Sodium Ion Batteries for Vehicles // Analysis

Show video

Welcome back everyone! I’m Jordan  Giesige and this is The Limiting Factor.  In my first video on Sodium Ion batteries, I said  that CATL’s upcoming Sodium Ion batteries were   well-suited to grid storage. That led to some  pushback in the comments that emphasized the   viability of sodium ion batteries for vehicles. In my view, sodium Ion batteries are definitely   viable for vehicles, but that’s dependent on  the type of vehicle that you’re referring to,   whether you’re talking next year or 5-7  years from now, and how other chemistries   on the market will develop during that timeframe. So today, I’ll walk you through my thinking and   give you a high level overview of how I expect  the sodium ion battery market to develop for   vehicles between now and 2030, how that might fit  in with Tesla’s plans for a robotaxi or small car,   and why all that doesn’t change the fact  that sodium ion batteries are better suited   the much larger energy storage market. Before we begin, a special thanks to my   Patreon supporters and YouTube members. This is  the support that gives me the freedom to avoid  

chasing the algorithm and sponsors. As always,  the links for support are in the description.  What seems to have first triggered the discussion  on Sodium Ion for vehicles is the fact that CATL   has advertised that their first generation Sodium  Ion cells are hitting 160 Wh/kg of energy density   at the cell level and that in the next year or two  they expect to get that up to 200 Wh/kg. This is   roughly the same gravimetric energy density that  CATL is getting out of their LFP battery cells. 

However, there’s a reason why CATL is advertising  the gravimetric energy density and not the   volumetric energy density. Although there aren’t  publicly available volumetric energy density specs   for CATL’s sodium ion battery chemistry, based on  similar chemistries, I figure their energy density   is around 260-320 watt hours per liter at 160-200  Wh/kg, or 35% less than an LFP battery cell.  This is the primary reason why I said that  sodium ion is better suited for grid storage   than electric vehicles. For most electric  vehicles, volumetric energy density is the   first or second priority because the more space a  battery cell takes up for a given energy density,   the fewer cells that you can squeeze  under a vehicle, which limits range.  For grid storage, the space that the battery  packs take up doesn’t affect their commercial   viability and the priority is cost per kwh  per cycle. Commercial energy storage is all  

about cost control and that’s where sodium  will absolutely dominate other chemistries.  If all those assumptions are true, then  it should follow that the majority of   sodium ion battery cells will be used  for grid storage rather than vehicles,   and that’s where things appear to be headed.  ICCsino, a Chinese industrial consulting company,   estimates that there will be 160 GWh of  Sodium Ion production by 2026. Of that,  

70% is forecast to go into energy storage, 11.3%  into scooters, and only 18.3% into vehicles. That   is, sodium ion can be used for vehicles, but it’s  really more of a side show than the main event.  With that said, things will get more interesting  over time in the vehicle space. Let’s look at how   the use cases for sodium ion in vehicles will  evolve from now until 2030 for a variety of use   cases. I’ll start with the Chinese market, then  Western markets using the Model 3 as an example,   and then Tesla’s upcoming compact vehicle. There’s two vehicles that have shown up in  

reporting out of China. The first  is a test vehicle created through a   joint venture by Volkswagen and JAC, which  uses a Hina Sodium Ion battery. That is,   this vehicle isn’t in production yet. It has  252 kilometres of range, but that’s on a Chinese  

test cycle. On an EPA test cycle, that would be  more like 187 kilometres or 116 miles of range.  Hina advised that the first vehicles to use  sodium ion batteries will be A00 class EV’s   like the VW/JAC test vehicle. A00 class EVs  are ultra-compact vehicles that weigh less   than 800 kilograms. They didn’t say why sodium  ion is suited to cheap ultra-compact vehicles,   but in my view, it’s for 3 reasons: First, small vehicles are more efficient   and can squeeze more range out of the low  energy density sodium ion battery packs.  Second, LFP batteries for the time being will  be considered a more premium battery option than   the sodium ion batteries because they’ll offer  more range. That means the LFP batteries will   be used in vehicles that are larger, more  expensive and have larger profit margins. 

