Simple solution for safer, cheaper and more energy-dense batteries

BYD Blade Battery prismatic cells
BYD Blade Battery prismatic cells

In the previous article we compared the energy density of many battery packs used by popular electric cars. Most batteries were either NCM 523 or NCM 622 and on average had a gravimetric energy density between 140 and 150 Wh/kg, which is disappointing considering that’s a lot less than what we get at the cell level (230-250 Wh/kg).

The poor gravimetric energy density of current EV batteries can be explained by unnecessary complexity and in this article we’ll see an easy solution to make battery packs simpler, safer, cheaper and more energy dense.

First, to give you a bit of context, we’ll see where we are now.

Currently battery packs are like matryoshka dolls, inside them we have modules and inside the modules we have the important stuff that stores energy, the battery cells. This means that the weight of the battery cells represents only part of the total weight of the batteries.

Let’s see some examples of GCTPR (gravimetric cell-to-pack ratio) to better understand how inefficient weight-wise are current battery packs.


Renault ZOE (old ZE 40 battery)

This battery weighs 305 kg, of which 185 kg (61 %) are from the cells. The rest 120 kg (39 %) of the weight is from the metal cases, cabling, BMS (Battery Management System) and TMS (Thermal Management System).

  • Battery pack weight: 305 kg
  • Gravimetric energy density at battery pack level: 145 Wh/kg
  • Battery cells weight: 185 kg (61 %)
  • Gravimetric energy density at battery cell level: 245 Wh/kg


Renault ZOE (new ZE 50 battery)

This battery weighs 326 kg, of which 206 kg (63 %) are from the cells. The rest 120 kg (37 %) of the weight is from the metal cases, cabling, BMS (Battery Management System) and TMS (Thermal Management System).

  • Battery pack weight: 326 kg
  • Gravimetric energy density at battery pack level: 168 Wh/kg
  • Battery cells weight: 206 kg (63 %)
  • Gravimetric energy density at battery cell level: 265 Wh/kg


Nissan LEAF (40 kWh battery)

This battery weighs 303 kg, of which 175 kg (58 %) are from the cells. The rest 128 kg (42 %) of the weight is from the metal cases, cabling, BMS (Battery Management System) and TMS (Thermal Management System).

  • Battery pack weight: 303 kg
  • Gravimetric energy density at battery pack level: 130 Wh/kg
  • Battery cells weight: 175 kg (58 %)
  • Gravimetric energy density at battery cell level: 224 Wh/kg


Nissan LEAF (62 kWh battery)

This battery weighs 410 kg (estimation) of which 263 kg (64 %) are from the cells. The rest 147 kg (36 %) of the weight is from the metal cases, cabling, BMS (Battery Management System) and TMS (Thermal Management System).

  • Battery pack weight: 410 kg
  • Gravimetric energy density at battery pack level: 151 Wh/kg
  • Battery cells weight: 263 kg (64 %)
  • Gravimetric energy density at battery cell level: 224 Wh/kg


BMW i3 (94 Ah battery)

This battery weighs 256 kg, of which 193 kg (75 %) are from the cells. The rest 63 kg (25 %) of the weight is from the metal cases, cabling, BMS (Battery Management System) and TMS (Thermal Management System).

  • Battery pack weight: 256 kg
  • Gravimetric energy density at battery pack level: 132 Wh/kg
  • Battery cells weight: 175 kg (75 %)
  • Gravimetric energy density at battery cell level: 175 Wh/kg


BMW i3 (120 Ah battery)

This battery weighs 278 kg, of which 215 kg (77 %) are from the cells. The rest 63 kg (23 %) of the weight is from the metal cases, cabling, BMS (Battery Management System) and TMS (Thermal Management System).

  • Battery pack weight: 278 kg
  • Gravimetric energy density at battery pack level: 152 Wh/kg
  • Battery cells weight: 215 kg (77 %)
  • Gravimetric energy density at battery cell level: 207 Wh/kg


As I mentioned many times before, the BMW i3’s battery is by far my favorite EV battery.

Here’s why:

  • Made with extremely durable prismatic NCM Samsung SDI cells (94 and 120 Ah versions, the 60 Ah isn’t that great).
  • The disposition of the cells is what it should always be, which is side by side, for better heat dissipation. Having cells on top of each other – like the Nissan LEAF battery pack has – is not a good idea, the cells on top will get hotter and degrade faster.
  • Good active TMS (Thermal Management System) with liquid cooling for keeping the battery at recommended temperature is very important.
  • KISS (Keep It Simple, Stupid) principle, since it only has 96 cells all connected in series, none in parallel.


