Hopefully this is an easy design question to answer. I'll start with some background information.
The case is with lithium iron phosphate or LFP batteries that are "48 volts" used for household photovoltaic battery backup.
There are many brands and there are different lithium chemistries, but we will be talking about the 16 S configuration using the 3.2 volt (nominal) LFP batteries. This creates a battery pack that is 51.2 volts nominal. (LFP 51.2V systems have a maximum operating range of 40V to 60V with 43.2V and 58.4 more typical for longevity.)
These battery systems typically use a balancing BMS to provide safety to the cells. However most of these BMS systems cannot be connected in parallel if you want to add battery storage. Additionally, the ones that can be connected in parallel must be connected to the same brand of BMS. The main reason for this is that if one battery pack has a higher state of charge than another battery pack, then when they are connected in parallel they will produce very high amounts of current causing the BMS units to disconnect then reconnect over and over rapidly which damages the BMS (in cheaper commodity type BMS, not Mil-spec).
Now the challenge. How would you design a bi-directional current limiter that could be placed on the common port terminals of the BMS output allowing no more than 50A, 100A, 150A, or 200A to be continuously delivered to/from the battery pack without the BMS shutting down? This added device would enable any brand battery to be placed in parallel with another without harming the cells or the BMS (given the same chemistry and voltage).
The system example is simply adding a second 10kWh battery pack capable of 100A to an existing one of the same specification, but 5 years older, and a different BMS. These will be connected to the same solar inverter with built-in MPPT charger.

 

Hello there welcome to AAC!

How would you design a bi-directional current limiter that could be placed on the common port terminals of the BMS output allowing no more than 50A, 100A, 150A, or 200A to be continuously delivered to/from the battery pack without the BMS shutting down?

Please clarify?

 

Bi-directional Current Limiter for common port BMS at 48V nominal. How would you design it to be lowest cost?
Better?

 

I wouldn't... paralleling LiPo at the BMS level is always problematic, as you don't know how the BMS will react. It would be better to pull the LiPo cells out and/or junk the existing BMS and apply a single new BMS over the whole cell-bank.

 

Not true. There are a many BMS units that 'talk' together on CAN bus. As long as they communicate together, they can work together. The most common are the server rack variety. The idea here is to create a limiter for the ones that don't talk or talk to different protocols. As in units from different manufacturers.

 

That's exactly my point... You need to be inside the BMS to know the current SoC of the cells to be able to determine charge rates etc else you'll be fighting the BMS every time. You aren't the first to try this... Why are there AFAIK no commercial units that do this? Do you know of any?

 

LFP cells have an inherent stability in charging in the middle part of their curve. Give a low SOC cell it's desired charge current (say 0.3C) and the charge voltage will remain relatively constant for 80% or so of capacity when the voltage starts needing to be elevated to continue to deliver the current. Stopping that voltage a safe range within the cell's chemical limitation (say 3.65V), you then see the current tail off significantly in the final percentages. Because the cells like to have this constant current, why would you not want a current controller/limiter to be part of the battery system? Hybrid inverters perform this exact same function outside the battery system already. What I'm asking is why not have this function built into the battery system itself in order to provide universal compatibility?