Some applications still require higher energy densities than what LFP can provide. Are you saying silicon anodes are more promising than solid state for those applications?

My position is as follows:

-LFP wins on price where it fits the use case, so grid storage, current-gen EVs.

-Silicon anode competes for market share with solid-state in areas where LFP can't, namely aerospace and next-gen EVs or even aviation, should they ever reach 1kWh/kg. The former has easily a decade lead considering that they're already selling packs and the latter exists largely in the lab or on paper.

I also believe even 400Wh/kg at pack level is already good enough to be a serious proposition as replacement for internal combustion in land transportation, provided pricing is adequate.

I totally agree with your position on LFP, and about 400 Wh/kg being enough to unlock some applications.

But I’m not fully convinced on silicon anodes versus solid state. Yes, energy density is very promising, but I understand capacity fade is still a huge problem due to silicon volume expansion/reduction while charging/discharging?

That's the problem they've solved here using nanowires - the cells have a 200-1200 cycle life depending on operating conditions - I suppose the range is so wide because of the temperature range - -30°C - 55°C. It's not surprising to see cells degrade fast at the hotter end of this spectrum.

For EVs they advertise cells with no less than 410Wh/L, 4C rate of discharge.

Is there still a fire risk with LFP? From my understanding, solid state batteries have a large safety advantage here because of this.

Yes - LFP still uses a flammable, liquid electrolyte - usually ethylene carbonate.

That being said LFP survives the nail penetration test without causing a fire, so it's considerably safer than the usual alternatives.

LFP cathodes and silicon or 'solid state' aka lithium metal anodes are not incompatible.

5 GWh in what time? Gigafactory 1 was at 30 GWh per year in 2019 and they planned to ramp up to 54 GWh, according to Wikipedia. So probably per year? Which would make their production capacity 570 kW. Also according to Wikipedia, a production of 35 GWh per year which is 4 MW was estimated to require 300 MW of energy input.

What is the point of comparing 300 MW energy use for production with 35 GWh/year (which is technically 4MW indeed) new battery capacity though?

These are technically the same units, but they don't measure the same things.

I find that comparison quite interesting with respect to the energy intensity of producing lithium ion batteries. A factory that can produce two oil barrels per hour or seven 60 liter gasoline tanks per hour or will also produce 35 GWh storage capacity per year and would probably not consume anywhere close to 300 MW in the process.

They measure comparable things. The ratio is how many times you would have to discharge the battery to output enough energy to manufacture the battery. It's the capacity independent measure of embodied energy. An important metric.

> Which would make their production capacity 570 kW.

I understand where you got this number, but I would rather stick to units which are easy to interpret.

Yes, 5GWh per year or 570kW - in a sense.

Energy input can be green.

If not now, at least doable.

You are right. The company you mention is focusing on the aviation industry. Drones and light airplanes. Weight matters a lot there. To the point where saving weight is worth a lot to manufacturers.

Basically every kilo you save is more useful load and range. It's also a very conservative market. Most of the planes in the process of being certified, or already certified (like the Pipistrel trainer) are using batteries that are quite unimpressive in terms of wh/kg. That's because the technology is typically at least half a decade old by the time a plane gets through the certification process.

Manufacturers can't just switch battery supplier without triggering a lot of re-certification activity and most of them would have locked in their supplier many years ago; long before even starting the certification process.. It's their next models that are being designed now that would get something that is state of the art now. Like this battery. We won't see those in the market until 6-7 years from now.

The exception to this is probably going to be experimental aircraft. I think that should become a growth market pretty soon. A 500wh/kg battery of say 100kwh would weigh about 200 kilos, give or take. That should get you some usable range and be light enough to put in a small plane. And it would put a nice little dent in the cost of what used to be a 100$ hamburger flight. Fuel cost is high and planes use a lot of it. 5-10 gallons per hour typically. And combustion engine maintenance on planes is very expensive. All that goes away with battery electric. Servicing becomes a lot simpler, less moving parts that can break and the few remaining ones last a lot longer.

Irresistible for a lot of private pilots, I would imagine. This is a market that doesn't currently exist but I don't think it should take long for people to start experimenting with battery electric.

Who’s the current market leader on EV private pilot planes?

Cessna I guess.They bought Pipistrel recently. There are a few companies getting closer to production with various planes and drones. But realistically, I think volume sales is still a few years out.

Companies like Amprius scaling their high end battery production (like they announced recently) is interesting though.