BMW i3 battery interior


The battery pack of the BMW i3 has the highest GCTPR (gravimetric cell-to-pack ratio) of mainstream EV batteries due to its simplicity. Few big prismatic cells, all connected in series require less cabling and cases for modules.

A GCTPR of 77 % is very good for a conventional battery pack, however it could be even better with the CTP (cell-to-pack) technology.

With the CTP technology instead of having battery cells inside modules, then modules inside battery packs, we remove modules altogether. We end up with long prismatic battery cells connected in series that are put in an array and then inserted into a battery pack, making it as simple as it can be.

Various Chinese battery cell makers such as BYD, CATL and SVOLT already have their own versions of CTP battery packs.



BYD Blade battery with CTP technology


The simplicity of the BYD Blade Battery is visible in the image above. Imagine how simple it is to assemble or replace cells in this battery pack. BYD says that this battery has at least 100 cells (all connected in series).

Moreover, with CTP technology battery packs made with cobalt-free LFP/LFMP cells achieve energy density levels of around 140-160 Wh/kg, which are equivalent to what we currently get with EV batteries made with more expensive and less safe NCM 523 and NCM 622 cells.

CTP is just another technological breakthrough that helps the comeback of LFP battery cells to electric cars. I think that now there’s no doubt that LFP/LFMP battery cells will play an important role in the massification of electric cars.


BYD Blade Battery highlights:

  • VCTPR (volumetric cell-to-pack ratio): 62,4 %
  • GCTPR (gravimetric cell-to-pack ratio): 84,5 %

This means that in the BYD Blade Battery, battery cells represent 62,4 % of the volume and 84,5 % of the weight. Mainstream battery packs made with modules have on average a VCTPR of 40 % and a GCTPR of 60 %.


BYD unveils the volumetric and gravimetric cell-to-pack ratio of the new battery packs


Nonetheless, the BYD Blade Battery isn’t only about increasing the energy density of battery packs. Regarding safety this battery is hard to beat. Not only the LFP/LFMP chemistries are extremely safe on their own, the long rectangular shape of the cells provides a large cooling area and reduces the ability to generate heat during a short-circuit.


Nail penetration test by BYD of different battery cells


Furthermore, what’s really interesting is that we won’t have to wait years to see this technology being implemented. The upcoming electric car BYD Han EV arrives this June and will come equipped with a BYD Blade Battery.


The Han EV, BYD’s flagship sedan model slated for launch this June, will come equipped with the Blade Battery. The new model will lead the brand’s Dynasty Family, boasting a cruising range of 605 kilometers and an acceleration of 0 to 100km/h in just 3.9 seconds.


The BYD Han EV is a really interesting electric car, the 605 km range in NEDC should translate in around 450 km (280 miles) in the more realistic WLTP test cycle.


BYD Han EV with CTP battery technology


While BYD’s electric cars aren’t very popular outside its domestic market China yet, BYD’s electric buses are already extremely popular worldwide and these electric vehicles will probably get CTP batteries down the road.


BYD’s energy density goals: 140-160 Wh/kg for cobalt-free LFP/LFMP chemistries



CATL CTP technology


While BYD is more focused in cobalt-free LFP/LFMP batteries, CATL is working in two fronts and wants to apply the CTP technology not only to cobalt-free LFP/LFMP batteries, but also to more energy-dense NCM batteries.


CATL is already producing CTP battery packs for the Chinese carmaker BAIC.


BAIC EU5 EV with CTP battery technology from CATL


CATL’s energy density goals: 145-160 Wh/kg for cobalt-free LFP/LFMP chemistries and 200 Wh/kg for NCM chemistries



SVOLT CTP technology


SVOLT is focused to apply CTP technology to more energy dense NCMA batteries. Unfortunately there isn’t much information available about it.


SVOLT’s energy density goals: above 200 Wh/kg for NCMA chemistries


Summing up.

It’s just a matter of time before CTP becomes the mainstream technology to build simpler, safer, cheaper and more energy-dense battery packs. Moreover, we don’t have to wait years to get cobalt-free battery packs with decent energy density that are extremely safe and cheap. BYD Blade Battery is really impressive. Warren Buffett has reasons to be really happy about his bet on BYD years ago.

Furthermore, the kWh cost of cobalt-free LFP/LFMP batteries is around 20 % cheaper than high nickel-content batteries like NCM 811. Nonetheless, even without LFP/LFMP battery cells Volkswagen already has the kWh cost below 100 euros, which proves that automakers could make electric cars with decent range affordable right now if they were interested in selling them.


