Comment by dmitrygr | Hacker News Reader

Comment by dmitrygr | Hacker News Reader

Let us try some napkin math. How thick does a cable need to be to move that much energy in that amount of time without overheating or being too heavy for a user to carry and plug into their car. Let's say 100KWh battery, to be charged at 99% efficiency via a 99% efficient DC-DC converter in 10 minutes. That is a rate of 600 KW. Current HV chargers use 600volts, so that is 1000A current. 0000 gauge wire is rated for 302 amps, so we'll use four such wires for VCC and four for GND, thus 8 total. That is eight wires, 0.46 inches diameter each. Must be heavy and unbendable, but let's persevere. Copper's density is 8.96 g/cm^3. This cable assembly will thus weigh 11.7 kilograms (and that is without insulation) Given copper's resistance (0.046 ohms per 1000 feel of length) at 0000 gauge, our cable assembly will be shedding ~24W of heat simply conducting this power. Toasty for user hands... I cannot even fathom the kind of connector required to allow for this current AND specced for thousands of insertion/removals. Contact areas in connectors are a constant issue, and here it will be even more so.

You'll also need to bend 8 half-inch-thick copper wires to plug/unplug this... I am not sure many people are ready for the force that will require. Also copper tends to break if you bend it a lot, so you'll need some other kind of conductor that bends, which means it will conduct less well, which means it will be thicker and heavier...

The alternative is a lighter cable with active cooling and higher losses. This has issues too (failure of cooling can snowball into a fire quite fast, coolant leaks are no fun, etc)

Now, back to our 99% efficient DC-DC converter. It will be shedding 6KW of heat, good luck cooling this, the battery charging at 99% efficiency will also be shedding 6KW of heat...same issue...

Basically, i am sure a small such battery can indeed charge in 10 minutes in the lab (~a 6C charge rate), and that is VERY cool, but moving 100KHw safely in ten minutes from anywhere to anywhere near live humans is a nontrivial engineering difficulty.

EDIT: why not higher voltage? arcing, insulation material limits, cracking, rain, costs

I've worked on very high current automotive applications in the past. While you raise some valid points, there are generally solutions to them. I've found its best to think of these things on a unit level - which is why so many battery characteristics are quoted in that way. If you can think of solutions for how to get around these issues for a single battery on a bench, there tend to be tractable ways to scale them.

Heat dissipation is one of the only things that doesn't scale linearly. 6kw sounds like a lot of power, but over 10 minutes, you are only talking about 1kwh of energy - about enough to boil 10 liters of water. Given the mass of the 100kwh battery, a large fraction of that could just be absorbed as temperature gain. If you assume the specific heat of iron (which is an underestimate) and the weight of the current 100kwh batteries, its about a 13 degree C change.

You're confusing energy and power which is leading you to the wrong conclusion. 6KW is 6KW of power. Over 10 minutes it's 1KWh of energy but typically you only care about instantaneous heat dissipation, which is power.

I think all of my units are correct. The primary point of the second paragraph was that the kind of analysis you are pointing to is inappropriate here. Thinking of this from a server heat-sinking perspective isn't helpful because the car side of the system doesn't have to operate indefinitely in steady state and is a large thermal mass.

They are borderline commercialized already (searched one up):

https://www.phoenixcontact.com/online/portal/us/?uri=pxc-oc-...

The power level is in active use on electric transit buses (using a charger that drops down onto contacts on the roof). The video here shows it:

https://new.abb.com/ev-charging/products/pantograph-down

That is a very expensive (almost 4 kilodollars: https://www.alliedelec.com/product/phoenix-contact/1101642/7...) cable (with ACTIVE COOLING) that can only do half the current. We need 2x that

It's 500 kW, you started at 600kW, not 1000 kW. They are absolutely going to have active cooling.

$4,000 is high enough I guess, but it matters more how much it costs per charge delivered (which depends on how much it is used and how long it lasts and so on).

Current Teslas can charge at up to 250kW using 480V DC at 520 amps. Efficiency is somewhere between 90-95%, so that's around 10kW of heat that the car has to dump. The car's cooling system runs through the battery pack, allowing it to pump liquid through the battery, then run it through the radiator to get rid of waste heat. IIRC, Tesla plans to increase charging rates in the future by using liquid cooled charging cables. They also have a Megacharger in the works for their semi truck.

