realities and/or mobility practicalities of the general populace, is that these
policies can be undone via popular dissent. It is in this vein that we pay particular
interest to French politics11 and the election of its President in May 2022.

10
Two possible reasons may include: (i) the rise of low-emissions zones across the
continent is incentivising fleet operators to actively avoid the costs of operating an ICE
(internal combustion-engine) fleet within a payment area. But the truth is, these low
emission zones are not that wide-spread yet; and/or (ii) EVs theoretically require less
servicing and lower maintenance than an equivalent ICE potentially lowering overall fleet
expenditure. But we note that EV resale figures suffer substantially higher rates of
depreciation, and given their higher initial cost of purchase, must considerably offset any
conceivable maintenance savings.
11
Rise of the “gilets jaunes” protest movement was initiated against high fuel prices
resulting from “eco taxes” meant to dissuade the French from using cars; unfortunately
penalised the poor disproportionately.
18
 Lithium Demand & Supply

Effect of Hybrid vs BEV Uptake on Lithium Demand

Historically, it has been very difficult to accurately model future lithium

consumption, not only to account for the variance of different BEV battery types
and capacities, but to also incorporate the rapid growth in hybrids. A Honda Fit
has a 20kWh battery, whilst Tesla S can now be specified to have a capacity of
up to 90kWh. Looking at 11 different BEV models, the average capacity was
~39kWh, driven by the arrival of more recent cheaper models, typically with
smaller battery capacities; with plug-in hybrid output typically 25% that of an EV
(see Figure 28).

Figure 28: European car registration by fuel-type, updated to March 2021. Note, we have attempted
to separate BEV and plug-in-hybrid numbers, attributing the latter to overall hybrid sales, believing it’s
a closer definition fit.

50
Plug-i n Hybrids BEVs 45
45 43
41
Estimated battery capacity (Kh)

40

35

30

25

20

15
10.2 10.6 11
10

0
2017 2021 2025
Source: Statista (2021), Janus Analysis

Evaluating pure hybrid vehicle ranges, we found that they vary dramatically (13 to
50km), with battery capacities ranging from 4-15kWh. The mean/median from
our compilation appears to be ~4 to 6kWh. Meaning, if we take a snap-shot in
time, it takes approximately eight hybrid vehicles to equal the Lithium
equivalence of a single in-situ BEV. The implication being, despite hybrids
making up 18% of EU market sales (see Figure 29), in BEV-terms, they only add
~36% of additional overall battery demand, despite being >200% more prolific.
Why is this distinction important? As previously mentioned, in the UK, hybrid
cars are mandated to be outlawed by 2035, which if enacted, will dramatically
increase underlying lithium demand.

19
 Lithium Demand & Supply

To demonstrate the possible effect on lithium demand12 using different EV types,

we have assumed three scenarios assuming the UK continues to sell 2.2m
vehicles per annum:

• In the first scenario, if 100% of unit sales were BEVs13, annual UK LCE
demand would be ~57kt pa.
• In the second scenario, if 20% of unit sales were BEVs, 20% plug-in
hybrids14, and 60% vanilla hybrids15, annual LCE demand in the UK
would be ~19kt pa.
• In the third scenario, we assumed 100% of sales in the UK were vanilla
hybrids, annual LCE demand would be <8kt pa.

The discrepancy between these three scenarios implies that the variation
between different types of EVs (using current capacity averages) can have up to
a 640% theoretical difference in the quantum of underlying Lithium demanded.
We remind that EU hybrid sales have increased 700% over the past six years
(without State financial intervention). By comparison, over the same period, BEV
sales have risen a more modest ~163% despite substantial Governmental grants
in the form of tax breaks and subsidies. Our working narrative that if a
technology is useful and applicable, uptake is inevitable and usually rapid;
conversely, subsidies only influence consumer choices whilst they endure.

Figure 29: European car registration by fuel-type, updated to March 2021. Note, we have attempted
to separate BEV and plug-in-hybrid numbers, attributing the latter to overall hybrid sales, believing it’s
a closer definition fit.