Third, when sodium ion batteries do  hit scale, they’ll be cheaper and   therefore better suited to budget vehicles. Let’s look at the next vehicle example:  The BYD Seagull is rumoured to be the first  vehicle in China that’ll offer a sodium ion   battery. Although BYD denied the rumours  and I don’t like operating off rumours,   I think it still serves as a good example. The sodium ion version of the vehicle is expected   to have 300 kilometres of range versus the 400  kilometres of range for the LFP version. That’s   about 138 vs 184 miles of real world range, which  is a 33% range hit for the sodium ion battery   pack and exactly what we’d expect as a result of  sodium ion’s lower volumetric energy density. So,  

even if the rumour isn’t true, it’s at least  technically feasible. The cost information   from the rumour is also feasible. The rumour indicates that, currently,   at low volume, sodium ion batteries are $96/kWh  and that BYD hopes to get that price down to   $74 per kilowatt hour when they hit high  volume production. Those prices roughly  

align with prices I’ve seen touted by other  sodium ion companies. BYD doesn’t specify what   high volume means, but I’m assuming they  mean in the low tens of gigawatt hours.  That is, at low volume, Sodium Battery cells  will actually be slightly more expensive than   LFP batteries cells, which have a production cost  of between $75 to $85 per kilowatt hour. Then,   at volume production, they’ll be roughly the  same price. But, that’s just the beginning.  After that, with each cumulative doubling of  production, which should happen every 1-2 years,   I expect the price should decrease by about 18%. I  chose 18% because that’s the Wright’s Law learning   curve for batteries. With that in mind, by  the end of the decade, the production cost  

of sodium ion battery cells using primarily  iron and manganese will probably bottom out   at around $40 per kilowatt hour, which would  around $50 per kilowatt hour at the pack level.  So yes, sodium ion can and will be used  in vehicles, but I expect that the first   vehicles to use sodium ion batteries will  be budget Chinese commuter vehicles which   need to be as cheap as possible and  where range isn’t a big consideration.  Then as the energy density of sodium ion  increases, it’ll expand beyond cheaper   Chinese commuter vehicles to Chinese  Sedans and vehicles in western markets.  On that note, let’s move on to  markets like North America, Europe,   and Australasia. What’s the best case scenario  for sodium ion in a vehicle like the Model 3?  Let’s take the 260-320 Wh/l figure I estimated  for CATL’s sodium ion batteries and combine it   with their current and upcoming battery  pack technologies, which are referred to   as cell to pack 2.0 and 3.0. Cell to pack 3.0 is  also known as Qilin. These pack technologies have   packing efficiencies of 55% and 72% respectively. Yes, CATL did say that the packing efficiency of  

the Sodium Ion battery could be over 80% due  to the fact that it’s such a safe chemistry,   but I view that as a long term aspiration rather  than something we’ll see in the first generation.  After factoring the 260 to 320 Wh/l figures with  cell to pack 2.0 and 3.0, the pack level energy   density ends up between 143 to 230 wh/l. If we  use the long range Tesla Model 3 as a benchmark,  

that would provide about 214 to 345 miles of  range. However, that comes with some big caveats.  First, the estimates on screen don’t take  into account the extra weight of the lower   energy density packs, which would handicap  the upper end of the range figures. Second,   most of the estimates are based on marketing  material or internal strategy documents,   which I’ve come to find are highly  aspirational for all battery companies,   Tesla included. Let’s look at the figures  extracted from CATLs marketing material and   internal documents vs reality with two examples. First, the Standard Range Plus Model 3 using CATL   LFP battery cells has 272 miles of range. That  battery pack should be getting a bare minimum of   285 miles of range if the marketing material was  correct. And as you can see in the video on screen  

from Munro and Associates, except for the row of  cells removed for analysis, with this architecture   the pack appears to be at its cell capacity limit. Second, CATL advised that their Qilin battery pack   would be capable of hitting 255 Wh/kg with a  high nickel chemistry. The first implementation   of the Qilin battery pack with a high nickel  chemistry by Zeekr is said to be hitting 200   Wh/kg. That’s an excellent number and the best  on the market as far as I’m aware, but it’s only   about 10% better than a Tesla, despite CATL’s  claims that they would be hitting 255 Wh/kg,   which would mean besting Tesla by 30-40% With that in mind, I’d be surprised if the   first generation of CATL’s Sodium Ion battery  pack hits Tesla’s minimum requirement of 250   miles of range in a Tesla Model 3. In fact, if  it hits 200 miles of real world range in any   vehicle this year, I’d be very impressed.  So, in my view, it’s likely we won’t see  

a CATL Sodium Ion battery in a Tesla Model 3  until the 2nd or 3rd iteration of the pack,   which may not happen until later this decade. For an entry level luxury vehicle like the   Tesla Model 3, instead of using Sodium Ion, I  see it as more likely that CATL builds an LFP   battery pack that provides up to or even beyond  300 miles of range. LxFP is also a possibility,   but it might initially increase the pack cost  and could have eccentricities such as higher   resistance and reduced cycle life compared to  LFP. By LxFP I mean derivatives of LFP that   are doped with elements such as manganese to  increase voltage and therefore energy density. 