Cessna didn’t buy Pipistrel, Textron did. Maybe that’s being pedantic, but Cessna has zero development going on in the electric plane space, and Pipistrel still runs almost completely independent of the other brands in Textron’s portfolio.

No idea who the technology leader might be, but I hear about Joby most frequently.

Manned airplanes are a conservative market, I would imagine that unmanned drones are not? Especially cheap consumer ones?

EVs are already very heavy, and when the model 3 switched to LFP it got even heavier. This mass has lots of subtle consequences that aren't always appreciated. Like accelerated tire wear (which Tesla owners notice after they buy) and increased particulate pollution from that, increased road wear, increased danger to pedestrians and cyclists, etc.

So, I'm less excited about LFP. There are a number of ideas in the works that might double the energy density over the current leading lithium ion chemistries and if any of them work out its going to unlock a lot a things that don't quite work yet like:

1. light weight sports car EVs (like a miata) 2. EVs that can tow things or carry 4 mountain bikes on road trips 3. Some aircraft 4. EVs that can do a track day for a reasonable amount of time 5. EVs that are lighter than ICE counterparts rather than heavier

Wh/Kg is the relevant metric to optimize though. Whether it’s LFP or solid state doesn’t matter. If that company is building higher density LFP batteries, that addresses every concern you mention.

Independent of this, this also gives more and.better options for the energy grid in general.

And economy of scale: if we can produce it, the chance that we can scale it might only be a question of few years.

I’ll know we’ve made it when you can buy a 1000kg EV Miata.

Before my Model 3, I had a BMW M3. Curb weights were the same (2009 BMW M3 vs 2018 Model 3 LR RWD). Even comparing a 2020 Model 3 AWD vs. a 2020 BMW M3, curb weights are about the same (~4000lbs).

> increased danger to pedestrians and cyclists, etc.

Why do seemingly so people here understand clipping functions?

The lethality of a pedestrian-vehicle collision with a 4,000lb vehicle at 60MPH and a 400,000lb vehicle at 60MPH are both 1

In practical terms, the additional weight of an EV has zero effect on pedestrian collision lethality, whereas something like bumper design would.

I'm not sure why clipping functions are relevant, because the probability of death isn't 100% even for semis hitting pedestrians at highway speeds. It's actually more like 80-90%. There's still quite a bit of long tail for higher speeds beyond that as well. SUVs and other "heavy" consumer vehicles are somewhere in the 75% range at comparable speeds, while sedans are down around 65% mortality. Feel free to find your own numbers, but I'm using [1]. Vehicle weight and size clearly have a significant effect on mortality even at high speeds.

However, not all collisions are at highway speed. Most collisions are at relatively low speed, where weight is an even more significant factor in the energy involved. There are tradeoffs you can make between weight contributed by safety equipment (e.g. bumpers), but to a large extent those tradeoffs are mandated by regulations rather than consumers or OEM engineers.

[1] https://www.moneygeek.com/insurance/auto/analysis/pedestrian... [2] https://www.nber.org/digest/nov11/vehicle-weight-and-automot...

Vehicle mass matters a lot in vehicle-to-vehicle collisions, where the energy and momentum of each vehicle is going to change a lot, but when a car collides with a person, the car's momentum is barely affected, because the car is already significantly more massive than the person. A 20x mass ratio is already practically infinite.

In a direct collision between a moving 1000kg car going 10 m/s and a stationary 50kg person, the person will end up moving somewhere between 9.5 m/s (perfectly inelastic collision) and 19.05 m/s (perfectly elastic). The car will end up moving 9.5 m/s (perfectly inelastic) and 9.05 m/s (perfectly elastic).

In a direct collision between a moving 10000kg car going 10 m/s and a stationary 50kg person, the person will end up moving 9.95 m/s (perfectly inelastic) and 19.90 m/s (perfectly elastic). The car will end up moving somewhere between 9.95 m/s (perfectly inelastic) and 9.90 m/s (perfectly elastic).

Even though we 10x'd the weight of the car, we only increased the velocity of the person after impact by about 5%, and energy imparted on the person by about 10%.

I suspect vehicle size (or rather, the size/shape of the front of the vehicle) has way more effect on pedestrian survivability than the weight of the vehicle. SUVs and trucks will impart the force of impact directly to your torso and head, and then subsequently run you over. Sedans hit you in the legs and then roll you over the top of the vehicle.

Higher curb weight correlates with longer breaking distances