Like Bitcoin, electric vehicles (EVs) are topics du jour and fascinating subjects. EVs, though, I find more interesting in that they have more moving parts and are a topic the shipping industry should be looking at. Short term, there is likely to be little impact. But beyond five years, I think we are going to see a marginal increase in related bulk materials needed for EV batteries as well as increased use in aluminum across the auto industry which will have an impact on demand for mid-size vessels, as well as a few larger bulk vessels.
The biggest headline to clue us in that bulk shipping will see increased volumes from EV growth is that up to 20,000 pounds or some nine tons go in to an average EV battery. Of course, not all these materials will move by ship. But in cases like aluminum we may see multiplier effects on many levels: aluminum in batteries, EV car bodies as well as for the network of charging stations. Near term EV production could rev up at 2m EVs a year, but within 10 years we are looking at 12-17m units. A high case 17m units (out of some 100m auto units) would imply some 150m tons per year of materials for EV batteries, which represents about 7.7% of about 2bn tons transported in minor bulk, and this figure does not include all of the related volumes in bauxite and alumina.
The first problem that hit me in the face upon reviewing hundreds of documents over the last month was that forecasts for EVs are all over the map. What goes into the batteries not only will change substantially and rapidly, but prices of these inputs are sometimes in short supply and we don’t understand the price dynamics at certain volume levels. But assuming most of these issues get solved one way or another, we are looking at volumes of bauxite and alumina, nickel, graphite, cobalt and lithium to go up very substantially.
In this shortened version of my findings I list two sections which may summarize many of the issues. First is a summary of issues in bullet form. Second is a list of materials and volume demand levels from EV battery growth.
First the bullets.
Facts to begin forming a view of where EV production could be in 2025, 2030, 2040 through to 2050:
• About 88m cars and light vehicles were produced in 2016
• About 115m cars and light vehicles could be produced in 2030, including impact of ride shares
• There were about 2m EVs on the road in 2016
• By 2018 China aims to be producing 1m/yr EVs, 3+m/yr by 2025 (or 5m on some forecasts)
• Could there be 12m EVs/yr by 2027 (or even 17m?) … and 20m EV sales/yr by 2030?!
• Half of new car sales by 2040 as EVs?… That would be 50+m/yr
• Battery factory capacity could rise from 68 GWh to 332 GWh by 2021
• 332 GWh would limit annual EV production in 2021 to about 3.4m/yr 300mile range equivalent EVs with current tech, not adjusting for PHEVs
• Some forecasts call for higher unit production at 4+m units by 2022
• Goal of 1,165 GWh capacity in a decade… or the 12m EVs/yr stat we see for about 2027
• It took ‘consumers’ about 30 years in the US to convert from horse and buggy to auto
• 20,000+ pounds or some 9-10 tonnes of materials can go into an EV battery
• Modules can be 5-7kg per kWh as finished battery product
• Currently we are looking at 100kWh for 300miles, or about 2x 500+kg battery packs for 600 miles
• -20c below or 48c above are not seen as friendly temperatures for EV batteries. Tesla appears to have built in cooling/heating, but Leaf EVs have experienced issues
• Currently only about 5% of Li-Ion batteries are recycled compared to 90% for car batteries
• Battery tech will see EV prices fall, but EVs pollute more than headline perception. ICEs may last a little longer…
• EV battery plants may be obsolete quickly and bring on write downs as soon as new technologies move to production.
• Carbon based batteries, among others, are just beginning to appear over the horizon
And, yes, we might know conceptually:
1. Demand forecasts, prices for materials like cobalt are all going to be quite different than we think now compared to in 10 years, especially price ranges.
2. Battery designs will shift radically – and yet makers have to commit on EV and EV battery factories and time periods to recoup at least some of the costs. So we will have some periods of fixed production designs.
3. Nickel, copper, bauxite, graphite, cobalt, lithium demand volumes may be higher in gross numbers than we are ready for, and used in different proportions per EV battery unit after production jumps from say 1m EV batteries per year to 5m then 20m.
4. No one seems to have proper near term EV/year sales forecasts. Many studies quoted appear to prefer far out 2040 forecasts.
5. Vehicle growth will slow as more autos become autonomous and shared, possibly 10% of autos sold (30% of miles driven in new cars) could be shared by 2030 (McKinsey 2016). This figure itself could be as much a stab in the dark as other forecasts.
Based on some of the above facts and possible outcomes, here is what I found may be the relationships with raw materials going into EV batteries and EVs.
About 60+% of nickel is used in production of steel in a 2m tonne production capacity market. Nickel in EVs may get to 15% of total demand or 300,000 tonnes per year by 2025, up from maybe 3% in 2017. Short term, nickel supply is ample. Longer term, who knows?
About 17kg of nickel products will go into a Tesla NCA battery, but only about 8kg in a NMC battery. Batteries in future are expected to use more nickel rather than less. There is some question as to how much battery grade nickel could go by sea, or processing locations. But likely there will be some industry displacements and unexpected shifting impacts.
In gross numbers, therefore, new demand could potentially require demand for a couple of additional handy bulk ships (or a little larger), historically often part of the global fleet of ‘trampers’. More interesting may be how demand patterns could change. Increased nickel EV demand could also put a squeeze on nickel used in stainless steel production.
