diy solar

diy solar

2nd Gen Powerwall/ 280 (272Ah) LiFePO4 16s 48V

CU17

New Member
Joined
Jan 12, 2021
Messages
7
Skip this if you're only interested in the new system.
********

I am new to this forum so I'll take a quick second to talk about my 1st Gen system.

Right now I have a 9.8kWp PV System with an 8kWh DIY Powerwall consisting of 18650 cells. It runs as a grid-tied self-consumption optimizing system only (i.e. no backup/island/transfer switch functionality) The setup is your standard 14s80p for some 160+Ah. The PV system was built before I added the battery and since the PV system runs on AC tied SMA inverters, I decided to add a Home Manager and a Sunny Island to simplify things.

The battery is being protected by a Batrium which also communicates through CAN with the Sunny Island, controls the main connection relay (EV200) and fans to cool the cells / Longmons for longevity.

I also built a little PCB that includes a buffer circuit to run the pickup current of the relay through an external transistor instead of the Batrium, manual pre-charge, and a buffer circuit for the fan control. All subcircuits also have colored LEDs attached to them to indicate the current operating state.

In addition to the relay, the battery is passively protected by 1.5A cell fuses and a 125A main breaker.

*********

The new system will consist of 16 LS272Ah cells. I will basically reuse all the equipment from the previous system, make it a nicer / neater built, and add two Longmons.

Here's my thinking for the enclosure and connections:
My Sunny Island can provide a max current of 110A, but is unlikely the battery will see that often. The average discharge current is around 12A and the charge current varies by season but will rarely exceed 90A. While the SI can run 4.6kW continuously, 6kW for 30 min, and 11kw for 3s, I am planning on limiting the battery current to 120A - which will keep me below 0.5C at all times and my average current will be around 0.05C. This should ensure a high cycle life and little swelling.

I am planning on using the provided bus bars and 50mm2 cable to make the connections because that is plenty for my average loads.

I have read through a couple of manufacturer datasheets, talked to my old contacts in the EV industry, and read through the posts here, and decided NOT to compress the cells. However, I will build an enclosure for them:

The plan is to build a concrete enclosure for two rows of 8 cells. There will be a little extra space to include either insulation sheets or extra tape around the cells as I don't trust the blue wrapper they come in. Other than that the cells should fit in tightly to prevent expansion. So I am not compressing them, but I am enclosing them at a low SOC without the ability for them to expand. The concrete will be sealed with epoxy or something similar to prevent it / the cells from becoming wet in humid conditions. I will also include four little "baskets" to put SUPERSORB in. I have a bunch of that stuff lying around and get more every time I buy shoes (it's in the little bags that come with shoes etc and meant to keep the humidity out of the product) so that will work great and it absorbs crazy amounts of humidity as well.

Other than that, the Batrium shunt and the main relay will go into that box as well as all the Longmons. I am aware that I won't get accurate temperature readings of the cells with the Longmons since I can't put them directly on the cells, but I am not expecting a lot of heating at my planned C rates. This way I can use 8 of my 14 fans for 2 Longmons each and 4 fans to provide circulation across the top of the cells (2 for each row) and bus bars.

The Batrium and my own PCB (it's actually just a perfboard at the moment) will stay in the box they are in right now.

Let me know what you guys think.
 
Back
Top