The comeback of LFP batteries

ETC battery energy density evolution

Once upon a time, not many years ago LFP (LiFePO4) batteries were very popular, including for electric vehicles. Most LFP batteries were produced by Chinese battery cell makers such as Thunder Sky Winston Energy, but it was a North American company that made this battery technology famous.

A123 Systems was the battery cell maker that made great improvements to the LFP battery technology – at a time when Chinese companies only copied things without improving them.

This battery cell maker was able to increase the energy density, while at the same time increased the power energy density and the cycle life of the cells. I always say that developing batteries requires compromises, but A123 Systems successfully improved the LFP technology in every single way.

Unfortunately A123 Systems battery technology never became mainstream. When the company was preparing mass production it was plagued with quality control problems and early on GM replaced it by LG Chem as its battery cell supplier. Then the company declared bankruptcy and was purchased by the Chinese auto parts maker Wanxiang Group.

This meant that for a long time LFP developments stopped, while NCA and NCM chemistries kept improving. Let’s see some differences between mainstream cathode chemistries.


Lithium Ferro PhosphateΒ (LFP)

  • Energy density: (β˜…β˜…) 2/5
  • Power density: (β˜…β˜…β˜…β˜…) 4/5
  • Cycle life: (β˜…β˜…β˜…β˜…) 4/5
  • Safety: (β˜…β˜…β˜…β˜…β˜…) 5/5
  • Cost: (β˜…β˜…β˜…β˜…) 4/5


Lithium Nickel Cobalt AluminiumΒ (NCA)

  • Energy density: (β˜…β˜…β˜…β˜…β˜…) 5/5
  • Power density: (β˜…β˜…β˜…) 3/5
  • Cycle life: (β˜…β˜…β˜…) 3/5
  • Safety: (β˜…β˜…) 2/5
  • Cost: (β˜…β˜…β˜…β˜…) 4/5


Lithium Nickel Cobalt Manganese (NCM 333 or 111)

  • Energy density: (β˜…β˜…β˜…) 3/5
  • Power density: (β˜…β˜…β˜…) 3/5
  • Cycle life: (β˜…β˜…β˜…β˜…) 4/5
  • Safety: (β˜…β˜…β˜…β˜…) 4/5
  • Cost: (β˜…β˜…) 2/5


Lithium Nickel Cobalt Manganese (NCM 523)

  • Energy density: (β˜…β˜…β˜…β˜…) 4/5
  • Power density: (β˜…β˜…β˜…) 3/5
  • Cycle life: (β˜…β˜…β˜…) 3/5
  • Safety: (β˜…β˜…β˜…) 3/5
  • Cost: (β˜…β˜…β˜…) 3/5


Lithium Nickel Cobalt Manganese (NCM 622)

  • Energy density: (β˜…β˜…β˜…β˜…) 4/5
  • Power density: (β˜…β˜…β˜…) 3/5
  • Cycle life: (β˜…β˜…β˜…) 3/5
  • Safety: (β˜…β˜…β˜…) 3/5
  • Cost: (β˜…β˜…β˜…) 3/5


Lithium Nickel Cobalt Manganese (NCM 811)

  • Energy density: (β˜…β˜…β˜…β˜…β˜…) 5/5
  • Power density: (β˜…β˜…) 2/5
  • Cycle life: (β˜…β˜…) 2/5
  • Safety: (β˜…β˜…) 2/5
  • Cost: (β˜…β˜…β˜…β˜…) 4/5


As you can see LFP seems almost perfect for every application, but with poor energy density isn’t suitable for electric cars.

Fortunately, for a while now the Chinese government determines the amount of subsidies for electric vehicles depending on range and battery pack energy density. Requirements for electric passenger cars are more demanding than for electric buses.

This made most battery cell makers completely replace LFP with NCM cathodes in batteries for electric passenger cars, however a few decided to improve LFP. That was the case of BYD, which introduced manganese to the cathode and managed to increase the energy density to 165 Wh/kg. However, this figure still isn’t enough to get the maximum subsidy, more improvements need to be made.


