Home PV - Reconfiguring the Leaf battery modules

  Now that the battery pack has been disassembled and the modules balanced, they can be reconfigured from the Leaf's 96S2P factory setup, suited for providing about 400V to the car's drive system, to a 12S16P setup compatible with my inverter's 48V nominal DC input. Assuming a 4.2Vmax and a 3Vmin working voltage per cell (I still need to confirm these values for that 2015 battery pack type), the working voltage for the reconfigured pack will be within 50.4Vmax and 36Vmin (actually, 40Vmin due to the inverter's specs, more on that below).

 For those interested in the logic behind the 12S16P setup, see the details at the end below. But for now, let's get straight to the rewarding part, i.e. assembling the modules and finally having something to show after months of preparation, ordering hiccups and disassembly !

 The 48 modules are split into 2 subpacks of 24 modules each. Each subpack is configured as 6S16P. This allows to reuse the existing Leaf end plates and compression rods for at least one of the subpacks. And, more importantly, two humans together now have a chance to lift one of those almost-300lbs babies... 

 Above is a picture of the first subpack. The first step was to align flat on the same plane all the modules' electrical connectors to ensure a stress free connection and alignment with the bus bars later. In the picture, the connectors are facing the floor and are now aligned in the Z direction thanks to the 2x5 plank. Then 2 flat 2x5s were used to press the modules laterally and ensure that the connectors will also be aligned in the Y plane. Finally the rod screws were tightened until the pack length was the same as the length of the Leaf's original 24-subpack (i.e. ~32" 1/16).

 For the second subpack, new end plates had to be created and compression rods procured. The rods are just 1/4" 3ft threaded rods from the hardware store. For the end plates I simply cut in half three of the 2-stack end plates (on the right side of the pic) that the original pack uses for the 24 modules that lie flat in the Leaf pack. 3 of those were then stacked on each end (center and top of pic):

 Second subpack done:

 To keep each subpack well held mechanically, in addition to the compression rods, I bolted on 3 flat 0.25" x 1" aluminum bars to the few modules that sport 2 front and 1 back studs/nuts. This provides 2 benefits:

1. Mechanical stress relief. That's the function they were serving in the Leaf's original 24-subpack. They will now serve the same function, but across 2 subpacks. The subpacks weighing almost 300lbs each, should they be bumped or dropped during installation or maintenance, it'll help prevent the modules from shifting. This is important to avoid putting mechanical stress on the electrical contacts that will be connected to the rigid bus bars used for the center terminals
   
2. Guiding rails. This is a critical function for the coming bus bar setup. The 2 front bars can now be used as guiding rails supporting the holders that carry the floating bus bars (more on that in the next post, hopefully)

 Below is the fully assembled pack with the bars installed. In the second picture, note the 3 consecutive groups, each made of 8 modules with the same red contacts up or down. Since each Leaf module contains 4 battery pouches in a 2S2P configuration, each subpack is now organized into 3 groups in series x 2S, and 8 modules in parallel x 2P, i.e. 6S16P total.

 Finally, prior to installation, the front bars were wrapped in black electrical tape, to lower the risk of accidentally making contact with the floating bus bars, tools, connectors, etc.

And that's it, we now have 2 new 6S16P subpacks.


PS - battery reconfiguration considerations:

• 36Vmin (12 x 3Vmin) would be a problem ! Because the GS8048A Outback inverter's minimum DC input is 40V (64V is max). Note: the FM80 charge controllers support that range as well. 
• So I am planning to cut off discharging the batteries at 40V. Of course that will result into a loss of battery capacity. I don't know yet how much, and am planning to measure that either with a data logger (Leaf Spy) connected to a working 2015 Leaf car with the same 24KWh battery type, or with a State Of Charge controller on the completed system
• Another option would have been to consider one more group of modules in series, like 13S14P, but it'd have led to unused modules (10 in the case of 13S14P or 14S13P), and even less capacity (182 modules cutting out at 40V, instead of 192)
• BMS controllers are readily available up to 12 cells in series. Beyond that the choice of controller or board is more limited, or requires serializing them, which is more expensive
• I could also have explored going to a 60V pack setup since the GS8048A & FM80s support that as a nominal battery pack voltage as well
• However, I only intend to use this setup for a few years because upgrading to the newer Lithium car packs that are now, or soon to be, on the market (Bolt, 2018+ Leaf, etc). Those will reach about 60KWh capacity, more than double the "old" 2015 pack's 24KWh. So, at that time, once salvaged packs are readily available I'll revisit the setup and optimize some more
• For now my goal is to get this lead-to-lithium initial conversion done fairly fast and simply, so onward with 12S16P !