Battery costs roadmap by Volkswagen


Anyway, I’m really interested to know more about the SVOLT and CATL’s implementations of the CTP technology. For the moment we have more details about BYD’s own version of CTP.

Moreover, I also wonder how long will Korean battery cell makers take to realize the importance of cobalt-free batteries and start producing them. Right now Chinese companies like BYD and CATL are the uncontested experts in these chemistries. However, it would be great to see LG Chem and Samsung SDI working to improve the LFMP chemistry.

Finally, I’m very optimistic and expect that in the near future (one or two years) most electric cars will be available with LFMP (optimized for cost) and NCMA (optimized for range) battery packs made with simple CTP technology.



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.

32 Responses

  1. carlos says:

    Elon Musk agrees with you and has stated that Tesla will move forward and do away with the modules in a pack. Great article. Between economy of scale, emissions regulations and cheaper kWh, EV prices will drop significantly over the next years. I just hope the Osborne effect doesn’t kick in within the EV market.

    • Pedro Lima says:

      Thanks Carlos. In the beginning Tesla had no choice but to make a very complex battery pack to protect thousands of tiny cylindrical NCA battery cells, they successfully made a very secure pack with very delicate cells. I’m definitely curious about Tesla’s future battery packs with other chemistries and cell forms.

  2. Earl Colby Pottinger says:

    Thank you for this and the previous article, good job.

  3. Maximilian Holland says:

    Thanks for another great article Pedro. Would be good also to discuss the potential recharging time/ C-rate advantages of LFMP, which – after cost improvements – is the next area for progress towards mass EV adoption.
    Once we get 10-80% charging in a reliable 15-20 minutes (and later towards 10-15 minutes) all the doubts around “range anxiety” (which is mostly a function of the perceived difficulties/delays of recharging) will dissipate.
    LFMP should be able to outperform NCx in c-rates / cycle life.

    • Pedro Lima says:

      Thanks Max. Indeed, the higher power density is definitely an advantage for LFP/LFMP chemistries. Moreover, the extended cycle life makes V2G reasonable to implement.

  4. Pajda says:

    Hi Pedro,

    Thanks for this article. As always I am very sceptical about announcing new technologies particularly in Batteries.Now some remarks:
    1/ CTP: as very long prismatic cells. I think that we can found some studies which says that this particular dimensions are not optimal for cells with known manufacturing technologies. The idea is great but I’m afraid that there will be already full market of this if it were that simple.
    2/ LFMP: I am extreme sceptical about LFP chemistry at all. Again there will be already full market of LFP but it does not. My opinion is that LFP is pushed for “political” reasons that commercial. China invests a lot of resources into the LFP technology and so they don’t want to lose those. China even “banned” the production of other technologies (NMC, NCA) in recent history. So they makes from LFP something like “Panda bear” symbol. Hi-tech China electric vehicle manufacturers already switches to NMC. The main problem with LFP is that is not as cheap to produce. The cost of raw materials was never the main issue of overall product cost. They are cheap only due to the made in PRC. Another benefit of Safety is not significantly important for onroad electric vehicles or majority of other apps. Both NMC/NCA are still much safer than ICE. And of course the whole world is overfloded with LCO chemistry in consumer electronics. LFP safety is practicaly usable only in few specific apps like submarines or ships.

    So my conclusion is that LFP will “work” only in China and LFMP improvement does not significatly helps. But we will see. 🙂

    • Pedro Lima says:

      Hi Pajda, thanks for sharing your always interesting inputs.

      The biggest problem with long prismatic or pouch cells was that they couldn’t fit in compromise platforms designed primarily for ICE cars. Imagine putting them in a MQB platform, but in a skateboard platform like MEB you can start being creative…

      GM/LG Chem Ultimum NCMA battery cells are also very long.

      Most reports say that LFP is at least 20 % cheaper than NCM, even the recent internal report from Samsung SDI says 21 % cheaper than NCM 811. Did you read it? What do you think?

      In practice the main difference between LFMP (3,75 V) and LFP (3,2 V) is higher nominal voltage.

      By the way, the Chinese government tried to stop NCM chemistries to protect its industry back then because Chinese battery cell makers didn’t have the technology yet, once they acquired the technology and were ready to produce NCM batteries the Chinese government was all onboard with it. It even demanded that passenger electric cars required more energy-dense chemistries like NCM to get the subsidies, LFP was then relegated to electric buses…

      The likely LFP/LFMP comeback will only be possible because Chinese battery cell makers managed to increase the energy density.

      Best regards.