Lithium Ferro Manganese PhosphateΒ (LFMP)

  • Energy density: (β˜…β˜…β˜…) 3/5
  • Power density: (β˜…β˜…β˜…β˜…) 4/5
  • Cycle life: (β˜…β˜…β˜…β˜…) 4/5
  • Safety: (β˜…β˜…β˜…β˜…β˜…) 5/5
  • Cost: (β˜…β˜…β˜…β˜…β˜…) 5/5


ETC is another Chinese battery cell maker that is successfully improving the energy density of this battery technology.

BYD and ETC share the same goal of reaching 200 Wh/kg in batteries with a LFMP cathode and a silicon/carbon anode, when this happens we’ll finally get cobalt free batteries with decent energy density, that are extremely safe, durable and cheap.


ETC roadmap cell planning


Let’s compare the future LFMP battery cell made by ETC to Samsung SDI prismatic battery cells used in BMW i3.


Future LFMP battery cell by ETC

  • Gravimetric energy density: 200 Wh/kg
  • Volumetric energy density: 417 Wh/L
  • Cycle life: 4.000 cycles


Samsung SDI 94 Ah battery cell

  • Gravimetric energy density: 174 Wh/kg
  • Volumetric energy density: 352 Wh/L
  • Cycle life: 4.600 cycles


Samsung SDI 120 Ah battery cell

  • Gravimetric energy density: ???
  • Volumetric energy density: 449 Wh/L
  • Cycle life: ???


The specs of the future LFMP battery cells from ETC seem suitable for electric cars. They are very similar to Samsung SDI prismatic NCM cells, with the advantage of requiring no cobalt.

Let’s see the roadmap of ETC in more detail.


ETC roadmap


For now just ignore those interesting NCM battery cells from ETC, especially the ones in PHEV2 format. The LFMP cells from ETC that we want are expected to become available during the second or third quarter of 2020. Therefore, we are about a year away until we see if ETC achieved its goals and its corporate customers can finally use LFP/LFMP battery cells not only in electric buses, but also in electric cars.


ETC customers


A good thing about Chinese battery cell makers is that sooner or later, their cells become available on AliExpress for private buyers. Corporate buyers get the cells first, but for example private buyers like you and me can already buy the second generation cells for around 190 euros per kWh.


Summing up…

In this article I tried to show why I think that when LFMP reaches an energy density of 200 Wh/kg it could be a good cobalt free battery technology for many applications, including electric cars. I’m curious to know which battery cell maker will reach the goal first. BYD has more resources than ETC, so it has an advantage.

However, while LFP/LFMP developments look promising and will likely be enough for the technology to regain its lost space in EV batteries, NCA and NCM are still the most energy dense chemistries and will continue to evolve until we get a mixture of the two chemistries called NCMA.

NCMA cathodes get the best of modern NCM and NCA chemistries, they are a real breakthrough and a exception to the rule that battery technology developments almost always require compromises. You can get an idea with the chart below.

Performance of different advanced battery cell cathodes


I can’t promise a date, but I’ll write an article exclusively about NCMA cathodes, their advantages and why I think they’ll be the final nail in the coffin for ICE (Internal Combustion Engine) cars. Good things are coming.



More info:

Pedro Lima

My interest in electric transportation is mostly political. I’m tired of coups and wars for oil. My expectation is that the adoption of electric transportation will be a factor for peace and democracy all over the world.

19 Responses

  1. Rafael says:

    Muchas gracias por el gran articulo Pedro Lima. En estos dΓ­as se esta hablando de nuevo de las celdas de grafeno de Samsung “Usted como no, hizo un gran articulo sobre el tema” Hablan de que Samsung esta luchando para aumentar la capacidad de la celda y reducir costos ΒΏCree usted que acabara llegando al mercado para el prΓ³ximo 2020?. Respecto al ΓΊltimo clavo de los ICE creo que esta practicamente dado……Sion Power ya tiene celdas comercializables de 500wh/kg-1000wh/l y 500 ciclos “antes eran 450 pero su CEO hace unos dΓ­as dio la cifra de 500 ciclos”. Y ojo tienen un PDF muy completo donde dan detalles de unas con 700wh/kg y 200 ciclos. Voy a poner fechas para 2030 las ventas de ICE seran testimoniales. Pedro Lima le voy a dar una sugerencia podria hacer un artΓ­culo a largo plazo hablando de fabricantes de celdas y su evoluciΓ³n en tecnologΓ­a de Celdas. Se que es mucho trabajo pero seria un artΓ­culo muy brillante.