90
Diesel Petrol BEV AFV/Hybri d
80

70

60
52
50
%

40

30 24
20 18

10 6
0
1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

2012

2014

2016

2018

2020

Source: ACEA (2021), JATO (2021), Auxiliarie de L’Automobile (2020), European Automobile Manufactures Assoc. (2019),
Cargreencongress (2020), Janus Analysis

12
A single Tesla S model 90kWh battery uses 63kg LCE, therefore assume 0.7kg LCE
per kWh.
13
Assume a 39kWh capacity average
14
10.6kWh capacity average – see Figure 28
15
5kWh capacity average
20
 Lithium Demand & Supply

Although outside the scope of the analysis of this report, the growth of E-
commerce, especially during this recent pandemic has led to a dramatic increase
in demand for individual address deliveries, typically using the ubiquitous white
delivery van. Previous analysis of their replacement hypothesised smaller
electrified drones incorporating delivery, app and aerial-based autonomous
services, although realistically, this appears to have stalled due to logistical
difficulties and may now be a more distant aim. Volkswagen, Daimler, PSA
Group and Fiat have commenced converting diesel vans models to electric drive,
with quoted BEV ranges from 100 to 280km.

Critically, the UK Government is offering a plug-in van grant of up to £8k (£5k

more than the equivalent car grant). Critically, real world range, however, is their
greatest shortcoming, typically based on theoretical open road distances
travelling at 100kmh, non-stop, with no air-conditioning, in summer conditions.
Anecdotal evidence suggests that in winter, with frequent stop starts, many
drivers struggle to reach 60% of quoted ranges. However, as battery prices
continue to drop, capacities are likely to double to allow a realistic alternative to
current diesel powerplants.

Figure 30: Global Lithium Production (metric tonnes of contained lithium) with forecast growth
expected to double in the next decade.

Source: Janus Analysis

21
 Lithium Demand & Supply

Near-term Recycling as a Secondary Source? Uneconomic!

It is a reasonable assumption to believe that recycling will inevitably play an

important component in overall global lithium supply-chain, for example, an
estimated ~90% of lead-acid batteries are recycled. Lithium batteries have the
capacity to hold and generate considerable amounts of energy, and therefore, if
left in storage indefinitely, would pose a significant fire and explosion risk. In
Europe, over the past decade, lithium as a percentage of total portable battery
takeback, has risen from one to more than six percent. There are two main
routes for recovery available: (a) hydrometallurgical, via leaching, where cobalt,
copper and lithium is extracted and recovered. The other is (b) a
pyrometallurgical process, where the batteries are melted in a furnace then
reprocessed. Despite this “inevitability”, as yet the recycling of Li-ion batteries
effectively does not yet exist.

Unlike other battery types (e.g. Ni-metal-hydride) that are relatively homogenous,
Li-ion battery heterogeneity is the result it being a relatively new industry, with
various companies creating a plethora of different componentry and chemistries
in an attempt to gain a competitive edge. Moreover, in regards to Li-ion batteries,
the technology is continuously evolving, meaning that, apart from a generic
hydrometallurgical process, extraction could be outdated inside several years.
Whilst not presenting the same environmental hazards associated with lead-acid
batteries, lithium-ion receptacles have inherent electrical charges, chemical
dangers and burning reactants, particularly when exposed to water (e.g. UPS
Airlines Flight 6) 16.

Figure 31: Summary of a Scoping Study for a Li-ion recycling plant, based in Perth, WA.

Source: NMT (2021)

A study released by Neometals (ASX: NMT) summarises the costs for a 50tpd
plant (see Figure 31). It involves a two-stage shredding process, followed by
drying and beneficiation to separate coarse metal and plastic from the feed for
processing, utilising a hydrometallurgical methodology. The resultant metal
materials to be sold as scrap metal. In the modelled financials, there is no
allowance for tax, debt or any other type of funding; with an assumed 88%
recovery of Co, Ni and Cu contained in the battery feed material, and a 70%
recovery for Li. Although this project is being marketed as a Li-recycling venture,
Co contributes approximately 70% of the modelled revenue, underlying the
importance of by-product credits. Overall, the modelling implies that the total

16
https://aviation-safety.net/database/record.php?id=20100903-0
22
 Lithium Demand & Supply

cost of recycling lithium from batteries is ~400% higher than that from primary
supply, underlying the financial challenge.

Economics aside, we cannot envisage a future reality whereby a modern society

would allow Li-ion batteries to build up indefinitely without a disposal solution.
Concluding that eventually some form of recycling will be mandated, with a likely
impost levied Li-ion on the end-user.

23
 Lithium Demand & Supply

Effect of Carpooling on Global Vehicle Growth? Substantial!

