Is there enough lithium on earth ?
We often hear or read articles related to this topic, mainly using the number of cars, the size of the battery they are equipped with, an estimated lifetime for the cars (and/or batteries), and then comparing the outcome in kg of lithium with the known reserve around the world.
This calculation implicitly assumes that the car battery size is optimal, or at least well designed, and neglects the recycling of old batteries, which is today available and already reaching 99,5% at actual industrial mass process level.
These 99,5% recyclability means that once the worldwide car fleet will be completely electrical, the yearly need in new battery will be of 0,5% of the yearly fleet replacement, plus the net yearly grow of the fleet.
The battery sizing is another matter, much more depending on the way they are charged than the yearly mileage combined with the electricity consumption of the car.
Let us illustrate this through some examples.
Most of the cars, in Europe, travel yearly less than 15.000 km, which result in a daily average of less than 41 km. With a consumption of 16 kWh/100km, this means some 6,6 kWh a day.
Since the car is parked during at least 20h by day, it does not require more than 345 W (yes W and not kW) of charging power, considering a charging yield of 95%.
If we double these figures, to be able to afford special days where the travelled mileage will be double, we get a battery size of a bit more than 13 kWh and a charge of 1,3 kW (13 kWh in 10 hours, I know the number of hours should not be divided by 2 but anyway, this is conservative and remains low).
These elements reveals that the 3 important parameters are: mileage by day on a peak day (rather than a yearly average), the car consumption and the charging strategy.
The first example, illustrated above, is probably exactly not representative of the reality (statisticians would probably classify it as a stochastically impossible event). Even in Belgium, where the distances are short, a lot of commuters drive 120 to 150 km on a daily basis (over their workweek). While other people are only using their car in the weekend. So, averages are always distorting the analysis.
Yet, a commuter who travels 150 km a day, quite stressed by the traffic density, will consume rather 20 kWh/100 km, thus will need 30 kWh by day. The car will probably be driven during 5 or 6 hours, leaving 18 hours for charging, requiring 1,67 kW of charging power (same result with the half of the distance, one way, and the halve of the time which matches the daily time spent at work premises).
Now, a lot of people are using their car to go on holiday or for a weekend out. We deal here with exceptional days, where the total mileage can be of 1.000 km. The question is here whether the car battery must be sized to cover this distance, or could be charged in little time. Is it sound to equip the car with batteries that are 5 to 10 times the real need, except for 2 days a year ?
For this point the charging time is key. Everybody could be happy with a battery sized for the common days, if it could be fully charged (thus from nearly 0% till nearly 100% charge) in a couple of minutes.
Taking into account a recommended stop after 3 hours driving, thus after roughly 360 km (on a fluid highway), batteries of 60 kWh will do the job, 40 kWh would be a valuable alternative with a stop every 2 hours.
Further, the weight of the car is an important factor influencing its energy consumption, and batteries (even the famous lithium-ion ones) are quite heavy (150-200 Wh/kg, thus roughly 200 kg for 40 kWh), so that a smaller battery will allow lower consumption, what will enable to drive more km between 2 charges (relatively to the battery capacity).
Aerodynamic plays a major role when driving on highway (rather from 50 km/h up).
The power train yield, thus the ratio of the energy delivered to the wheels divided by the energy offtaken from the battery, impacts, in turn, every travel.
The lowest losses would be reached with motors placed in the wheels (in wheel motor), fed by high voltage (thus low current and low ohmic losses) system.
The lowest consumption figure, as of today, observed on long distances with a commercial car, has been achieved by a Lightyear, and was of 8 kWh/100 km. Less than the half of the average of the existing electric cars consumption. This enables to more than double the achievable distance with the same battery when a car consumes 16-20 kWh/100 km.
Of course, charging 40 kWh in a battery, in a few minutes is a challenge.
Usually when one searches for a solution to a situation, she/he transposes the present existing one (the known paradigm) for a similar case to the new one, but adapted for the new situation. So written is it quite abstract, but let us apply it to our subject.
For charging, this consists in developing chargers able to charge the battery in, let us say, 5 minutes. Roughly calculated, the power necessary to supply 40 kWh in 5 minutes is of 40 kWh × 60’/5’ = 480 kW, already quite huge. But, since the charging of a battery shows a (counter exponential) saturation curve (similar to: A×[1 – e-kt] with A an k constant an t the elapsed time since the start of the charging), the time to reach 95% of the full charge will be rather the triple (thus 15 minutes), or conversely the power necessary to charge this battery in 5 minutes would be of around 1.500 kW (1,5 MW). Enormous ! can you imagine a charging station along a highway where 20 chargers are installed: 30 MW, the power of a turbojet generator. Crazy !