However, many people are wondering about the  viability of Sodium Ion for the upcoming small   Tesla or Robotaxi that Tesla’s been teasing for 3  years. For the purposes of this video, I’ll call   that car the Tesla mini. As this electric vehicle  core efficiency chart shows, the smaller the   vehicle the greater the efficiency. That’s because  smaller vehicles weigh less and have less drag.  If we extend the efficiency line for Tesla up  and to the left for a Tesla mini, we’d easily be   looking at over a 15% improvement to range. Then,  if we take my low end figure of 214 miles of range   for a sodium ion battery pack for a Model 3 and  add 15%, that could mean a Tesla mini using sodium   ion would hit 246 miles of range, which would  be acceptable range for an urban commuter car.  Given that Sodium Ion batteries are expected  to go into GWh scale production this year when   CATL forms a basic industrial supply chain  for it, and given many people in the Tesla   community are expecting a Tesla mini class  vehicle in 2024, what are the odds that   a Tesla Mini would use sodium ion batteries?  I’d put the odds at less than 10%. Heres’ why: 

Tesla’s said that they expect their next  generation platform will be produced at   higher volume than all of their vehicles combined,  that they’re on a 2 for 1 target, and that they’re   trying to get to a production cost 50% below  the Model 3 and Y. In other words, the next   generation vehicle will be small and produced  at a volume of millions of units per year.  Tesla’s Model Y lines currently produce  about 250,000 vehicles per year. Those   production lines will likely be able to  produce 500,000 vehicles per year when   fully staffed. If production effort is cut in  half, we might expect Tesla’s next generation   lines to produce 1 million vehicles per year.  As for the size of the battery pack, I expect  

it will be about 40 kWh, plus or minus 5 kWh. If the vehicle launches in 2024 and we assume   12-18 months to hit volume production at the  first factory in Mexico, that means each Tesla   Mini line could require 20 GWh of total battery  cells by 2025. I say each line because if Tesla   wants to continue growing at 50% per year, they  may need to either concurrently or successively   launch multiple lines in North America, Asia,  and Europe. If that’s the case, the 20 GWh   figure could be conservative and it might be more  like 20 to 60 GWh of cells that are required.  Whether it’s 20 or 60 GWh of cells in 2025, it  seems doubtful that Tesla could secure all the   sodium ion battery supply it would need for a ramp  that aggressive. There’s several reasons for that.  First, there are a number of players entering  the sodium ion battery space and CATL won’t   have a monopoly. That is, there may not be  a single manufacturer that will be able to  

supply the 20 to 60GWh of sodium ion batteries  Tesla would need in 2025 for a Tesla mini.  Second, if that’s the case, it’s doubtful Tesla  would be willing to use multiple sodium ion   battery packs from multiple manufacturers  for the Tesla Mini when they could just   use one LFP battery pack from one supplier. Third, even if I’m wrong and one supplier alone,   such as CATL, could supply 20-60 GWh  of cells in 2025, cell suppliers tend   not to put all their eggs in one basket for  reasons that range from stability to profit.  Fourth, and finally, the next generation Tesla  will be as important as the original Beetle to VW,   or the Corolla to Toyota. Whether it’s  one manufacturing line or several,   it doesn’t make sense to me that Tesla would  make the launch of arguably their most important   vehicle ever dependent a sodium ion chemistry  that’s never seen scale production before.  To wrap things up on the Tesla mini, to me, the  chemistry that makes the most sense for the first   few years of production of the Tesla Mini is tried  and trusted CATL LFP battery cells. LFP is CATL’s  

flagship chemistry and they’re scaling like  mad. 20 GWh or even 60 GWh would be a fraction   of their annual growth in production capacity  from 2024 to 2025. Then, later this decade when   sodium ion does hit serious scale, Tesla could  offer it as an even cheaper alternative to LFP.  On that note, when sodium ion does hit the current  energy density of LFP, will it eventually dominate   the EV market in the late 2020s and into the  early 2030s? In my view, no. To understand why,   let’s zoom out to get a broad, long-term view of  battery cell chemistries, demand, and use cases. 