We have written extensively on Carpooling, largely based on quantitative

research published by Shaheen & Cohen (2020). Not wanting to repeat
ourselves, other than to point out that Lithium demand is strongly reliant on the
future numbers of vehicles produced. The idea that in 2018 we could have
reached global “Peak-Auto”, until recently, would have been thought
preposterous; with the introduction of several billion consumers from the rise and
development of China and India, simultaneously. And yet, sales in many markets
had either been stagnating or declining 12-18 months prior to this recent
pandemic. To determine what sales projections were for 2030, before carpooling
started having a material effect, we refer to McKinsey (2017)17, who estimated
that total annual global vehicle sales in 2030 would be ~115m units (including
10m shared vehicles!). This compares with 63.8m sold in 2019 (56.3m sold in
2020 under exceptional circumstances), not dissimilar to levels sold in 2013.

Figures 32 & 33: Asian number of rideshare members vs number of operated vehicles by providers (left); and number of members divided
by vehicles, versus 2-year CAGR (right). With an overall 75% CAGR over the past 13 years, it demonstrated 61% CAGR over the past
two-years, indicating market growth is still very strong. N = 10: China, India, Indonesia, Japan, Kazakhstan, Malaysia, Singapore, South
Korea, Thailand, and United Arab Emirates.
25.0 120 250 250%

100
20.0 200 200%
No. Vehicles (k)

80
Members (M)

15.0 150 150%

Member/Vehicle ratio

60

10.0 100 100%

40

5.0 50 50%
20

0.0 - - 0%
2006 2008 2010 2012 2014 2016 2018 2006 2008 2010 2012 2014 2016 2018

Source: Shaheen & Cohen (2020)18, Shaheen et al. (2018)19, Janus Analysis

What is carpooling? Carpooling and/or ridesharing, is a collective group

arrangement that allows individuals to book time slots to use a single vehicle.
The proliferation of UK platforms include Drivy, Zipcar and Liftshare, which
underlies the fact that at various times, a vehicle is still a critical requirement.

17
Autovista Group (2017) “Global auto revenue pool to almost double by 2030, with
recurring revenue surging to 20% share, says McKinsey.”
https://autovistagroup.com/news-and-insights/global-auto-revenue-pool-almost-double-
2030-recurring-revenue-surging-20-share
18
Shaheen S. & Cohen A. (2020) “Innovative Mobility: Carsharing Outlook; Carsharing
Market Overview, Analysis, and Trends”. UC Berkeley, DOI 10.7922/G2125QWJ. 6 p.
https://escholarship.org/uc/item/61q03282
19
Shaheen S., et al. (2018) Innovative Mobility: Carsharing Outlook. DOI
10.7922/G2CC0XVW. 7 p.
https://cloudfront.escholarship.org/dist/prd/content/qt49j961wb/qt49j961wb.pdf?t=pa6fa3
24
 Lithium Demand & Supply

However, increasingly, for many, short notice transportation platforms are now
the more flexible, affordable and realistic alternative to car ownership. Weighted-
average carpool vehicle to membership ratio using data from Shaheen and
Cohen (2020)20 in 2018 was ~180; whilst Asian membership/vehicle ratios are
already at 210. If we assume that 70% of these members forego the purchase of
a vehicle, then increasing membership (we estimate 2020 membership at ~59m,
see Table 2) will increasingly have a material impact on levels of global car
ownership.

Furthermore, we hypothesise that the pandemic will, in fact, accelerate, rather

than stymie the transition from car ownership to car use; primarily as a result of
economic considerations:

• The largest negative financial impact will be felt by younger adults and
the less educated, whose jobs are often in the service and hospitality
industries; businesses which in many instances, have either shuttered,
and/or will need to reorganise operations to reopen.
• Vehicle ownership is increasingly no longer considered aspirational for
younger generations. The reasons are multi-faceted and include
increased debt levels, underemployment, inability to afford payments,
petrol, insurance, maintenance, repairs; and
• The proliferation of App platforms allow carpooling members, at short
notice, to utilise transportation that is flexible and affordable; in the long-
term, this presents a realistic alternative to car ownership.

Table 2 & Figure 34: Growth of carpooling/rideshare globally, assuming membership ~59m by the end of 2020 (left); and graphical
representation of Carpool/Rideshare members at different CAGR to 2030 (right). NB: 2018 CAGR was ~53%.

900
809
800
15% CAGR 20% CAGR 25% CAGR 30% CAGR
No. Members (millions)

700

600 546

500

400 363

300 237

200

100 59

-
2020F 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030

Source: Janus Analysis

How is this important in regards to predicting future lithium demand? When

forecasting future lithium demand, not only will it be critical to predict the levels of
BEVs and Hybrids (and type), but increasingly, also the number of vehicles that
will be manufactured. Underlying this global paradigm shift of “car ownership to
car use”.