Charging at 600kW doesn't sound easy, but it seems possible with current technology.

Such a connector can certainly be made, and might look like this:

https://globetech.jp/products/standard-high-current-connecto...

They aren't very cheap though.

This is Toyota. I doubt they'll target 100KWh batteries to start (though that's probably table stakes for when this will actually debut beyond prototypes).

If we reduce to something small but prototype ready (20KWh) and walk back their 10m to be "10m to 80% SOC" then the numbers are surely doable with our current tech.

Cable weight seems irrelevant if the user is not handling it. For instance, the cable is raised and attached to the vehicle from below while the user remains seated inside.

that is an option, but that is not how things operate now. this introduces other difficulties. We do not have space for 30 different charging stations which each has a different plug for bottom of 30 different cars. This will not scale.

Think MagSafe for cars (but physically connected), but industry-standardized.

So standardize the plug. Any electric car capable of high capacity and fast charging will have similar requirements.

If the bottom of the car is going to be armoured anyhow can't we use plates there to transfer the charge? the problem there would probably be dealing with dirt and other contaminants.

What about doubling the voltage?

Why assume this has to be a traditional cable. I'm imagining something in the middle of the car park that latches on to the bottom of the car. Maybe it could be 10 charging cables in a base that rises up which gives you the amp rating without massive copper strands.

It's easily doable if you increase the voltage. 600kw is easy if you're using 10kv, only 60 amps

then insulation cracks + rain become a problem, as does arcing

10kV is indeed a lot, but say 1kV 600amps or 1.5kV 400amps is much more doable.

Let's consider for a second though. With the size savings of HV wire you now have room for a _lot_ of nice insulation that could last a long time. Arcing would not be a problem if you don't do live disconnects/reconnects. Rain would not be a problem with extra insulation. If arcing does occur this could be detected and power interrupted for safety, and a faulty cable replaced.

High speed satefy mechanisms could mitigate this. Put a grounded pad under the car and cut power if anything flows for more than a microsecond.

I think it's doable

Let's not charge the whole battery via a single point. Real batteries tend to charge groups of cells, or individual cells, separately anyway.

Let's have 4 cables, or maybe 6 cables, charging separate sub-batteries. Since the surface to mass ratio grows as mass decreases, passive cooling becomes more viable. The mass of each cable also becomes manageable. Of course, now you need to have several connectors on the car, likely under a dedicated retractable panel.

Cars aside, I'd apprentice a phone that can charge in 10 minutes, even if it needed to be under a fan during that.

The only thing that solves is the cabling issue, which quite frankly is the easy part of charging. You can always throw a larger connector, liquid cooling, etc at the problem. You're still going to charge them all the same way. The amount of heat generated by charging the battery would remain exactly the same, which is the biggest limitation.

I suppose batteries already have cooling. During charging, it could run at top performance. Likely air cooling with ambient air would be enough.

Yeah, that's absolutely not true. Ambient air cooling for batteries is not enough when you're talking about ridiculously fast charging. Tesla's implementation has the AC running full blast (with radiator fans going full tilt) when supercharging even at the 140kW rate.

Getting the heat out of batteries is the hard part. And if you can, then keeping up with the heat generated is hard too.

Wow, really nice to see it broken down like that. What if there was some inductive solution with a giant plate (or plates) that would center itself under the car?

Yup, it's merely with assumption the electric PS network is solidly built guaranteeing availability in most places

What happens if you move to, say, 2kV? Higher differences in potential create bigger risks of arcing, and need thicker insulation. Can this be mitigated in connector design? Or is this just considered too risky to be anywhere near people?

I realize this doesn't address the heat shedding of the converter or battery.

What are the issues with moving to a higher voltage?

American Superconductor sells wire for $200 per kiloamp per metre. It would worth it getting a lighter superconducting cable given it wouldn't need cooling.

Remember superconductors are the epitome of needing cooling.

Reply on news.ycombinator.com