And, anyway what you do in (possibly revolutionary) chargers innovation, the physics is there, not any possibility to reduce this power.
No solution thus ? Fortunately, there is (or are), but, as always, the real solution to a new problem requires another sight on the matter, another way to approach the need.
Geely (no, I have not any share in this company) developed and has already installed thousands of commercial battery exchangers, which are capable of exchanging the (standard) battery of its cars in a bit less than 1 minute. The exchanger is similar to a small car wash box (just a bit longer, larger and taller than a car), with a robot underground which exchanges the (discharged) battery by another (charged) one, in 59 seconds. The drivers pay only for the difference in charge (thus the equivalent of kWh that would have been supplied to charge the battery).
The advantages are multiple. I will just mention 3 very important ones. First, the battery charging (rather replacement) occurs in much less time (1 minute), even less than the usual tank time of a thermal car. Next, the (exchanged discharged) batteries can be charged continuously (24 × 7), spreading the charge over time and also enabling its distribution geographically without requiring enormous (and thus expansive, or even more probably, unrealistic) grid adaptations. Finally, the battery would no longer make part of the car ownership (neither do the fuel in the fuel tank, the tank self not being in the ownership would not change a lot in the consideration), so that the cars can also be less expansive. The battery can be a kind of hiring service (battery + access to guaranteed exchange service), a fixed monthly fee thus, of course not including the consumed kWh (or well, like Tesla used to).
This system does, on another scale, what you do when you change the battery of your (computer) mouse.
Now, with all these elements in mind, how much lithium is necessary to equip all the cars, have enough battery in stock for the exchanges, and replace the 0,5% recycling losses (after end od live) ?
Over the world, there are roughly 1,5 billion cars (250 million in the EU). But in order to estimate how much lithium is necessary we have to take into account the specials (days. Those days on which a lot of batteries will be exchanged (swapped).
For the transhumance days, let us make the following (conservative) assumptions:
- the travelling distance to destination is of 1.000 km (it is difficult to drive more than that on a single day)
- 1/3rd of the people is travelling this distance (one way) by car
- the transhumance is spread over 6 weekends, with 2 peak weekends showing a double concentration, thus 4 weekends with 1/8th of the transhumance and 2 weeks with 1/4th of the transhumance
- the transhumance occurs (one way) on the same day for everybody (thus concentrated on 1 day of the weekend)
we have 1,5 billion × 1/3 × 1/4 = 125 million cars in the peak weekend. At the same time 1,25 billion are driven as usual, which can be charged with low power chargers as explained above.
A consumption of 16-20 kWh/100km, thus 160-200 kWh for 1.000km, applied to 125 million cars gives 20-25 TWh. Considering 200kWh (conservative figure) and that the car starts with a completely charged battery, the travel will require 160 kWh, thus 20 TWh (20 billion kWh) worldwide.
A lithium-ion battery contains 113g of lithium, for each kWh of capacity. Thus 20 TWh will require 2,26 million ton of lithium.
And 1,5 billion cars equipped with a 40 kWh battery (thus 60 TWh) will require 6,780 million ton of lithium.
The total is thus of 2,26 + 6,78 = 9 Mt of lithium.
The worldwide terrestrial estimate lithium reserve lays between 26 and 98 Mt, thus taking the 26Mt into account, the reserve represents 3 times the needs. No problem at that level, provided applying the battery exchange system.
The global figures are O.K, we seem to be safe. Really ? Disposing of enough reserve is a good point, but is the production capable to follow the transition curve ?
Using a binomial distribution for the transition over 10 years (with a probability factor of 0,5), in the most solicited year (year 5) 25% of the thermal cars will be replaced by electric ones. This will require a production capacity of 6,780 Mt × 25% = 1,7 Mt, while today it lays around 43kt/year, thus 1/40th (43.000/1.700.000 @ 1/40) of the necessary level. In reality it is still worst, because we have to cover the transhumance needs on top (for already 50% of the fleet, in year 5), thus 1/65th of the necessary level.
So, we are facing a huge challenge, but not at reserve level, well on production capacity level.
An upscaling factor of around 40 to 65, means that the competition is still very broadly open, the race is not yet won by any party (nor any country) and also that the whole potential of innovation will be welcome to succeed in it.
The global picture goes much further than the production of lithium, it includes also the development of energy efficient cars (sobriety), new battery generation (lighter), charging infrastructure paradigm and deployment… and, why not, a change in the mobility paradigm (shared travels).