DC Charging of the Model S from a DC Home PV Solar Installation

[InTheShadows]

Wk057's solar panel circuits are optimized for charging a low voltage battery pack like he has installed and like the ones we install for our battery back up and off grid customers.

Standard grid tie systems usually operate at 300-450vdc. This is more in line with the tesla battery in the car.

Wk knows the float voltage of our car packs, I think they are in the 400v range but am only going from memory. So to have a standard grid tie solar panel pushing power into your battery pack in your car you need at least 110%, preferably 125% of the float voltage to get the juice to go into the batteries.

With those voltages you are bumping against NEC code for max residential PV voltage of 600.

IMO this may be an area that the powerwall is going to fill/address. It operates at 350vdc easily keeping within the bounds of NEC. if they are putting a step up DC-DC converter on it, that could easily regulate the voltage needed to charge our cars with out having to worry about the 600vdc limits.

Just my personal opinions though.

 

[Chris TX]

You'd need to build a Chademo that accepts DC input, they're mostly expecting to receive three-phase AC, aren't they? At least, the one ABB unit I just checked the specs on does.

[edit] And variable DC input, at that. Ick. You would need a very flexible (and efficient) DC/DC converter. [/edit]

This is the CHAdeMO charger I was looking at. The "DC - 450V or below" is what I was looking at.
Electric Motor Werks, Inc. - QuickCharge-25000 HV - a 25kW PFC charger for HIGHER voltage batteries

 

[Bangor Bob]

This is the CHAdeMO charger I was looking at. The "DC - 450V or below" is what I was looking at.
Electric Motor Werks, Inc. - QuickCharge-25000 HV - a 25kW PFC charger for HIGHER voltage batteries

Well that's cool! The price of the cable and CHADeMO connector though, YOWZA!!!

 

[wk057]

Well, for Tesla I would personally skip the CHAdeMO connector and just butcher a Tesla CHAdeMO adapter...

 

[Chris TX]

Well, for Tesla I would personally skip the CHAdeMO connector and just butcher a Tesla CHAdeMO adapter...

I would too, but people would question my concern for safety

 

[wk057]

I would too, but people would question my concern for safety

Safety? That's what electrical tape is for!

 

[Chris TX]

Safety? That's what electrical tape is for!

My homeowners (and life) insurance is paid up. I'm game.

 

[Ingineer]

It is certainly possible to do this, and from my analysis the Model S is fine with reduced input on DC. This is essentially the same thing that happens when you share a Supercharger cabinet and get less than "full" power. The car tells the Supercharger the max charging power, and it's up to the Supercharger to attempt to deliver as much as it can as long as it doesn't exceed the limits coming from the BMS in the car.

The only argument I can see for such engineering time and hardware expense is in the rare edge-case of a grid outage. Most of us charge at night so direct PV to car doesn't make sense. In the event of an outage it's also likely you wouldn't be going to work, so you might be able to take advantage of the excess PV at home. However, you'd then be at home and would likely want to have that energy for heating/cooking, etc. Since you aren't going to be doing much driving, you don't really need the charge. There's just no much logic to it, elegant or not.

You are much better off taking the money that you'd waste on the implementation of such a system and just buying a propane-sourced generator for emergencies. Propane keeps almost indefinitely in storage unlike gasoline, and you would then be able to fuel your car and provide make-up energy for your home in the event that bad weather reduces your solar contribution. I don't even think it's worth investing on a PV system that can island unless you have very unreliable grid power.

I see an inverter that can run off the large pack in the Tesla and provide back up power to the home as a more useful device. A High-voltage inverter that can take the ~400V pack voltage is smaller, cheaper, and more efficient than one that runs off low voltage DC (12v). It's just nobody has started making one. I have a prototype I've been perfecting that produces pure sine-wave output, is highly efficient, and is compact. I hope to offer it to everyone soon in some form.

It's also handy on trips and for camping, and would be even more at home in the Model X. It's so efficient that it can be left on even for small loads, such as charging laptops and tablets, but can also easily run a fridge, microwave oven or other appliances and tools such as an electric chainsaw.

 

[Oba]

Of course, this has (not surprisingly) already been done:

CHAdeMO Solar Powered Charger
Andromeda Power Solar Charger | Joshua Tree, CA | Electric Car Charging Station | PlugShare

 

[qwk]

About the only logical thing that comes to mind as why anyone would do this, is for efficiency reasons. Doing this eliminates conversion losses(not huge by any means), but over time they would add up.

 

[widodh]

About the only logical thing that comes to mind as why anyone would do this, is for efficiency reasons. Doing this eliminates conversion losses(not huge by any means), but over time they would add up.

And, because you can. Not everything has to be cheaper money wise.

 

[green1]

And, because you can. Not everything has to be cheaper money wise.

Exactly! If I wanted the cheapest option I'd certainly not be driving a Model S!

 

[Ingineer]

To me this efficiency argument smacks of the same argument of whether or not to mount your PV on a motorized tracker. Almost nobody does now, as more panels solve the problem with greater reliability and much less obtrusiveness.

