800V fast charging discussions

[ajdelange]

in what I think is a first for you is your quotes got messed up.

Went back and checked and everything look OK,

A couple of follow-ups:
1. from what I see the CCS2 standard is currently limited to 500A (or less). You don't think that's the case?

That is absolutely the case and that is what I said. To be clear: it is true,

2. Yes, the chargers themselves are currently limited to less amperage higher voltages as you noted.

The chargers are power limited. That means less current at higher voltages.

3. This is where I think we may be speaking slightly differently. So I'm going to pull in a video recording of a few systems quickly here. In it are an EQS, E-tron GT (same 800v platform as the taycan), Mach e, and a model S.

For a full charge, I agree something like 1C is probably about as good as you'll get. But for the "optimal" charging range of a battery 2C isn't unheard of.

I don't think we are at odds here. I expect 1.2C or a bit above as OEMs become bolder with their batteries but that is an AVERAGE rate which means that the peak rates are higher and the minimum lower. This is taper.

Sponsored

 

[ajdelange]

RE: No. 13: I remember when vacuum cleaners were advertised as having "more amps" than competitors.

 

[Craigins]

What kills me in the end, even if we get to 350-400kw charging, as EVs become normal, the grid is not going to be able to absorb that type of random and unpredictable demand increase.

I guess that ties back into the chargers not being able to provide the power. Everyone will need onsite storage to smooth out the instant demand (with storage it could even be that the grid demand doesn't change, just redirects from the stationary battery to the vehicle).

But this requires a fundamental change in grid operations or charging infrastructure, or both, to fully support when evs become more widely adopted.

 

[SeaGeo]

Went back and checked and everything look OK,

That is absolutely the case and that is what I said. To be clear: it is true,

The chargers are power limited. That means less current at higher voltages.

I don't think we are at odds here. I expect 1.2C or a bit above as OEMs become bolder with their batteries but that is an AVERAGE rate which means that the peak rates are higher and the minimum lower. This is taper.

Just want to be careful over what range we're averaging over. If you and I are charging from 10 say 75% in ~20 minutes still an average of 2C over that time period. If you're going deeper into that pack, that definitely drops down to an average of 1.2C or less pretty easily.

also, as a side note: the max pack would be a bit less than 2C over that 10 to 60 to 80% range, even with an 800v system with current chargers given their max output is 350kw.

 

[Zoidz]

Come on now.

Not sure what "come on now" means? Are you disputing that if a manufacturer doubles the voltage while keeping the the same battery capacity, the charge time can't potentially be cut in half?

 

[ajdelange]

Not sure what "come on now" means? Are you disputing that if a manufacturer doubles the voltage while keeping the the same battery capacity, the charge time can't potentially be cut in half?

No. Not disputing that at all. What I am disputing is that it is necessary to go to a 800V architecture to double the charging rate and thus halve the time.

Consider a 100 kWh battery pack configured for 400V with an internal impedance of 0.01Ω being charged at a 400A rate. That would require 404 V out of the charger and the losses would be 1600 W out of the 161.6 kW being delivered by the charger i.e about a percent. Now suppose we want to double the charge rate (not that we'd want to do that because we don't want to charge the battery at 3.2C). We now need to push 800A requiring 408V from the charger. Voila! The power delivered is now 326 kW but, hold on, the losses are now 6.4 kW or 2% of the power delivered. So it is a little less efficient and, of course, we have to go to a cooled or huge cable from the charger. This is the approach Tesla has taken.

Now of course we can reconfigure the cells and charger into an 800V architecture. As the cells are now more in series we might assume Z = 0.02. To get 161 kW into the battery so configured will take about half the current and accordingly the losses will be less. The same is true if we double the charge rate. Having to run half the current to the car means much less robust charging cable and connector. This is the approach Lucid has taken.

HP350 chargers will deliver up to 920 and up to 500A in any combination that produces less than 350 kW and will, thus, charge either 800V or 400V architecture cars. But they can't charge 400V cars at 326 kW because they can't deliver the current.

Clear?

 

[ajdelange]

Just want to be careful over what range we're averaging over.

My observations on Tesla charging averaged over a bunch of charges over whatever ranges were used for convenience on the trips taken is that the average rate is about 1C. I have certainly seen rates over that early in charging sessions and rates lower than that later. The interesting thing is that if you plot SoC vs time the curves are pretty close to being linear meaning that it doesn't matter that much where in the SoC range you charge. Don't get me wrong. It DOES matter but just not that much assuming you are a prudent driver staying out of the top 20%.

