I like the idea of reducing charging losses by a direct DC-DC link. Some things to ponder:
- Because of high DC transmission losses, the stationary storage should be located near to the parking spot, as in the picture above. That may not be a suitable location for many customers.
- To get a high charging rate, you must either have a large battery or discharge the battery at a high rate. The former is expensive, the latter contributes to cell degradation. If we cap the continuous draw from the battery at 1.5 C (about the same rate as 120 kW supercharging on an 85 kW battery), the battery would have to be 33 kWh to support 50 kW charging. That's about twice as much battery as typically needed to manage peak/off-peak power shifting and solar integration.
Ok running with that... I imagine this system would incorporate 1 main battery bank 30 to 50kwh in size, 1 main inverter, 1 DC to DC controller for controlling solar to batteries, 1 DC to DC controller for charging external batteries (such as an electric car). The goal is to have as few DC to AC and back type power changes in order to maximize efficiency.
The inverter would be much like what is already in the model S that can Pull or Push power between the battery and the AC side (grid), just locked at 60hz. That inverter would be sized appropriately to manage loads that a typical house might have (HVAC and such). In the model S this inverter can run over 240kw at peak and over 30kw sustained (I don't know the real specs). In my case 20Kw (80amps @240V) of power would be sufficient to run all my house loads at peak (not including car charging). This allows efficient on the fly power flow between battery, grid and loads. The system would also have a DC to DC controller to take DC power from the solar array and directly charge the house battery. Solar power would always be used to charge the batteries or get split between charging the batteries and maintaining the AC loads (via inverter). If AC loads and battery charging loads are less than what is being provided by solar, the extra power can be pushed into the grid. All programmable to sort out what is the best use of power, incorporating factor such as TOU, battery state, expected loads (predictive), actual loads, and even maybe using weather forecasts predict solar power availability.
Lastly would be the focus on the DC to DC controller for charging external batteries (home supercharger if you will). I envision a few scenarios:
-In scenario 1
The home owner may only want to "slow" charge their electric car at the end of a day. In this case they want to charge efficiently as possible with the least amount of power purchased from the grid. Time is not an issue so it can take a while. If solar power is available it would mix that into the DC charging. If the house battery falls below certain level, it could stop charging the car until more favorable TOU.
-In scenario 2
The home owner wants to recharge his car now, costs be damned I need to get somewhere type thing. In this scenario all of the systems resources can be brought online simultaneously. The inverter can be pulling from the grid to dump power onto the DC side. The solar panels can be brought into the mix as well if available. Such that the max power available from the battery + max power available from solar + max power available from grid - AC loads from the house such that the total power available from the system can be brought online to charge the car as fast as possible.
I am thinking about these things because it seems like all of these things are possible. Most of the hardware is out there and are already in use in these systems as stand alone parts. It's just that no one has a combined complete solution that brings it all together both hardware and software. If any company would be capable of doing this it would be Tesla. Solar city can install it for me. I am trying to come up with a more broadly focused product, not something niche for only certain customers in high utility rate zones.
Am I missing something fundamental here?