  2. Pedro says:

    Obrigado, nunca me tinha apercebido do conceito de densidade de potΓͺncia atΓ© agora:
    “For example, a tiny capacitor may have the same power output as a large battery. However, since the capacitor is so much smaller, it has a higher power density. ” (

    • Pedro Lima says:

      That’s normal, most people never heard about power energy density. It’s more important for hybrids that have small battery capacity than for full-electric cars.

  3. Hi Rafel, need your suggestions and insight on a couple of things.
    So we are working on a swappable battery ecosystem for Indian 2 and 3 wheeler requirements.
    1.5 kWh battery pack, that can work for 2 wheelers and 3 wheelers in single and multiple configurations.

    In any case, the vehicles do not have hi-speed or high torque requirements as max speed is less than 50 Kms per hour. What is more important is high cycle life of the batteries (2000 and above).

    As far as I understand, NCM or NCA is not viable chemistries for us due to limited cycle life. LFP/LFMP looks much adequate for such a solution.
    Can you recommend which cells should we use for such an application where cycle life is more important than charge density?

  4. Manuel says:

    Bring on the NCMA cathodes article. You made me very curious, as I’m sure many people are. Thanks!

  5. Gaston says:

    Strange enough, the ETC battery website do not mention LFP or I did not look right
    Have you try to directly have info from them ?

  6. Christian says:

    Competition is always good and will make sure we are not too much dependent on some materials. They will keep the cobalt price stable too.

    These parallel advancements are very welcome.
    Great article, Pedro.

  7. Marcel says:

    Thanks for the article Pedro. The LFMP tech is very exciting. Hopeful ETC can start producing them at scale. This kind of thing could be a tipping point for electrifying local delivery trucks, city buses and school buses. That would make serious dents in oil demand and air pollution.

    Do you think the NMCA technology is further away than the LFMP?

    And it seems like Tesla has something up their sleeve too, so their Battery Day (I think that’s when people think they’re going to announce something) is something to look forward to, too.

    • Pedro Lima says:

      Hello Marcel. As Christian wrote, it’s good to have alternatives. Especially now that cobalt and nickel prices are rising. NCMA still requires some cobalt a lot of nickel.

  8. Gareth Claase says:

    Fascinating articles thank you

  9. Ugo says:

    “at a time when Chinese companies only copied things without improving them”.
    Amusingly, A123 Systems was a spin off of US tech created by a Chinese researcher at MIT.
    Or, China is very good at “stealing” US IP created by Chinese

    • Pedro Lima says:

      One thing that US companies used to do very well was recruiting the best foreign scientists around the globe. Now Chinese companies are adopting the same strategy by recruiting the best battery engineers from Korea.

  10. Marcel says:

    Hi Pedro,
    I was wondering if you had thoughts on this new Tesla 1 million mile battery release.

    The abstract says something about pouch cells with NMC 532 chemistry, but there was this other twitter thread saying that it would be low Cobalt and very dense volumetrically.

    I haven’t had time to read the paper, and I’m sure a lot of it would be beyond me, but if this battery is actually NMC 532, wouldn’t energy density be lower than Tesla’s current batteries, and wouldn’t costs be higher due to more Cobalt being needed?

    If it is in fact so dense -795Wh/L -, requiring low amounts of cobalt, and with super long cycle life then it is actually a battery breakthrough. Although vol. density is only 10% better than their 2170 cells.

    Tesla would have to redesign their battery packs to accommodate pouch cells. I’m guessing they would keep selling Model 3 with cylindrical cells that they currently make, and with these new cells in a different pack for customers willing to pay extra for a robotaxi battery. Maybe put the pouch cells in the S and X as an upgrade?

    Actually, I just did a skim of the paper, and it appears that this cell would work well in the Semi. Extremely long life pouch cells that can tolerate heat and full DOD cycling.

Leave a Reply

Your email address will not be published. Required fields are marked *