 

[Zoidz]

No. Not disputing that at all. What I am disputing is that it is necessary to go to a 800V architecture to double the charging rate and thus halve the time.

Consider a 100 kWh battery pack configured for 400V with an internal impedance of 0.01Ω being charged at a 400A rate. That would require 404 V out of the charger and the losses would be 1600 W out of the 161.6 kW being delivered by the charger i.e about a percent. Now suppose we want to double the charge rate (not that we'd want to do that because we don't want to charge the battery at 3.2C). We now need to push 800A requiring 408V from the charger. Voila! The power delivered is now 326 kW but, hold on, the losses are now 6.4 kW or 2% of the power delivered. So it is a little less efficient and, of course, we have to go to a cooled or huge cable from the charger. This is the approach Tesla has taken.

Now of course we can reconfigure the cells and charger into an 800V architecture. As the cells are now more in series we might assume Z = 0.02. To get 161 kW into the battery so configured will take about half the current and accordingly the losses will be less. The same is true if we double the charge rate. Having to run half the current to the car means much less robust charging cable and connector. This is the approach Lucid has taken.

HP350 chargers will deliver up to 920 and up to 500A in any combination that produces less than 350 kW and will, thus, charge either 800V or 400V architecture cars. But they can't charge 400V cars at 326 kW because they can't deliver the current.

Clear?

Thanks for detailing, agreed. My comment about 400v vs 800v rates was intentionally not considering limitations of any particular DCFC hardware. As 800v architecture evolves and DCFC evolves, higher capacity chargers are inevitable. ChargePoint and others are introducing 400kw DCFC. If/when Solid State batteries come to market, it's conceivable we could see pack voltages increase to 1000v - 1200v to support even higher charge rates that solid state batteries reportedly will support.

 

[camaroz1985]

What kills me in the end, even if we get to 350-400kw charging, as EVs become normal, the grid is not going to be able to absorb that type of random and unpredictable demand increase.

I guess that ties back into the chargers not being able to provide the power. Everyone will need onsite storage to smooth out the instant demand (with storage it could even be that the grid demand doesn't change, just redirects from the stationary battery to the vehicle).

But this requires a fundamental change in grid operations or charging infrastructure, or both, to fully support when evs become more widely adopted.

It's not going to stop there. Batteries are getting bigger and automakers are targeting 600 kw charging...

 

[ads75]

What kills me in the end, even if we get to 350-400kw charging, as EVs become normal, the grid is not going to be able to absorb that type of random and unpredictable demand increase.

I guess that ties back into the chargers not being able to provide the power. Everyone will need onsite storage to smooth out the instant demand (with storage it could even be that the grid demand doesn't change, just redirects from the stationary battery to the vehicle).

But this requires a fundamental change in grid operations or charging infrastructure, or both, to fully support when evs become more widely adopted.

The electric grid already handles swinging electric loads. Demand constantly changes throughout the day, and days have different load patterns (weekends/week days/holidays), and season change that also. An arc furnace at a steel plant is considerably more load, and varies considerably more, than a plug in truck. As does a wind tunnel, or any other countless number of industrial operations. At the residential level, everyone runs their washing machine or dishwasher at different times.

Now if you are discussing charging at home, that can be a challenge to your distribution feeder in certain situations. If you lose power, and the rest of your neighborhood for a decent amount of time, not just a trip and reclose of the circuit, but a real fault that takes time to repair, when the power company tries to restore your circuit, it can immediately trip back open because everything in the neighborhood is trying to come back online immediately (everyones AC, or heat, lights, TVs, computers, refrigerators, cars that are charging). Then the utility may need to send someone to the substation to "help" the circuit reclose, since the circuit normally can't handle everything coming on immediately. Those loads normally are more diversified in their cycling time, providing a lower, smoother demand, not a spike when power is immediately available. But those problems already exist today.

As for when EVs are more widely adopted, yes, electric utilities may need to eventually increase circuit capability for distribution (neighborhood level) circuits. But if they charge less to to charge at night, that can be an incentive to charge at night (after say 10pm, or midnight) , when demand is typically lower than the day. Charging stations (like EA) should be being built at places that can already handle the demand. It is possible if you see a new charging station being built, it may need the local utility to upgrade something before it comes online, but I can't imagine there would be a significant delay between the station being built and the utility needing to do any upgrades.

