🔍
#1 A Brief History of LFP, Patents, Licencing Costs, Pricing // and Tesla - YouTube
Channel: The Limiting Factor
[5]
Welcome back everyone! I’m Jordan
Giesige and this is The Limiting Factor.
[10]
Elon recently confirmed that North American
made, standard range plus Model 3s would be
[16]
using lithium iron phosphate, or LFP, battery
packs from China. As a result of that,
[22]
I’m kicking off another video series, this time
on LFP. I was planning on starting an LFP series
[29]
later this year, but Tesla is incorporating LFP
in their US battery packs sooner than I expected.
[36]
Today we’ll cover a brief history of LFP; the key
patents, licencing costs, when the patents expire,
[43]
and why that matters; the cost difference between
an LFP and Nickel battery pack; whether this
[49]
is bad news for Tesla’s 4680 ramp; and some
cliff’s notes on how LFP compares to Nickel.
[57]
Note that these will just be cliff’s notes,
and I’ll go much deeper and highlight specific
[62]
manufacturers in the upcoming series.
Before we begin, a special thanks to my
[67]
Patreon supporters and YouTube members.
This is the support that gives me the
[72]
freedom to avoid chasing the algorithm and
sponsors, and I hope will eventually allow
[77]
me to do this full time. As always, the
links for support are in the description.
[81]
The story of LFP is quite long and convoluted,
so I’ll be giving a high level overview from
[87]
the perspective of an investor or consumer.
It’s also worth noting that there are missing
[93]
pieces in the story and I’ve done my best to piece
together a timeline here that has a logical flow.
[99]
LFP is a cathode material that was discovered
by John Goodenough in 1996 while working at UT
[107]
Austin. Not long after, in 1999, UT
Austin was granted a patent for LFP.
[114]
However, LFP had a critical flaw, which was
that it had electronic conductivity low enough
[120]
to be considered an insulator. As a result, UT
Austin didn’t have an interest in developing LFP
[127]
and sold the patent rights to Hydro-Quebec
through an exclusive patent licence agreement.
[133]
Hydro-Quebec is a public utility in Canada with
a large R&D arm. That is, they had the know-how
[140]
and funding to further develop LFP. Hydro-Quebec
then began working on the conductivity problem
[146]
and by the 21st of September 2001, Michel
Armand filed this patent through Hydro Quebec.
[153]
Michel solved the conductivity problem with a
carbon coating, which opened the door for LFP
[159]
as a viable commercial chemistry.
Note the date, the 21st of September
[164]
2001. Patents are granted from the original
filing date and are valid for 20 years. There
[170]
appears to be some rules which allow extensions,
but we’ll come back to this later in the video.
[175]
In the same year, 2001, A123 systems was
founded. Their focus was also on developing
[183]
and commercialising LFP. In 2006, Yet-Ming
Chiang, a founder of A123, patented improved
[191]
nanostructures and dopants to further improve
the ionic and electronic conductivity of LFP.
[197]
Around the same time, 2005 to 2006, A123 quickly
moved to sell LFP cells to Black & Decker for
[206]
DeWalt power tools and subcontracted BAK Battery
in China to do the manufacturing. Unsurprisingly,
[213]
Hydro-Quebec claimed A123 infringed on their
patents and a patent battle ensued that lasted
[219]
for about 5 years, until about 2011. Hydro-Quebec
ended up winning the patent dispute and A123
[227]
ended up being brought under Hydro-Quebec’s
patents through a sublicense agreement.
[232]
During the same time period, Hydro-Quebec
developed industrial partnerships and
[237]
formed a consortium to manage the patents for
LFP. The name of the consortium was LFP plus
[243]
Carbon Licensing. For me, this is where the trail
ran cold, but it seemed like a huge part of the
[249]
story with China was still missing. What ended
up happening in China with licencing? Why has
[255]
every major Chinese battery company developed
LFP and why are LFP cells mostly sold in China?
[263]
Several weeks ago, Steve LeVine came to the
rescue with a great article that, for me, brought
[268]
new information to light. If you aren’t following
Steve on Twitter, I recommend it. In the article,
[274]
he states that the LFP + Carbon Consortium made a
decision a decade ago to allow Chinese companies
[281]
to make LFP with no licencing fee as long
they only sold the batteries domestically.
[287]
I asked Steve where he got this information
and he directed me to this article by Roskill.