Graphite will often ship in kegs or bags and can go by container or bulk. In bag form it can go in sacks up to one tonne each. Graphite can also move by barge, truck and rail. China remains major producer and polluter here. Graphite is moving partly to synthetic which can be made using petroleum coke, among other sources (and could still contribute to pollution indirectly and can be 10x multiple cost), but also new mines are seen opening up, such as Mozambique (reported 90% ready as of July 2017) with production capacity over 300,000 tonnes/yr.
Current output is about 1.3m tonnes/yr with China producing 66% in 2016, but only about 35% of consumption. What will output be in 2022 and 2027?
We should perhaps expect additional vessel needs, potentially a couple or a few per year closer to 2022-2027.
Graphite usage for EV batteries appears to run about 2-5x that of nickel, indicating 30%? of total demand could be made up by Graphite by 2025. There are 50-85kg in many EV batteries, with a Tesla S using about 85kg. At 12m EVs/yr this could represent up to 1m tonnes of new demand.
Cobalt is mined as a by-product of copper or nickel, and could be over 50% taken up for the production of EV batteries by 2025. Many in the market believe cobalt could be in shortest supply in coming years. An Tesla Model S 85-kWh battery needs about 8 kg of cobalt.
The Democratic Republic of Congo is the largest producer, but impact on shipment sizes appears negligible.
Cobalt is more of a price story than a volume discussion.
Lithium’s boom has been well followed for many years. Generally, most of its production is expected to go toward batteries, potentially as early as 2021-22. Lithium remains a big story, but impact on bulk may be mixed if Lithium carbonate higher grades can be sourced in country, although eventually Latin America is the largest source of reserves.
Nonetheless, any and all materials going into EV batteries, including materials we may not have yet focused on, could shift demand and supply patterns over 5-10 year periods. Some 60+kg of lithium carbonate go into a Tesla Model S. Mined lithium/lithium carbonate from sources such as Australia or Chile may travel by bulk. Lithium batteries could go by container vessel. Lithium in EV batteries as a percent of usage is espected to go from about 15% to about 40%. Some off the wall calculations, assuming no changes in technology, see global reserves of lithium going from 365 years to 17 years with full conversion of autos to EVs.
With about 16m tonnes a year in production, manganese ore, mostly from S Africa can certainly have a small impact on bulk shipping. Manganese could take up 61% compared to 4% lithium in a lithiated manganese dioxide (LMD) battery. Mainly used for Nissan Leaf and Chevy Volt at the moment. About 7kg of manganese go into NMC (Bolt) and LMO NMC (Leaf) batteries.
Copper production of 21.2m tonnes at 5.3% growth in 2016 continues to represent good cargo for the minor bulks. An EV battery can use 3x more copper than an ICE, or over 70 kilos. Copper concentrates transported in 2016 and 2017 are running about 29m tonnes based on reported numbers.
Production of over 12m EVs per year by 2025 compared to 2m/yr in the near future represents strong a significant driver for copper volumes. One report quoted by Reuters sees copper used in EVs rising from 0.18m tonnes/yr to 1.7m tonnes/yr by 2027 (IDTechEx commissioned by Int’l Copper Assoc).
EVs use between 2-3x more copper than ICE vehicles, or between 40-60kg per EV. EV trucks could use up 370kg of copper.
1m tonnes of new cargo by vessel could represent a few additional ships. But there are unanswered questions on needs for copper derivatives as well.
Bauxite/Aluminum (5.5 tonnes into 2 tonnes alumina and finally into 1 tonne aluminum)
Bauxite is a 100m tonne per year market for bulk shipping, with growth in 2017 expected at about 10%, but with year to year demand shifts potentially volatile, as with many commodities, but with higher volumes than most minor bulks. This market could more than double in 10 years.
Aluminum is looking at multiple avenues of increased demand: potential new avenues in battery technology; increased demand in EVs relative to ICE autos; and yet also increased demand in conventional autos. EV charging stations are also expected to focus on high aluminum content, and this likely is not added to forecasts yet.
There were about 180kg of aluminum per auto in US production in 2015. By 2025 this is expected to rise to about 235kg, an increase of about 30% in 10 years. The corresponding estimates in Europe are 150kg in 2016 going to 235kg by 2025.
It takes about 4-7 tonnes of bauxite for conversion into about 2 tonnes of alumina, which in turn are equal to about 1 tonne of aluminum.
A 50-80kg increase in aluminum is equal to about 350-400kg bauxite. This would represent an increase about 1 tonne of bauxite per 2.5 – 3.0 EVs or autos when comparing 2025 to 2015. The calculations can get a little silly when considering current shipping volumes and could result in a significant amount of increased vessel requirements in this niche category. 100m autos and light vehicles could be equivalent to 40m tonnes of increased bauxite demand and an 15m or so additional tonnes of alumina coming off of an increase in 7m additional tonnes in aluminum.
This works out to some 150 mid-size vessel equivalents or more before additional needs for related as well as other uses. Bauxite is increasingly going in larger bulk ships, so the maths will shift.