      • Pajda says:

        Pedro, thanks for reply. Byd Blade design (and others “very long” format no matter if pouch, prismatic or cylindrical) by my opinion must solve fundamental problem with current collectors, respectively current path. They design the output terminals on one side (which makes perfect sense) but this leads to the uneven current density thru the cell. It is the same as you make for example 10p pack from 18650. You will like to have also output wires(terminals) on one side and so you will have very high current density in nickel strip (with constant size) near wires(terminals) and low density at the end of the pack. But the cells itselves will be fine in this configuration. Maybe they already solved this (and other) problems but I do not trust to this design until it will be seen in mass application, or I was able to get sample for doing laboratory tests.

        2/ to the price of LFP/LFMP. If Samsung says that LFP is ca 20% cheaper why they do not manufacture this? The same Panasonic, Tesla, LG Chem, SK Innovation and almost all big players outside China? The second major question is if it is 20% per cell enough to compensate the need of use higher amount of cells, wiring, more containers, higher shipping and handling costs and other costs due to the lower energy density even in applications of stationary ESS which does not need the high energy density?

        The last but for me the most important is what should I say to the “poor DIYer”, when he looks at the popular NKON battery shop in EU??? You can find:

        LFP JGNE 1800mAh cell in 18650 with 5.75Wh with the price of 238 Eur/kWh per 600pcs without VAT
        NMC K-Tech 2500mAh cell in 18650 with 9.25W with the price of 133 Eur/kWh per 600pcs without VAT

        even “quality brand” like Samsung ICR18650-26J 2600mAh costs ca 186 Eur/kWh!!!

        • Pedro Lima says:

          As far as I understand from the poor automatic translation from Korean to English of Samsung SDI’s internal report is that LFP isn’t considered a “threat”, because NCM 811 still has the energy density advantage and Samsung SDI will adopt this chemistry next year.

          I think that the higher cost of LFP cells in cylindrical forms is due to low volume production. Historically cylindrical cells were always expected to have the best energy density – to use in laptops -, therefore the production of NCA, LCO and NCM has always been higher. However, if you go to AliExpress you’ll see that in prismatic and pouch forms, LFP has the price advantage over NCM.

          Ultimately you’re right, if the LFMP chemistry is the near future of cobalt-free batteries as I think they are, Korean and Japanese cell makers will eventually get on board. I think that LFMP and NCMA cathodes will coexist for a while.

  5. Cildar says:

    Your premise that heat rises in a battery is not necessarily true. Heat only rises in a fluid, within a solid heat doesn’t care about gravity and conducts in all directions equally.

    If the battery is air-cooled it will have a flow of air around the cells to dissipate the heat and offset any rising heat.
    If it is water cooled then there shouldn’t be any stagnant fluid (gas or liquid) around the cells to convect upwards.

    • Pedro Lima says:

      Hi Cildar.

      Remember the Renault Fluence ZE? It had the exact same battery cells as the Nissan LEAF but had an entirely vertical battery pack. It was very common that the capacity degradation of the Fluence ZE was so bad that Renault had to replace the battery multiple times under warranty for each customer. Renault did bought back some cars from customers and resold them without battery leasing so it didn’t have to keep replacing the battery for the new owners. Battery leasing for the Fluence ZE was a terrible deal for Renault, unlike with the ZOE that rarely needs a battery replacement.

      • Cildar says:

        Ok. Are you asserting that the Fluence battery’s problems were entirely because of being verticle?

        All I’m saying is that battery modules are solid, and heat dissipates in every direction through a solid. Heat only convects in a fluid.

        • Pedro Lima says:

          No, not at all, the first generation AESC LMO battery cells were terrible. But the degradation on the Renault Fluence ZE was even worse than on the Nissan LEAF that got the same cells. The Nissan LEAF battery pack has some cells side by side (like it should) and others on top of each other.

  6. Inst says:

    I looked it up, the BYD Han apparently has a pack density of 140 Wh/kg, which is worse than the other stuff you’ve listed. I’m not even sure if BYD is trying LiFeMnPO4, or if they’re just going with LiFePO4 for enhanced stability.

    For subsidy levels, BYD is only getting about 90% of the full subsidy; it looks very much like a hail mary attempt to get as much subsidy as they can for their LiFePO4 chemistry.

    It’s very disappointing, the Model 3’s battery pack is 167 Wh/kg. At 160 Wh/kg, the Han would have been more competitive, but it looks like the blade battery design (abuse the LiFePO4 cells more since they can take it) isn’t fully up to snuff.