Just add a few more panels. Or take out the interior of your model S. =)

 

[israndy]

I charge my two Model S with my off-grid solar system. When there is sufficient sun the cars do charge directly from the PV power without using battery storage. PV->Charge Controllers->Inverters->Model S AC to DC Charger->Battery. Long road

That might be interesting. I have finally joined the high capacity electric vehicle club and find that charging at home is going to be expensive. I pay 28₵ per kWh, if I go to a SuperCharger it's only 26 cents. I called Tesla about getting panels to reduce the price I pay (even before the car I had a $500 bill in December for heating my hot tub and stuff). They said that my town was already saturated so NetMetering was no longer available, thus the payback for a solar system would be decades. SunRun confirmed that it's not feasible to do solar in Alameda, CA

I started thinking about all those cheap solar systems that people are taking down to upgrade. If I could just get a bunch of panels I don't need to feed the grid for no payback or feed my house just to run my TV. I need to feed my battery pack, in my car.

Has anyone designed an Off Grid Solar System that uses little or no batteries? When the sun comes up I could have the car start charging, but I don't know what happens when a cloud goes over the panels...

-Randy

 

[wk057]

That might be interesting. I have finally joined the high capacity electric vehicle club and find that charging at home is going to be expensive. I pay 28₵ per kWh, if I go to a SuperCharger it's only 26 cents. I called Tesla about getting panels to reduce the price I pay (even before the car I had a $500 bill in December for heating my hot tub and stuff). They said that my town was already saturated so NetMetering was no longer available, thus the payback for a solar system would be decades. SunRun confirmed that it's not feasible to do solar in Alameda, CA

I started thinking about all those cheap solar systems that people are taking down to upgrade. If I could just get a bunch of panels I don't need to feed the grid for no payback or feed my house just to run my TV. I need to feed my battery pack, in my car.

Has anyone designed an Off Grid Solar System that uses little or no batteries? When the sun comes up I could have the car start charging, but I don't know what happens when a cloud goes over the panels...

-Randy

Seems like the better solution would be to get out of California and stop overpaying for everything in general.


Barring that, there's not really a way to charge directly from PV with off-the-shelf components and without hacking the HV system of the car.

In theory this could be done with a DC-DC and the supercharger protocol, but honestly this would be a pretty poor use of PV if this is all it were used for. You'd be better off just building a battery-backed off-grid system capable of handling your power needs and saying 'screw you' to the utility and that that $0.28/kWh (except as a backup, if your keep-the-meter fee isn't too ridiculous).

Doing off-grid instead of grid-tie pretty much doubles the total cost... but competing with $0.28/kWh it should be a no-brainer. It worked out for me with about a 15 year ROI here where I have $0.10/kWh grid power available... so would make a lot of sense in your area if done right. Off-the-shelf stuff like Tesla's setups are going to be back in that 15-year ROI range even in your area, though, since they're ridiculously expensive. But if you can either do the work yourself or know someone who can you can get a great setup that's cost competitive against the grid. Using recycled Tesla vehicle batteries (I can source them for you at under $250/kWh in decent quantities (80+ kWh)) for the heart of the storage setup, and hardware from existing reputable companies like Outback Power and Midnite Solar, you can make a great setup at half the cost or better vs something from Solarcity/Tesla directly.

My home setup has been at least partially online for three years now (fully online for about 2.5 years). I love it. Forget the grid.

 

[israndy]

Maybe I wasn't clear. I am hoping to generate enough power from the panels to generate 110 or I guess 220 volts thru an inverter. I then want to plug in a EVSE to that inverter and that into my Tesla to make the battery pack in the car be my battery storage solution. Yes I would need to charge during the day, but I am retired and can go many days between big trips.

Since I can tell the Tesla to charge certain hours I would just pick daylight hours. I just thought there might be some problems with that setup so I was asking, well, you, cause you have an off-grid system already. Perhaps I just build a normal off-grid system, but for storage, I would just have one old car battery. That would leave all the power the solar panels are generating to feed thru the inverter and power the EVSE.

Will it blend?

-Randy

 

[arg]

Maybe I wasn't clear. I am hoping to generate enough power from the panels to generate 110 or I guess 220 volts thru an inverter. I then want to plug in a EVSE to that inverter and that into my Tesla to make the battery pack in the car be my battery storage solution. Yes I would need to charge during the day, but I am retired and can go many days between big trips.

This doesn't work, because you need to match the varying power output of the panels to the constant power input of the car. Although you can manipulate the car's charging rate by changing the J1772 pilot signal, it takes a couple of seconds to respond. There's also the limitation that J1772 doesn't go below 6A, so at low outputs from the panels you need to shut down altogether; that's not a problem if you have a large number of panels, but then you would be wasting most of that output on sunny days.

A standard solar inverter (intended to be grid tied) is absolutely not what you want, as these are designed to extract maximum power from the panels on the assumption that the grid will accept whatever is thrown at it (and will shut down completely if it doesn't).

In theory you could build something with a very small battery (or supercapacitors or whatever) to store the few seconds worth of power needed to smooth the fluctuations from the panels to match what the car is taking, but now you've built most of what's needed for an off-grid setup, so more useful to fit a bit more battery and have something that can run things other than the car.

But in practice, unless your utility completely prohibits attaching more solar to the grid (rather than just refusing net metering), the practical option to do what you want is to do grid-tied solar, and manage your loads such that you hardly send anything to the grid. In effect, you would use the grid to provide that very small amount of storage to match the solar generation to the car. There are various EVSE designs around that will monitor the solar generation and charge the car only when power would be 'wasted' back to the grid. Similarly for other loads such as water heating.

 

[israndy]

Thanks for the input, I knew there was a reason it wasn't gonna work. Thank for making it clear. I'll keep an eye out for systems that control turning things on based on solar input...

-Randy