Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Looking for Knowledgable People on PV Inverters and Battery Augmented PV Systems

This site may earn commission on affiliate links.

lolachampcar

Well-Known Member
Nov 26, 2012
6,469
9,368
WPB Florida
Every now and then I've run into some very knowledgable people on different subjects here. I'm hoping to find such a person now with experience/knowledge on PV inverters and battery augmented PV systems.

Specifically...

I would like to install a reasonably large PV array on my house using a battery system for storage. The goal would be to keep my grid tie but reduce usage to zero.


My initial reading indicates that all current inverter systems are capable of dealing with in excess of 400 VDC on the PV input side but normally only manage battery banks in the 24 to 48 VDC range. My interest is in using a large Li-Ion battery pack (85 KWHr) which I believe runs between 330 and 370 VDC. I think a MS battery would augment a home PV system nicely. Does anyone want to discuss the hows and whys of "Y"ing PV and battery as input to an inverter? Of particular interest would be bucking unused (by the house) PV output to charge the battery during the day such that the energy can be employed during the night.

Thanks,
Bill
 
For solar with battery backup there are at least two inverters in the system. The first is the PV to grid inverter (DC to AC). This is the typical inverter on grid tied solar systems. The grid tied inverters are configured to shut down during power outages in order to prevent back feeding the grid. In other words when the power goes out you can't use the PV system to power your house, unless...

The second inverter (charge inverter) is used for the battery backup/off-grid. It is actually a combination of a DC to AC inverter, an AC to DC charger, and a transfer switch. The DC to AC inverter converts the DC current from the batteries to AC current for home use. The AC to DC charger converts the AC current from the grid or PV inverter to DC in order to charge the battery. The transfer switch isolates the system from the grid during power outages in order to prevent back feeding.

Most charge inverters are configured for charging lead acid batteries. The algorithms for lead acid charging are different than Li-Polymer batteries. So unless the charge inverter is designed for Li-Polymer batteries, you can't directly use the Li cells in place of the lead acid cells.

Assuming you have access to the DC output of the Li cells...

in order to use Li cells directly connected to the DC input of your system you would need a battery management system (BMS) capable of monitoring each cell and at a minimum shutting down the charging when a cell reaches a certain voltage. Depending on the charge inverter you may be able to hook up the BMS to a relay on the charge inverter in order to shut down the charger.

A second way to use Li cells as backup would be use a separate charger capable of AC/DC input and DC output. Under normal operation you plug the charger into an AC outlet and charge your Li batteries. Once charged you turn off the charger. The batteries will hold the charge for a long period, requiring only monthly "top offs". During an outage you would hook the Li cells up to the charger input and connect the DC charger output to the DC input of your charger inverter. Or you could have a conventional lead acid battery bank and charge them from the Li bank and charger.

Another scenario...some backup charge-inverters have an option to connect to either a generator or wind inverter (AC sources). Now if you could access the AC inverter for the MS, in theory you could connect the output to the backup charge-inverter secondary AC input. This would negate the need for a separate charging and BMS for the Li cells, since the MS has this already built in.

End of rambling...
 
My interest is in using a large Li-Ion battery pack (85 KWHr) which I believe runs between 330 and 370 VDC. I think a MS battery would augment a home PV system nicely. Does anyone want to discuss the hows and whys of "Y"ing PV and battery as input to an inverter?

Interesting. Are you planning to modify your Model S car to allow it to be connected to your house electrical panel, or have you somehow procured a Model S battery pack and want to hook it up to your house?
 
MS is so expensive to repair that buying a significantly damaged MS and retrieving the battery becomes an option. Given that I have about $45K in budget if I look at a straight line 10 year return, that gives me a lot of room to look at the MS battery as part of the solution. It really starts to make sense if you have three days of power reserve (on day one, diminishes with time and loss of capacity) for times when solar is not available. It also makes for a no compromise system where we do not have to think about what we do off solar hours.

Pollux,
The diagram SC shows has the PV and Battery "Y"'d into the inverter which holds promise. However, the battery is small which makes it unlikely they are dealing with MS battery voltage levels. I'll call and report back here.
 