[293]
According to Roskill, LFP cathode producers
are the source of the claim that Chinese
[298]
companies pay no licensing fee. Note that one
decade ago is when the patent dispute between
[304]
A123 and Hydro-Quebec was winding up and that
dispute involved a Chinese company, BAK Battery.
[312]
I couldn’t find further information on what
happened here. It’s a strange decision to forego
[317]
tens of millions of dollars in potential revenue,
but I’m willing to speculate. It appears that,
[322]
for one reason or another, UT Austin wasn’t
granted patents under Chinese jurisdiction.
[328]
Maybe UT Austin didn’t realise in the late 90s
that what they had on their hands was valuable
[334]
or didn’t realise that China would become
the world’s dominate battery player.
[339]
UT Austin’s Patents are the most critical patents
because they cover the LFP material itself,
[345]
which means that without China based patents, it
may have weakened their patent claim in China.
[351]
That is, the Consortium may have decided
against what could have been an expensive
[355]
legal battle that they had low odds of winning.
If that’s correct, the decision not to charge
[361]
a licencing fee if Chinese companies stuck
to their domestic market may have also been
[367]
a way to keep China out of the rest of the
major world’s major battery markets with LFP.
[373]
That is, giving the Chinese LFP market to
Chinese battery manufacturers may have been
[379]
a good deal for both parties. Each party
avoided expensive litigation and each party
[385]
walked away with the keys to a market. If you
know something about patent law or negotiations,
[391]
let me know your opinion in the comments below.
So when do these patents expire and when will
[396]
China be free to export LFP battery cells
and packs to the US without a licencing
[403]
fee? The information I’ve come across is
conflicting, but April 2022 looks the most
[409]
likely based on information from Roskill
and James Frith of Bloomberg. James Frith
[414]
is another good follow on twitter if you want
to keep an eye on battery industry trends.
[419]
Again, the information for the expiry
date comes from Cathode producers.
[424]
Although most of the dozens of patents governing
LFP will expire in the next couple of weeks, one
[430]
of Michel Armand’s patents for the carbon coating
received an adjustment, pushing its expiration
[436]
date to the 27th of April 2022, 7 months from now.
James also provided information on the licencing
[444]
cost. It costs millions of dollars per year for
the licence and there’s also an additional fee
[450]
that’s rumoured to be less than $1 per kilogram
of LFP produced. I did some back of the napkin
[457]
cost modelling and 80 cents per kilogram is
about a 3% cost premium at the cell level,
[463]
which is actually pretty hefty and doesn’t
include the fixed yearly cost of the license. 3%
[470]
is roughly the profit margin that Panasonic
makes off the battery cells it sells to Tesla.
[476]
With that in mind, how much will LFP cells
affect the production cost of the US made
[482]
standard range Model 3? The average price of
an LFP battery pack is typically 20% cheaper
[489]
than an average Nickel based battery pack, but
there are other costs we need to factor in here.
[495]
First, as James Frith points out, there’s
a 10% import tariff. Second, there would
[501]
be shipping costs, which as I understand it, would
be negligible adding about 1% to the price. Third,
[508]
if the patents haven’t expired, there would
be a licencing fee of about 3%, give or take.
[514]
After factoring all that in, the cost
savings for LFP batteries shipped to the US
[519]
offers about a 6% savings instead of 20%. Tesla’s
cost savings might be less because the NCA cells
[527]
that they put in their vehicles are purchased
at near the manufacturing cost from Panasonic.
[533]
Furthermore, Tesla’s been refining their
pack manufacturing for over 10 years.
[537]
In other words, I think the LFP cells Tesla is
purchasing from China might not be generating
[543]
significant savings compared to a nickel
based chemistry. Instead, the primary reason
[549]
for using LFP would be to increase production
output, allowing Tesla to increase their output
[555]
of vehicles and energy storage products.
Tesla’s been cell constrained for years,
[561]
and LFP cells from China give Tesla access
to profit they would’ve otherwise had to
[567]
leave on the table. It also broadens Tesla’s
supply chain and makes them less dependent
[572]
on Nickel to drive their growth, which will
become more important in the coming years.
[577]
With all that in mind, let’s start digging into
this tweet by Troy Teslike, which predicted Tesla
[584]
would soon switch the LFP battery on the 22nd of
August, just a few days before it was confirmed by
[591]
Elon. This tweet sparked concern for some people
that if Tesla is moving standard range vehicles to
[597]
LFP cells, it’s bad news for 4680 cell production.