 

[ajdelange]

As 800v architecture evolves and DCFC evolves, higher capacity chargers are inevitable. ChargePoint and others are introducing 400kw DCFC. If/when Solid State batteries come to market, it's conceivable we could see pack voltages increase to 1000v - 1200v to support even higher charge rates that solid state batteries reportedly will support.

I believe it is the combination of larger capacity (135 - 200 kWh) batteries and the desire to charge them at higher and higher C values that together will drive the trend towards higher voltage systems. Both those things mean higher power wanted from the charger and that pushes the designer towards higher voltage.

 

[ajdelange]

Now if you are discussing charging at home, that can be a challenge to your distribution feeder in certain situations.

I can't say that this would never happen but I suppose it could. The utility can always hang bigger (or more) transformers and string larger conductors from the substationif they needed to but I don't think that's very likely to be required. When the utility told me I needed 600 A service for my 150 A load (don't ask) they just popped in a bigger transformer (and billed me for that I'm sure).

If you lose power, and the rest of your neighborhood for a decent amount of time, not just a trip and reclose of the circuit, but a real fault that takes time to repair, when the power company tries to restore your circuit, it can immediately trip back open because everything in the neighborhood is trying to come back online immediately (everyones AC, or heat, lights, TVs, computers, refrigerators, cars that are charging).

The main load in the majority of houses will be heat pumps, in particular ones running in A/C mode in the summer but can be in winter too for those that heat with them. These do not all come on at once at power restoration after an outage. For starters your thermostat will have a timeout and in addition to that the compressor control in the heat pump adds an additional timeout. In many cases the latter is not a fixed timeout but a random one. This is done mainly in order to protect compressors from starting up against full head but of course it also prevents the problem you are describing. Reclosers don't trip for that second and third (final) time because of startup load. They do so because the fault hasn't been cleared.

This system has worked well for years but now there is a new potential twist in BEV charging. I can only speak to Tesla but would be very, very surprised if all the manufacturers don't do the same. When the power is applied (contactor in the EVSE closes) the vehicle gradually ramps up the power it takes from the mains. Again this is done more for the protection of the car than for elimination of startup transients upon power restoration but clearly it has that effect and thus removes EVSE from consideration in worrying about sudden startup grid load.

 

[Zoidz]

As for when EVs are more widely adopted, yes, electric utilities may need to eventually increase circuit capability for distribution (neighborhood level) circuits. But if they charge less to to charge at night, that can be an incentive to charge at night (after say 10pm, or midnight) , when demand is typically lower than the day.

California is facing the opposite problem. Too much power during the day, due to excess solar capcity, and not enough at night. So from a supply/demand issue, they should incent people to charge during the day - they are currently paying homeowners (in the form of an "excess power" credit) to not generate solar power.

 

[ajdelange]

Thus we see the main problem facing us is storage. Hydrogen in inefficient and batteries are still wicked expensive. Here in VA we are fortunate to have the particular geography to allow for a huge pumped storage facility (pump water from a lower reservoir to a higher one when there is excess and let it flow back when there is demand). It is one of few pumped storage facilities in the world because the geography doesn't exist in many places.

Some researchers in the Netherlands have what I think is a devilish clever new take on pumped storage. There is a rigid tank buried in the sea bed but open to surface pressure via a pipe. Above it on the sea floor is a flexible bladder. When there is excess generation (they are studying this for offshore wind) water is pumped from the tank to the bladder. When there is demand water from the bladder flows back to the tank through a turbine. Don't know whether this will pan out but I guess that's what the studies are to determine.

 

[Zoidz]

Thus we see the main problem facing us is storage. Hydrogen in inefficient and batteries are still wicked expensive. Here in VA we are fortunate to have the particular geography to allow for a huge pumped storage facility (pump water from a lower reservoir to a higher one when there is excess and let it flow back when there is demand). It is one of few pumped storage facilities in the world because the geography doesn't exist in many places.

Didn't realize there was one in Virginia, where's it located?

Our engineering company did an interesting one-off project for someone who is truly independently wealthy. He has a remote camp in the upper midwest with a large manmade fishing lake. We installed an off the grid solar/hydro system that pumps water into the lake during the day, and generates power via hydro at night.

Sponsored