3 ms·

Hi, engineer from an electric vehicle company here. This is not an issue because, even if a small wire is used, the efficiency loss is not all that considerabl

Hi, engineer from an electric vehicle company here. This is not an issue because, even if a small wire is used, the efficiency loss is not all that considerable. Energy is cheap, so liquid cooling the cables and dissipating i.e. 5% of the energy into the conductor is a straightforward trade-off. Compare that trivially to petrol, which costs petrol (diesel) to move to petrol stations, and it's actually not that surprising. Additionally, there has recently been a lot of investment into EV chargers with built-in batteries that charge at slower rates when a vehicle is not present, and operate independently from the grid once a vehicle is connected. This answers the obvious question about infrastructure to deliver the power required.

clouddrover 6y ago
> Additionally, there has recently been a lot of investment into EV chargers with built-in batteries Two such chargers: https://freewiretech.com/products/dc-boost-charger/ https://freewiretech.com/products/dc-boost-charger/ https://insideevs.com/news/398348/eon-volkswagen-fast-charging-everywhere/ https://insideevs.com/news/398348/eon-volkswagen-fast-chargi...
AnthonyMouse 6y ago
> Additionally, there has recently been a lot of investment into EV chargers with built-in batteries that charge at slower rates when a vehicle is not present, and operate independently from the grid once a vehicle is connected. Is this even going to be required in most cases? Charging at that speed is only really useful for road trips where you're sitting there waiting for charging to complete so you can go. If you're charging while the car is parked at home for the night, what does it matter if it's ten minutes or ten hours? So then you only really need the very fast chargers at truck stops on the highway and the like, and then couldn't those chargers just be given a hefty grid connection? The high voltage lines already run parallel to major highways.
octoberfranklin 6y ago
You basically dodged his question there. Increasing I^2R heat loss by 5% doesn't let you shrink the cable much. You go from "humongously phat cable" to "really phat cable". The problem remains. Also remember that the cost of the petroleum distribution infrastructure is shared between in-town drivers and road-trip drivers. If these "fast chargers" are only supposed to be used by people who need more than one batteryful of driving in a given day (i.e. people on a road trip), the budget available to pay for the refueling infrastructure is going to be tiny compared to the petroleum infrastructure. EVs will simply never "catch on" for non-commuter use if we go down that road.
- irjustin 6y ago
  Let's just go superconducting cable and we're good to go!
- sephamorr 6y ago
  I think you're overestimating the difficulty of moving 500A. As GP said, liquid-cooled cables are the solution and come in reasonably-sized cable thicknesses (CCS form factor). This is ITT Cannon's product as an example: https://ittcannon.com/core/medialibrary/ittcannon/website/literature/catalogs-brochures/itt-cannon-evc-dc-liquid-cooled-brochure.pdf https://ittcannon.com/core/medialibrary/ittcannon/website/li... Also note that 1000V isn't an upper bound; many vehicles under design now will be ~1000V, and I expect that to push up to at least 1500V within the next few years.
  - imtringued 6y ago
    I'm not an electrical engineer but I ran into the I^2R heat loss problem with a simple mosfet circuit. If I drive the mosfets at 48V (voltage is actually irrelevant) and 70A and the RDS of the mosfet is 0.002 Ohm then the mosfet will lose 9.8W of power to heat. For a small compact SMD mosfet that's a lot of heat and requires a heatsink. Once you scale up to 500A you are basically losing 500A x 500A x 0.002 Ohm = 250W of power just on a single mosfet. I'm not very experienced in this subject but I noticed that mosfets with lower voltage ratings tend to also have lower drain source resistance (RDS) which means doing this at 1000V can only get harder.
    - aetherspawn 6y ago
      As the parent comment said, moving high currents in cables isn’t that difficult. Mosfets are small devices, but cables have a lot of surface area and mass, so even dissipating kW into a cable is not an issue and can be solved with liquid cooling. Imagine we used 35mm2 at 0.55 Ohm/km and we have 2 poles. If the charging cable is 3m including inside the device, that’s 2x3 = 6m. That’s 6.6 mOhms total, so I2R on that at 500A is only 1.65kW, which is nothing.
conk 6y ago
5% of 450,000w is 22,500 watts. That’s also 5% of the energy not going to charge the battery further raising the power pushed over the charging cable. As for chargers with batteries, I don’t see how that’s viable. 75kWh would require each charger to have 6 powerwalls just to store 75kWh. If you wanted to dispense that amount of power in 10 minutes you would need 60 powerwalls per charger. Maybe a battery could be designed to provide higher power output, but it’s going to have the same issues that car will have when attempting to charge at a fast rate. Even with storage in the charger how quickly can it recover? Tesla superchargers run non-stop on busy travel holiday. Relying on a onsite battery would do no good on high demand days.