20
Op cit.
25
 Lithium Demand & Supply

Rise of Graphene/Solid-State Batteries? Not any Time Soon!

The reason why lithium-based batteries are not a certain long-term bet, is the
enormous speed at which scientists and engineers are experimenting with new
combinations of materials to lower the cost and boost capacity. As a result, it is
entirely unclear what the base electric car battery will look like in several decades
time, or even, which commodities they will rely upon?

Figures 35 & 36: Patents for graphene related applications (2005 to 2020E) over time have fallen out of favour due to the long
development times and costs, and, we assume, technical difficulties (left); and comparing various battery types, in particular Li-ion with the
new generation of graphene batteries (right).
400
300

350
250

300

200
250
Wh/kg

150 200

100 150

100
50

50
0
0
05

06

07

08

09

10

11

12

13

14

15

16

17

18

19

E
20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

Lead Acid NiCd NiMH Li-ion Graph. Li-ion Graph. Li-S

20

Source: European patent office (2020). Bloomberg (2018), NanoGraphene (2019), Janus Analysis

One the most potent future materials is graphene, the strongest material ever
recorded, more than 300 times stronger than A36 structural steel and six times
lighter; at 130 gigapascals, it is more than 40x stronger than diamond. The
thinnest material known, its crystalline structure elastic, able to stretch up to 20%
of its length. A very efficient electrical conductor at room temperature, it can
sustain densities six orders of magnitude higher than that of copper; its charge
carriers have the highest intrinsic mobility, having the best thermal conductivity of
any composite. Lithium sulphur (Li-S) batteries have low toxicity, are low cost
and potentially have an energy density of 2,567 Wh/kg-1, five times higher than
that of existing Li-ion batteries. Although challenges remain, including inorganic
salt deposition at the cathode due to highly soluble reactants within the cell, and
the inherent low conductivity of sulphur.

The advantages of a graphene batteries include it being lightweight, durable,

suitable for high-capacity energy storage, and potentially, have dramatically
shorter charging times (e.g. 80% recharge capacity in 8 minutes21) than the
current generation Li-ion batteries. Li-ion batteries typically cannot handle more
than 500 charge cycles, unlike graphene batteries which can handle 1,500 to
2,000. Cambridge scientists (a number of years ago) created a prototype lithium-
oxygen battery with a theoretical energy density up to 10x greater than that found
in current generation lithium-ion cells; researchers claimed that if
commercialised, the resultant battery would be 20% the cost and one-fifth the

21
https://www.graphene-info.com/gac-group-announces-its-aion-v-sporting-graphene-
battery-will-start-production
26
 Lithium Demand & Supply

weight of current Li-ion batteries, yet able to travel the 650km (e.g. between
London and Edinburgh) on a single charge.

The current global market for graphene is tiny by any measure, but it has the
latent technical potential to transform macro sectors from batteries, mobile
phones to electric cars, even energy harvesting. By contrast, Lithium has several
key challenges, it is a relatively poor conductor and physically deforms as it
discharges energy, resulting in shearing and cracking. Coating the lithium with
graphene oxides reduces both issues, its high conductivity helps keep its shape,
allowing the battery to last longer. As an interim step, Samsung engineers have
developed a graphene Li-ion battery technology that could result in substantially
longer-lasting power packs. Using a silicon anode, researchers grew layers of
graphene on top to improve the density and longevity, experiments extending
power outputs 50% to 80% greater than those commercially available;
theoretically extending a current smart phone battery from 12 to 21 hours. Scalar
in application, these batteries could equally be applied to electric cars, allowing
them to match the range of their current petrol-powered counterparts.

The closest technological analogy is the transition from incandescent light bulbs
to compact fluorescent lamps, which provided the same amount of visible light,
but used 20-30% of the electric power, lasting eight to 15 times longer (ignoring
health risks). A host of legislation was introduced in a number of developed
countries to restrict or outright ban incandescent light bulbs as a matter of public
policy, in favour of fluorescent lighting for environmental reasons. After the
commercial introduction of LED lamps (which are significantly more energy-
efficient than fluorescent lamps), the most efficient of which, are able to produce
200 lumens per watt (Lm/W)22, fluorescent lamps became quickly obsolete,
without a single Governmental directive anywhere globally. Which underscores
our macroeconomic narrative, that less, not more, governmental intervention into
technological innovation will ultimately result in better social and environmental
outcomes.