I want to contribute some arm chair considerations

"Y-ing" the DC solar generator with the battery pack in front of the PV inverter
+ can use smaller inverter
+ simple
- DC pack voltage defines operating point for PV generator, this is less efficient than MPP tracking inverter
- need special inverter that is hooked up with BMS (you don't want your PV generator to overcharge the battery pack)
- need additional transfer switch to isolate home from grid in case of power outage

grid tied solar PV with separate battery storage that is tied to AC mains
+ more conventional setup like UPS (uninterruptible power supply)
+ flexibility to have dedicated circuits on UPS
- less efficient because of multiple AC-DC / DC-AC conversions
- need prevention for PV generator not to overcharge the storage unit in case of power outage

I think there is potentially big market demand for such a system, but you have to engineer for slightly different requirements first.
 
If I hear you correctly, the distinction you are making is to separate the UPS function from the PV function.

I'm planning on doing the control and related electronics so as not to rely on existing technology (assuming it does not exist - that is, there is no quality high efficiency inverter that excepts 350 VDC nominal battery input and speaks Tesla BMS CAN).

VP
Your approach would allow me to isolate the project and concentrate solely on the UPS and related integration function which I like. I could also use one of the Tesla chargers to charge the pack which also is helpful although wasteful given the DC to AC to DC conversion.

Some questions-
Is there any reason I should not still consider syphoning off DC from the array, putting it through a programmable buck converter and using that to charge the battery? I would need the ability to pull just enough off the PV array at any one time so as not to degrade the AC being supplied to the house. This would eliminate the DC-AC-DC losses and is really not that big a challenge given that I have to talk Tesla BMS CAN and manage the battery anyway.

Are there UPS systems that sync with the grid? Are they capable of supplying the house with sufficient current and yet not too much as net metering from a battery would be stupid?

Is anyone familiar with the ins and outs of PV inverters supplying only that which is necessary for the house without excess (net metered)? Florida Power and Light pays near zero for net metered power thus I would prefer to stick any excess in the battery.

As I see it, the challenge is the control implementation that keeps the PV and Battery systems doing their part in harmony.

Thanks for the input on the UPS segmenting of the system. I'm not convinced yet that a simply "Y"ing would not be the best approach as it allows for only one inverter in the system (be it supplied by PV or battery on the high voltage DC input) but I will need to better understand the design implications of varying DC input voltage on inverter performance.


As for market demand, it is one of the reasons I posted here. I can easily handle the micro controller based work along with the buck DC converter for charging. I would think there may be others here interested in playing with such a system. Being retired, I have time. Finding others with experience in PV/inverter systems would dramatically speed the process.
 
Unless you are really dollar constrained by the project (which it doesn't sound like) or are doing it in order to build a system from scratch (hobby) why not use existing technology instead of reinventing the wheel? While Li-Polymer cells are great for cars (higher energy density, lower weights) they are not necessary for stationary backup.

My most recent battery backup bank using lead acid, deep cycle cells cost ~$5K for ~40kWh. I have a bank of Li-Polymer cells for my other electric vehicle that with BMS cost about ~$14K for ~30kWh.

Take a look at the Sunny Island inverter chargers (SUNNY ISLAND 4548-US / 6048-US.SMA America, LLC). They provide all of the features you are looking for (charger/inverter/backup/scalable) except for the Li cell requirement.
 
idoco,

I would think that 40 kW of lead acid would be a huge bank of batteries. Are there any out gas issues with those lead acid batteries? I must admit, the value seems to be there and the inverter would manage the battery which is also very nice indeed.

VP,
' do not know you well enough to know if you are joking or not. If you're not, that still means doing a DC-DC and manager that speaks Tesla BMS CAN. If I am going to do that, better to drop the losses and go PV to battery then battery direct to inverter when the PV is not producing. I think the panels have diodes built into them which would lead me to believe a simple diode or'ing in of the battery would do the trick. Ten to fifteen amps across a diode is not much of a loss.
 