Tesla provided us with two slides on  battery day about where they see the   demand for battery cells. The first slide  was for renewable energy, or grid storage.  They expect that 10 TWhs of annual production  capacity is required. So long as sodium ion   lives up to its promise, there’s no reason why  it can’t eventually be used for all these grid   storage applications and dethrone LFP as the  king of energy storage. As I said earlier,   that’s because sodium ion will have  the lowest cost per kWh per cycle of   any commercialized battery chemistry. Based on a  quick estimate, it would cost 80% less than an NMC   chemistry and 33% less than an LFP chemistry. Moving on to the second battery demand slide,  

Tesla showed that another 10 TWhs of battery cells  are required to electrify transportation. Which   of these vehicles can sodium ion be used for? The  Semi needs an expensive high energy density Nickel   chemistry to maximise its commercial viability. Luxury sedan, SUV, and long range truck owners   will demand maximum range, so they’ll also use a  nickel chemistry. That won’t always be the case,   particularly in China, but for Europe, North  America, and Australasia Nickel chemistries   will be the go-to chemistry for mass  sensitive and long range vehicles.  The last 5 GWh’s for small to mid-sized vehicles  is where things will be the most dynamic. For the   next couple of years, I expect LFP to remain  king for smaller classes while sodium ion goes   through an early scaling and product verification  phase in ultra-budget Chinese vehicles. Then,  

from 2025 to 2030 sodium ion will work its  way into most small vehicles in China and   then into small vehicles in Western markets. However, what about mid-sized vehicles like   the Model 3 and Y? As I suggested earlier, in  the next 2-4 years, sodium ion doesn’t make   sense for entry level luxury sedans and SUVs.  This is because as battery chemistries improve,   those customers will likely prefer improvements  to range over a cheaper vehicle with less range.   This will be especially true when Tesla  starts making a cheaper vehicle like a   Tesla mini to fill the commuter car niche. Beyond the next 2-4 years it’s difficult to   say how the market will evolve for entry level  luxury vehicles. But what I do know is this,   there’ll be a variety of chemistries on offer  and by 2030 sodium ion may be able to offer 300   miles or range or more in a mid-sized sedan.  If that’s the case, I’ll take a wild stab in  

the dark and say that by 2030 Tesla will offer  a sodium ion option for the Model 3 or Y that   gets more than 300 miles of range while  costing around $35,000 in today’s dollars.  In summary, the maximum market opportunity  for Sodium Ion appears to be about 30% of   the electric vehicle market, or 3 terawatt hours  by 2030. As the technology scales and improves,   between now and 2030, it’ll work its way from  budget electric vehicles in China to budget   vehicles in western markets and potentially  into mid-sized sedans. That’s in contrast to   energy storage, where the market opportunity  is the entire market, or 10 terawatt hours.  However, to saturate the total market  opportunity for vehicles and energy storage,   it will probably take until the end of the 2030s.  I think in the last video some people viewed the  

fact that I said that sodium ion will take time  to scale as slight against CATL or sodium ion   batteries. That certainly wasn’t the intent. I don’t think we can make a rapid transition   to sustainable energy without sodium ion.  Forecasts indicate that battery material   constraints will begin to weigh on the industry in  the next few years and due to those constraints,   growth in lithium ion battery cell supply  will start to wane. The materials required   for sodium ion batteries are abundant  rock forming elements, which would take   pressure off the materials constraint problem. It’s just the nature of manufacturing and the   scale involved here that means sodium ion won’t  take over the market overnight. It’s going to  

take millions of people, billions of tonnes of  material, and well over a decade to create 13   terawatt hours of annual battery cell production. On that note, one of the things I’m most excited   about this year in battery technology is to find  out how much capacity CATL is installing for their   first major sodium battery cell factory. When we  know that, we can start staking out with greater   accuracy what the S-curve for sodium ion will  look like and I’ll be happy to adjust my bullish   forecast if it turns out I’m wrong. But there’s  a good chance that revision may be downward.  If you enjoyed this video, please consider  supporting me on Patreon with the link at   the end of the video or as a YouTube member.  You can find the details in the description,   and I look forward to hearing from you. A special thanks to Tesla Boomer Mama for  

your generous support of the channel,  my YouTube members, and all the other   patrons listed in the credits. I appreciate  all your support, and thanks for tuning in.

2023-04-20

Show video