With regards to the first bullet, which states
[604]
that Standard Range vehicles out of Fremont
will be using LFP from October 1st, Elon
[610]
already stated earlier this year that Tesla would
be shifting their standard range vehicles to LFP.
[616]
In other words, the plans for a switch to LFP
pre-date any indications that the ramp of the
[622]
4680 lines could be delayed. It also means
that they would have had plans for what to do
[629]
with the 2170 cells the switch to LFP would
free up and brings us to the next bullet.
[635]
The second bullet states that Model Y production
in Texas and Berlin will start with 2170s instead
[642]
of 4680s. Again, Elon already said in the Q2 2021
earnings call that they have a back-up plan with
[650]
2170s. So, even if 4680 production is delayed, it
won’t affect vehicle production in the short term.
[658]
In fact, if Fremont does switch to
LFP battery cells later this month,
[663]
that probably frees up about 6-9 months’ worth
of production ramp for the Model Ys from Austin
[669]
and Berlin. That, in turn, gives Tesla’s
engineers quite a bit more breathing room to
[675]
work the rats and mice out of the 4680 lines.
As with the Q2 quarterly earnings call,
[681]
I’m eager to hear updates from Tesla on the
4680 in the Q3 earnings call, and if there
[689]
is useful information to dissect, I’ll produce a
video giving my take from a battery perspective.
[695]
In summary, there are a large number
of patents that govern LFP technology,
[700]
but most of the critical patents appear to be
filed on or before the 21st of September 2001.
[707]
Now, 20 years later in September of 2021,
most of those patents are due to expire.
[713]
There is still one lingering patent, but that will
be expiring in about 7 months. When that patent
[719]
does expire, it will potentially lower the cost
floor for LFP batteries by around a few percent.
[726]
More importantly, just as Drew and Elon suggested
on Battery Day, it’ll mean Tesla can produce LFP
[732]
batteries in house. It also means they’ll be able
to do so without any red tape or licence fees.
[739]
Regardless, as confirmed by Elon, Tesla
is now incorporating LFP battery cells
[745]
into their worldwide vehicle production
for standard range plus vehicles. This
[751]
fresh supply of cheap battery cells will bolster
the growth of both Tesla’s energy storage product
[757]
line and the vehicle ramps in Austin and Berlin.
Of course, what we’re all waiting for is Tesla’s
[764]
4680 lines to spin up, which will allow for
unbridled growth rates that far exceed 50%
[771]
per year. However, the 4680 lines
appear to be coming along nicely,
[776]
with Galileo Russell of Hyperchange reporting
that over the course of the last year the yield
[781]
rate at Kato Road has gone from 20% to 70-80%.
70-80% is on par with Giga Nevada as indicated
[791]
by this image from @JPR007. There is a catch
here, and it’s that despite the high yield rate,
[798]
Tesla is currently unable to sustain high
production rates. If you’d like to know more,
[804]
it was covered in my Q2 earnings call video.
Some people were wondering whether Tesla is
[810]
importing the entire vehicle, or just the
pack. @ranig on twitter did an inventory
[815]
check and there are vehicles with LFP battery
packs that appear to be assembled in Fremont.
[821]
And for those concerned about LFP batteries
being inferior to Nickel batteries, as Elon
[826]
has suggested in this tweet, they’re not. Like all
battery chemistries, LFP does have eccentricities,
[834]
but on balance will perform as well in Model
3s as the current Nickel based chemistries.
[840]
In short, if you live in a hot climate, LFP is
hands down a win. If you live in a cold climate,
[846]
there is a potential for lower performance,
but there are workarounds thanks to Tesla’s
[851]
superior software and thermal management
systems. If you live in any climate,
[856]
you’ll be getting a million mile battery that’s
much more robust to regular 100% charges.
[862]
As a final note, this is just the first of
three videos on LFP. For those keeping track,
[868]
we have one more video left in the QuantumScape
series, three videos left in the Gigacasting
[873]
series, two more videos in this LFP series, and
I’ll be working through these videos randomly.
[879]
If you enjoyed this video, please consider
supporting me on Patreon with the link at the
[884]
end of the video. I am also active on Twitter.
You can find the details in the description,
[889]
and I look forward to hearing from you.
A special thanks to Sten Karlson,
[894]
Mark Russcher for your generous support of the
channel, my YouTube members, and all the other
[898]
patrons listed in the credits. I appreciate
all your support, and thanks for tuning in.
Most Recent Videos:
You can go back to the homepage right here: Homepage