diy solar

diy solar

Audi E-tron battery into powerwall

SwedishDIYer

New Member
Joined
Apr 7, 2023
Messages
12
Location
Sweden
Hi! First time poster here.

I currently have an Audi E-Tron battery pack that I want to use to make a powerwall.

The battery pack is 71kWh, has 27 modules with 12 cells each. Each cell is 3,66 volt and 60Ah. It’s LG Chem and model LGX N2.1.

I’d like to build a 3 phase battery storage out of this, preferably compatible with my Growatt solar system, but I can also do with a “stand alone” solution.

Any pointers on what else I need? BMS, inverter and so on.

Best regards!




 
Yes, they can be re-arranged. I have not yet decided what system voltage I want, I was thinking about 48v, but 144v might work too.

Growatt has some nice looking inverters (SPH10000TL3 BH-UP) but they want 100-500v battery voltage and I’m not sure if I can find a BMS that works with that voltage and Growatts inverters.
 
I would simply arrange each module into a single "cell". build 2 13S batteries for 48V (or one 2P13S battery) and have a spare cell/module.

There are a couple of github projects that integrate JK and JBD (and others) BMS with growatt.
 
Thanks, I’ll have a search on GitHub ? Still not sure about the voltage, can the SPH inverters speced to 100-500v handle a 48v battery?
 
Thanks, I’ll have a search on GitHub ? Still not sure about the voltage, can the SPH inverters speced to 100-500v handle a 48v battery?

I would assume not, but things will get way more complicated and expensive if you go high voltage. Very few DIY BMS exist that can go to that level, and the ones that do are targeted for the EV market... $$$$ like the Orion.
 
Yes, going for the lower voltage setup seems to be the best path.

I have 324 cells to work with, so I have the option to make 22 packs with 14S1P to reach 51V. Is there any point in making 14S2P instead?

Also, what BMS is considered to be "state of the art" for this kind of project? I´ve been looking at both Seplos, Daly, Batrium and some no-names from Alibaba but I havent decided what to use. I want it to be as safe as possible and not only go for the cheapest possible solution.
Edit: I have decided to go with Batrium.

One thing that I have not yet been able to solve is what inverter I should choose.
Grid connected, 3 phase (3x 230v). I also have a 27 kWp solar array connected to a Growatt MID25KTL-3 (peak solar power is about 18kW due to roof angle and direction).

I want to be able to charge batteries from grid and/or solar panels and also export from batteries to grid (arbitrage).

A 20-25kW 3-phase hybrid solar inverter with support for 48-51V batteries would be the perfect match - but I havent been able to find one.

The options I have found so far is this:
1. Use one Deye SUN-12K-SG04LP3-EU only for batteries and keep my Growatt for the solar panels.
2. Use 3x Victron Multiplus-II 48/5000 for batteries only and keep my Growatt for the solar panels. Probably the most expensive solution.
3. Other suggestions?
 
Last edited:
I started to disassemble the e-tron battery today, and it seems as if I need to re-think this a bit.

The batteries are 3P4S and they are not easily taken apart. My idea was first to make some 14S1P packs to reach 51V. I guess I’ll have to use four of these packs in series to reach 58,72V nominal instead.

This makes it a bit more complicated when it comes to BMS. Do I need a separate BMS for each of the 27 packs?
 

Attachments

  • 9CFB8ADC-7458-40CB-BC97-9E9AB661A91A.jpeg
    9CFB8ADC-7458-40CB-BC97-9E9AB661A91A.jpeg
    244.6 KB · Views: 51
  • 51E7E4D6-173A-4068-98B1-D10EB9B527BD.jpeg
    51E7E4D6-173A-4068-98B1-D10EB9B527BD.jpeg
    219.8 KB · Views: 50
  • ABC5D8D8-9281-4EF5-8320-7CF4A3DEBCE8.jpeg
    ABC5D8D8-9281-4EF5-8320-7CF4A3DEBCE8.jpeg
    142 KB · Views: 52
  • 3FEDACFF-0924-477C-AE14-6FD85E76CD03.jpeg
    3FEDACFF-0924-477C-AE14-6FD85E76CD03.jpeg
    255.2 KB · Views: 48
  • 3F0AE733-BE3F-45C7-8B60-75251C84D307.jpeg
    3F0AE733-BE3F-45C7-8B60-75251C84D307.jpeg
    160.9 KB · Views: 48
  • CD00B05E-9539-4866-B2DF-7BF671DB2787.jpeg
    CD00B05E-9539-4866-B2DF-7BF671DB2787.jpeg
    179.2 KB · Views: 49
Given the count, 12S might be more desirable. There is very little capacity left below 3.5V/cell. I recently did a run to 0% on my PHEV NMC cells, which should be very similar to yours:

1682252922111.png

At 3.5V, I was sitting at 8% SoC

For 12S, that would be 42V - the default cut-off for 48V lead acid, and most can be set lower.

Tesla modules are 6S, and folks will put two in series for 12S/48V.

With 27 modules, that would give you 9 parallel batteries.

On the other hand, 16S would practically limit you to 4.0V/cell on the top end, and you would lose 20% at the top, BUT you would have a few spare modules.
 
I am currently looking at the Victron system, and I've started to do "shopping-list". If anyone with know-how want to chip in, I´d appreciate it - the Victron eco-system is still new to me and all the information and documentation available feel a bit overwhelming ?


3x Victron MultiPlus-II 48/5000/70-50
1x Cerbo GX
1x EM24
3x Current transformer 100/5

I also need
1x Shunt (suggestions?)
1x Contactor (suggestions?)

What else?
 
12 cells each. Each cell is 3,66 volt and 60Ah.
That will be the biggest challenge to find an inverter that can operate at the voltage range of one of those modules. Assuming NMC or similar chemistry with an individual cell voltage range of 3.5 to 4.1 volts that translates to a voltage ranges for the inverter of a low voltage of 42 Volts and a high voltage of 49.2 volts. As mentioned earlier that is similar to two Tesla Modules.
The next challenge will be paralelling them and monitoring with a BMS. You would need a separate BMS for each module unless there is some way to paralell the cell groups. In that case depending on the physical difficulty in doing that you could get by with one BMS but the interconnections might get tricky trying to keep them the same length and finding an electrically secured way of doing that.
 
This makes it a bit more complicated when it comes to BMS. Do I need a separate BMS for each of the 27 packs?
I confirm you than 12S will give you a good battery to work with 48V inverter. Many inverter can be program to have a 40V low discharge cut off (3.3V per cell).
As about connecting module in parallel to reach higher capacity, I can suggest you to hardwire cells in parallel to create a single 1080Ah cell.
In short, starting with 24 modules, you take 6 modules, you connect them in parallel at cell level, of course you also connect modules in parallel with the bolts a terminals and you do this for 3 group of 6 modules to end with 12S 6P or 44V nominal 1080Ah.
You can also decide to build 2 batteries for redundancy or/and lower weight. That way you will not relay on multiple BMS to have a reliable large battery, but instead you will relay on wires... I consider this reliable.

I did this several time to increase capacity of a battery. Sometime it's easy with bolts, sometime you have to drill holes or weld.
Below is a extreme example when, back to 2014, I welded two groups of 45Ah Chevy Volt cells together to create a 90Ah cells. It was to passed 800A during acceleration of an EV, so there the big connecting wires. In the case of a solar battery you can certainly end with wires able to pass 4-5A just in case there is small imbalance in capacity between cells and the current will flow.
Volt cells parallel.jpg
 
Below an example of Audi E-tron 14.6V 360Ah battery.
View attachment 149922

Thank you, that's valuable input.

Just to make sure that I understand you correctly - the orange lines you drew are for balancing the cells (if any imbalance occurs) and there won't really be any large currents passing by these cables?

And two of these 3P4S-packs, joined like that would be called....what? 6P4S? 3P4S2P?

Given the number of packs I have (27). Perhaps 12S 9P (1620Ah) would be more suitable? That would bring down the number of BMS's to 3.
 
Back
Top