clouddrover 6y ago

> Additionally, there has recently been a lot of investment into EV chargers with built-in batteries Two such chargers: https://freewiretech.com/products/dc-boost-charger/ https://freewiretech.com/products/dc-boost-charger/ https://insideevs.com/news/398348/eon-volkswagen-fast-charging-everywhere/ https://insideevs.com/news/398348/eon-volkswagen-fast-chargi...

AnthonyMouse 6y ago

> Additionally, there has recently been a lot of investment into EV chargers with built-in batteries that charge at slower rates when a vehicle is not present, and operate independently from the grid once a vehicle is connected. Is this even going to be required in most cases? Charging at that speed is only really useful for road trips where you're sitting there waiting for charging to complete so you can go. If you're charging while the car is parked at home for the night, what does it matter if it's ten minutes or ten hours? So then you only really need the very fast chargers at truck stops on the highway and the like, and then couldn't those chargers just be given a hefty grid connection? The high voltage lines already run parallel to major highways.

octoberfranklin 6y ago

You basically dodged his question there. Increasing I^2R heat loss by 5% doesn't let you shrink the cable much. You go from "humongously phat cable" to "really phat cable". The problem remains. Also remember that the cost of the petroleum distribution infrastructure is shared between in-town drivers and road-trip drivers. If these "fast chargers" are only supposed to be used by people who need more than one batteryful of driving in a given day (i.e. people on a road trip), the budget available to pay for the refueling infrastructure is going to be tiny compared to the petroleum infrastructure. EVs will simply never "catch on" for non-commuter use if we go down that road.

irjustin 6y ago
Let's just go superconducting cable and we're good to go!
sephamorr 6y ago
I think you're overestimating the difficulty of moving 500A. As GP said, liquid-cooled cables are the solution and come in reasonably-sized cable thicknesses (CCS form factor). This is ITT Cannon's product as an example: https://ittcannon.com/core/medialibrary/ittcannon/website/literature/catalogs-brochures/itt-cannon-evc-dc-liquid-cooled-brochure.pdf https://ittcannon.com/core/medialibrary/ittcannon/website/li... Also note that 1000V isn't an upper bound; many vehicles under design now will be ~1000V, and I expect that to push up to at least 1500V within the next few years.
- imtringued 6y ago
  I'm not an electrical engineer but I ran into the I^2R heat loss problem with a simple mosfet circuit. If I drive the mosfets at 48V (voltage is actually irrelevant) and 70A and the RDS of the mosfet is 0.002 Ohm then the mosfet will lose 9.8W of power to heat. For a small compact SMD mosfet that's a lot of heat and requires a heatsink. Once you scale up to 500A you are basically losing 500A x 500A x 0.002 Ohm = 250W of power just on a single mosfet. I'm not very experienced in this subject but I noticed that mosfets with lower voltage ratings tend to also have lower drain source resistance (RDS) which means doing this at 1000V can only get harder.
  - aetherspawn 6y ago
    As the parent comment said, moving high currents in cables isn’t that difficult. Mosfets are small devices, but cables have a lot of surface area and mass, so even dissipating kW into a cable is not an issue and can be solved with liquid cooling. Imagine we used 35mm2 at 0.55 Ohm/km and we have 2 poles. If the charging cable is 3m including inside the device, that’s 2x3 = 6m. That’s 6.6 mOhms total, so I2R on that at 500A is only 1.65kW, which is nothing.

conk 6y ago

5% of 450,000w is 22,500 watts. That’s also 5% of the energy not going to charge the battery further raising the power pushed over the charging cable. As for chargers with batteries, I don’t see how that’s viable. 75kWh would require each charger to have 6 powerwalls just to store 75kWh. If you wanted to dispense that amount of power in 10 minutes you would need 60 powerwalls per charger. Maybe a battery could be designed to provide higher power output, but it’s going to have the same issues that car will have when attempting to charge at a fast rate. Even with storage in the charger how quickly can it recover? Tesla superchargers run non-stop on busy travel holiday. Relying on a onsite battery would do no good on high demand days.