Figures 37 & 38: Total lithium-ion revenue breakdown by application in 2020E (left); and application demand for 2030 if the current
automotive demand growth differential is maintained (right).

2020E 2030F
Industrial Industrial
8% 5%
Consumer
electronics
Automotive 13%
30% Consumer
electronics
24% Energy
Storage
21%
Automotive
61%
Energy
Storage
38%

Source: Semiconductor Engineering, Frost and Sullivan (2018), Janus Analysis

22
With frequent switching on and off not reducing life expectancy.
27
 Lithium Demand & Supply

Figuratively speaking, graphene has been on the threshold of transforming our

known universe for years, and still, we are waiting! Despite (well founded?)
rumours that Tesla are actively developing and experimenting with a graphene
battery presently; Samsung’s experience is a salient lesson in subdued
expectations, primarily centred around production costs and quality control.
Despite spending billions in R&D, commercialisation of any of the purposed
applications has proven to be far more difficult; with all the graphene applications
mentioned in glitzy regular PR releases in years past, quietly shelved.

Li-ion batteries could be around for many decades yet. EV are not yet the largest
segment for ion batteries (see Figure 37), but we believe that to be only a matter
of time (see Figure 38).

28
 Lithium Demand & Supply

Appendix A – Lithium Geology and Extraction

Global lithium production is approximately 90kt pa (see Figure 1) of contained

metal. The two main sources are brine lakes and saltpans that produce the
soluble salts, lithium carbonate and lithium chloride; and the extraction of a
mineral termed spodumene (see Table 3), a silicate containing both lithium and
aluminium. Lithium brine production bodies are found in salt lakes (salars),
typically in hyper-saline regions within basins where the lithium is sourced from
surrounding rocks (predominantly granites). Extraction involves pumping brines
into a series of ponds, which are concentrated by solar and wind evaporation,
with a single tonne of lithium requiring ~750t of brine, with the process typically
taking 24 months. Further processing of this solution remove other minerals
such as B, Mg and K, resulting in a lithium carbonate (Li2CO3) product.

Figure 39: Map of global lithium reserves and method of extraction. Orange circles designate
pegmatite (hard-rock) lithium deposits, blue circles are lithium brine deposits, with black circles in the
US and Mexico depicting lithium clay deposits.

Source: Reuters (2019)

The key Li-mineral in hard rock mining is spodumene (lithium aluminium

inosilicate) which occurs within granitic intrusives called pegmatites. Unlike
brines, hard-rocking mining is typically well constrained in grade and recoveries,
whereby operating costs (excluding mining costs) are largely dependent on the
price of raw materials, such as sulphuric acid, soda-ash and energy prices. The
end product is typically lithium hydroxide (LiOH).

A third type of deposit (for which very little literature exists), not yet in production,
is best exemplified by Hawkstone and their delineation of a clay hosted lithium
deposit. Prior geological surveys have identified large areas of hectorite (a
magnesium-lithium smectite) in a number of areas: western United States, and
potentially, into northern Mexico. The current deposit of interest is interpreted to
be a geological hiatus, where intensive historical evaporation occurred (not unlike
modern-day Salars in Chile, Bolivia and Argentina). The potential resource
29
 Lithium Demand & Supply

tonnages of these deposits dwarf anything we currently know of. The lithium
equivalent of oil shales. It doesn’t require too much imagination for the
development of an in-situ leaching operation extracting the lithium via pregnant
solution23.

Table 3: Lithium minerals of commercial importance.

Li 2 O
Chemical weight
Mineral composition percentage
Amblygonite (Li, Na)AL[(F,OH)|PO 4] 10.0
Lepidolite K 2Li4Al2[(F,OH) 2|Si 4O10 ] 2 5.0
Petalite LiAlSi 4O 10 5.0
Spodumene LiAl[Si 2 O 6] 8.0
Zinnwaldite K(Li,Fe,Al) 3(F,OH) 2 (AlSi3O10 ) 5.0
Source: USGS 2010, Janus Analysis

From a macro-sense, the cost curve can be deceptive, as the majority of brine
producers, though with a lower C1, typically do not initially meet battery-grade
specification (99.5% Li2CO3), therefore needing additional processing. Nor does
the cost-curve account for royalties, which, dependent on jurisdiction and price,
can be substantial. What is clear, amongst the vast majority of lithium producers,
is that brine and hard-rock production have costs substantially below spot-price
(~$12,000/t). Moreover, the cost curve is extremely flat, with little or no
differentiation between various primary suppliers.

