I have been mulling over the inefficiency of taking PV DC power, converting it with 20% losses to AC for the electric company, and then converting it back to DC with another 20 percent losses to charge an EV. If that was not painful enough, I will have ample enough time waiting for the EV to charge to want a better way. Am I right in thinking that we should strive to charge the EV from DC current from the PV directly, and only convert the unused part to AC for household use and grid ? Seems to this electricity ignoramus that 5 kw charging with less than 10% total charging losses is feasible. Am I way off here ? I'd love to hear stories of people who are already diverting some of their PV production to local DC use, although I realize that in my case the efficiency nirvana will have to likely be delayed until my local utility is no longer paying me a production credit.
From a theory perspective, your desires can be met. From a practical perspective, it must be met if EV use is to be mainstream. I'm use to the best specialized converters being above 90% efficient, so all the technology is available for implementation.
To use the 480 DC port on the Leaf, you'd have to step up the PV Voltage a LOT. That's NO cheap thing.
480 volts DC can be acheived by putting 20 panels producing 24 volts, in series. Normally, 480V might be done on a large PV system. Of course, that will mean you need a smallish 480DC to AC inverter for the rest of you usage, they are not unheard-of, but not cheap. At that point, you need to know what the terms your electric utility is offering, to determine whether it is economically feasible.
I think this is a little trickier than just supplying the voltage. My Hymotion conversion has a fairly sophisticated charging cycle. I am not sure what would happen if the voltage that it sees were to drop at random (e.g. as the panels were shaded by passing cloud cover). I'd guess something would not be done properly. There's also the current draw issue. If you're going to charge a (say) 20 KWH battery in 4 hours, that's 5KW flow. If your system can't put out 5KW, there's a mismatch. I don't think the the manufacturers of existing systems would have bothered to make their on-board chargers robust to these types of issues. They're expecting a steady supply. So you'd either have to have a different on-board charger for this setup, one that is more robust to loss of voltage and can limit instantaneous current draw to whatever the output of your panels is. Or you'd have to supplement the output of your panels with grid electricity, to give the existing charger the steady input that it expects.
Absolutely, you will need an appropriate charge controller, but you need that anyways. Yes, the array needs to be correctly sized for the expected use. Standard PV charge controllers are designed for just these conditions, you just need one that matches the requirements of your battery pack. Perhaps so such animal currently exists. The biggest theoretical problem I see though, is having the PV panels and the car in the same place during peak sun. I suspect most people would find that hard.
I know of no current production charge controller that will support an output more than 48 vdc. Outback, Midnight Solar, Xantrex all, to my knowledge all max at 48 vdc. I suspect that has to do with certain UL/NEC issues regarding HV DC. I would argue that the OP's proposition that you lose 20% in each direction is a bit pessimistic. I think that grid tie inverters are closer to 90% depending on loading. So it may not be the big issue it seems. Icarus
Thanks to all for the discussion. I was wondering about this too. Charging during the day works just fine for me, but I wondered what fraction of EV users could say the same. One answer is direct DC charging at work. I know one large commercial PV installation in my city put the PV on shaded parking. Then there are all the home-based parents that shuffle their kids around town -- daytime charging suits them just fine. And then we can also include the obvious groups of work-at-home folks and those who are retired. Icarus, PVwatts suggests a 23% DC -> AC derating. If the inverter is 10%, then 13% is cable losses. Does that sound right ? How about the return trip -- how much energy loss will be saved in avoiding the AC -> DC conversion ?
Sage, I con fess I don't know exactly how PVwatts makes it's assumptions, but we generally use ~80% of name plate rating to compute averages. That would be a 100 watt PV might out out ~80 watts once it warms up. In the off grid world, I use a 50% derate from panel to end use, including wiring loses, controller loses, basic battery charging loss ( to those that don't know, I takes more energy to recharge a battery that that charged battery can return as energy,, about 20% loss in a FLA battery.) I know that my off grid inverters tend to be ~92% efficient DC to AC. Icarus