    • Inst says:

      Roughly, BYD’s cell energy density is about 165 Wh/kg. At 84.5%, there’s little BYD can do about further increasing mass efficiency; if they were somehow to get it up to 94.5% (a reduction of 2/3rds of the remaining module weight), it’d only be an increase to 157 Wh/kg. The expansion potential for LiFePO4 is pretty much finished at this point.

      • Inst says:

        BYD might be able to launch its newer chemistry at about 180 wh/kg, but that only pushes it up to 153 wh/kg on the pack level.

        BYD’s problem is basically one of scale; Tesla is far larger as a company (unless they finally trip and fall), CATL has 4 times BYD’s market share, while BYD can only content itself with being larger than Guoxuan, which has 190 wh/kg LiFePO4s.

        Pushing LiFePO4 is simply going to be a matter of cost; it’s roughly 23% cheaper than conventional NMC / NCA.

        With the CTP / Blade Battery stuff, both CATL and BYD are trying to further reduce battery pack costs, but BYD seems outmoded. They’re working in all sorts of directions, but while CATL is claiming to have increased its latest NCMs past 300 wh/kg, BYD doesn’t seem to have anything equivalent.

        And in a high-tech industry (read: capital intensive) like batteries, BYD just doesn’t have the size or scale to compete anymore. Poor BYD.

  7. Leo B says:

    Nice article!
    There is a picture of the CATL-BAIC CTP-pack that seems to suggest that CATL is not using very long blade-like cells:

    Another aspect of the BYD battery is they changed the form factor to reduce height. The usual prismatic cells are rather high, but the blade cells rival cilindrical cells so a more straightforward skateboard platform becomes possible, like Tesla.

    • Pedro Lima says:

      Thanks Leo.

      SVOLT’s CTP cells are also not as long as BYD’s. I think that BYD’s version of CTP is a more simple approach, which I like.

  8. Marcel says:

    It seems like there is still a lot of room for the industrial learning curve to improve batteries and battery costs, much like what is happening with solar PV and wind power. Which is a great thing, as it seems like we are already very close to batteries that are good enough to undercut ICEVs in many market segments. As those markets take off, prices will fall further, and the BEV rout of ICEVs will really get going.

    • sola says:

      Yep, it seems there is still a lot of room for improvement.

      I hope battery and car mfgs start scaling up massively in coming years.

      As a city cyclist I cann attest that the sooner we get rid of ICEVs the better. All of the dirty, polluting, poisoning diesels will be good riddance.

  9. NK2020 says:

    Great article. Wonder what is the BMS architecture for these CTP packs? If the architecture is centralized BMS, there will be long sense cables and a decentralized architecture is difficult without any modules!

    • Pedro Lima says:

      Yes, it does make sense to have a centralized BMS in this case where modules don’t exist.

      • NK2020 says:

        Yes, but the cell voltage sense cables will be very long and result in EMC &reliability challenges, not to mention being inefficient in terms of pack volume and weight. BMS architecture is vital and yet, a big unknown with these CTP packs.

  10. Marcos Henz says:

    This blog is talking about what really matters!
    Great post. I am also very optimistic about upcoming battery techonologies.
    Cheers from Brazil

  11. elmo says:

    Pedro, I was wondering about the earth supplies for the cobalt free batteries do we have enough to create an all green powernet with V2G and local storage of energy produced with green methods like wind and solar.

    Or do we need to have alternatives to like temporarily storage of H in times of overproduction of solar and wind.
    Just curious because a lot of anti EV people shout that we can’t even build al these batteries even if we want to.

    • Pedro Lima says:

      That’s a good question, but I don’t have a definite answer.

      My guess is yes and no. Yes, we could already produce enough ESS (Energy Storage Systems) and electric cars with cobalt-free LFMP batteries equipped with V2G (Vehicle-to-Grid) technology to store a lot of the green electricity produced. But no, we can’t and shouldn’t put all eggs in one basket.

      Pumped-storage hydroelectricity is a simple and cheap method to store green energy and countries like Switzerland, Austria and Portugal use it frequently.

      “Taking into account evaporation losses from the exposed water surface and conversion losses, energy recovery of 70–80% or more can be achieved. This technique is currently the most cost-effective means of storing large amounts of electrical energy, but capital costs and the presence of appropriate geography are critical decision factors in selecting pumped-storage plant sites.”

      When Pumped-storage hydroelectricity isn’t viable due to geography, then hydrogen could play an important role.

  12. Jorge Ferreiro says:

    Interesting information on how Chinese government subisdies are tied to meeting targets on GCTPR and vehicle price, thus shaping the battery industry

    Battle of the batteries – Cost versus Performance
    10 Jun 2020

Leave a Reply

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