A couple of thoughts for consideration

- Might want to check into whether time-of-use rates and net metering are available in your area. In places where they are available, it could be a money loser to store solar-generated power for use during off peak hours. I send solar-generated power to PG&E at 16-39 cents/kWh during the day, and charge my Model S after midnight at 10 cents/kWh.
http://www.pge.com/tariffs/tm2/pdf/ELEC_SCHEDS_EV.pdf

- Some utilities are not approving net metering tie ins for solar installs that have battery backup systems.
Fight Over Battery-Backed Solar in Southern California : Greentech Media
 
Last edited:
FPL pays squat for net metered power (2-3 cents if I remember correctly) while charging 12 cents. My goal would be to size the array such that it powers the house during the day with excess PV going to the battery. The battery then provides for nighttime needs as well as acting as an overall back up for up to three days should there be an issue with FPL and the sun :)
 
No joking but I am in no way qualified to advise en detail on this. It's just that I think building from existing parts is often more easy, and worth to shed a few % of efficiency.

BTW all PV and battery systems I know that can run in "island mode" (=grid outage) - they have a BIG resistor bank to burn surplus power in case production exceeds demand & batteries are full. Please consider if you would need such a component too, and how to integrate into your design.
 
idoco,

I would think that 40 kW of lead acid would be a huge bank of batteries. Are there any out gas issues with those lead acid batteries? I must admit, the value seems to be there and the inverter would manage the battery which is also very nice indeed.

For a vehicle, yes. For a stationary application where the batteries are in a garage or basement, no. The batteries I have weigh about 100lbs each (16 total). Here is a link to the page (U.S. Battery Solar RE L16XC Renewable Energy).

I've been using lead acid batteries for backup for over two years. Prior to that I was using them for my electric vehicle for several years.

Lead acid batteries do gas during equalizing charges, something that the inverter/charger is programmed to do a few times a year in order to balance the cells. When this occurs there is some release of hydrogen sulfide gas. As long as the space is well ventilated this should not be an issue.
 
For a vehicle, yes. For a stationary application where the batteries are in a garage or basement, no. The batteries I have weigh about 100lbs each (16 total). Here is a link to the page (U.S. Battery Solar RE L16XC Renewable Energy).

I've been using lead acid batteries for backup for over two years. Prior to that I was using them for my electric vehicle for several years.

Lead acid batteries do gas during equalizing charges, something that the inverter/charger is programmed to do a few times a year in order to balance the cells. When this occurs there is some release of hydrogen sulfide gas. As long as the space is well ventilated this should not be an issue.

If you are in the USA, code requires that the batteries be in a box vented to the outdoors.
 
I called Solar City and I thought for a moment that I was talking to Tesla :)

I spoke with a nice enthusiastic young gentleman who was perfectly happy to talk leasing a full house system for California. When we got to me being in Florida, wanting access the the battery back up, wanting information on the size and capability of the systems,,,, we ran out of run way. SC does not sell anything. They are still in pilot with the battery system and, from what I gathered, do not have enough batteries to sell systems even if they wanted. The person I spoke with had no idea if their battery was integrated at a higher than common 24V-48VDC level. I did learn that they are packaging someone else's inverter with their systems but he did not know who's inverter. Anyone out there with a SC installation that knows which inverter they are using?
 
Yes, it is a small battery. I failed to wonder out loud whether a bigger battery might be built out of these smaller packs.

MS is so expensive to repair that buying a significantly damaged MS and retrieving the battery becomes an option. Given that I have about $45K in budget if I look at a straight line 10 year return, that gives me a lot of room to look at the MS battery as part of the solution. It really starts to make sense if you have three days of power reserve (on day one, diminishes with time and loss of capacity) for times when solar is not available. It also makes for a no compromise system where we do not have to think about what we do off solar hours.

Pollux,
The diagram SC shows has the PV and Battery "Y"'d into the inverter which holds promise. However, the battery is small which makes it unlikely they are dealing with MS battery voltage levels. I'll call and report back here.