Figure 40: C1 Lithium cost curve, differentiating mineral and brine producers.

Source: Citi (2018), Roskill (2018), Janus Analysis

23
Similar to that in Kazakhstan uranium operations, with similar grades and related
mineralogies.
30
 Lithium Demand & Supply

Disclosures
Janus Analysis Ltd is an appointed representative of Messels Ltd which is authorised and regulated by the Financial Conduct Authority of the UK.

Important Disclosure Statement from Janus Analysis Ltd.

This document is issued by Janus Analysis Limited solely for its clients. It may not be reproduced, redistributed or passed to any other person in whole
or in part for any purpose without written consent of Janus Analysis. The material in this document is not intended for distribution or use outside the
United Kingdom. This material is not directed at you if Janus Analysis is prohibited or restricted by any legislation or regulation in any jurisdiction
from making it available to you.

This document is provided for information purposes only and should not be regarded as an offer, solicitation, invitation, inducement or recommendation
relating to the subscription, purchase or sale of any security or other financial instrument. This document does not constitute, and should not be
interpreted as, investment advice. It is accordingly recommended that you should seek independent advice from a suitably qualified professional
advisor before taking any decisions in relation to the investments detailed herein. All expressions of opinions and estimates constitute a judgement and,
unless otherwise stated, are those of the author and the research department of Janus Analysis only, and are subject to change without notice. Janus
Analysis is under no obligation to update the information contained herein. Whilst Janus Analysis has taken all reasonable care to ensure that the
information contained in this document is not untrue or misleading at the time of publication, Janus Analysis cannot guarantee its accuracy or
completeness, and you should not act on it without first independently verifying its contents. This document is not guaranteed to be a complete statement
or summary of any securities, markets, reports or developments referred to herein. No representation or warranty either expressed or implied is made,
nor responsibility of any kind is accepted, by Janus Analysis or any of its respective directors, officers, employees or analysts either as to the accuracy
or completeness of any information contained in this document nor should it be relied on as such. No liability whatsoever is accepted by Janus Analysis
or any of its respective directors, officers, employees or analysts for any loss, whether direct or consequential, arising whether directly or indirectly as
a result of the recipient acting on the content of this document, including, without limitation, lost profits arising from the use of this document or any
of its contents.

This document is provided with the understanding that Janus Analysis is not acting in a fiduciary capacity and it is not a personal recommendation to
you. Investing in securities entails risks. Past performance is not necessarily a guide to future performance. The value of and the income produced by
products may fluctuate, so that an investor may get back less than he invested. Investments in the entities and/or the securities or other financial
instruments referred to are not suitable for all investors and this document should not be relied upon in substitution for the exercise of independent
judgment in relation to any such investment. The stated price of any securities mentioned herein will generally be the closing price at the end of any of
the three business days immediately prior to the publication date on this document. This stated price is not a representation that any transaction can be
effected at this price.

Janus Analysis and its respective analysts are remunerated for providing investment research to professional investors, corporations, other research
institutions and consultancy houses. Janus Analysis, or its respective directors, officers, employees and clients may have or take positions in the
securities or entities mentioned in this document. Any of these circumstances could create, or be perceived as creating, conflicts of interest. Janus
Analysis’ analysts are not censored in any way and are free to express their personal opinions. As a result, Janus Analysis may have issued other
documents that are inconsistent with and reach different conclusions from, the information contained in this document. Those documents reflect the
different assumptions, views and analytical methods of their authors. No director, officer or employee of Janus Analysis is on the board of directors of
any company referenced herein and no one at any such referenced company is on the board of directors of Janus Analysis.

Messels Ltd is authorised and regulated in the United Kingdom by the Financial Conduct Authority for the provision of investment advice. Residents
of the United Kingdom should seek specific professional financial and investment advice from a stockbroker, banker, solicitor, accountant or other
independent professional adviser authorised pursuant to the Financial Services and Markets Act 2000. This report is intended only for investors who
are 'professional clients' as defined by the FCA, and may not, therefore, be redistributed to other classes of investors.

The content of this report is covered by our Policy of Independence which may be viewed at www.messels.com.

Analysts’ Certification
The analysts involved in the production of this document hereby certify that the views expressed in this document accurately reflect their personal
views about the securities mentioned herein. The analysts point out that they may buy, sell or already have taken positions in the securities, and related
financial instruments, mentioned in this document.

31