ElectricBikeSquatch
New here
Ok, I have a theory/battery bank topology design question for an extremely hypothetical electric hydrofoil (like everybody and their dog these days?)
Let's say you can buy a big pile of 173ah CATL LiFePO4 cells at an amazing bargain. Let's say 72v nominal (76.8v, if you actually multiply the 3.2 V nominal by 24, though) is perfect for the waterproof motors and ESCs you want to use. Let's say you can make a somewhat arbitrary number of 24s strings, like as many as 6 or 7 out of the salvaged battery backup unit you see for sale.
I've been trying to figure out the best way to put those cells to use for a boat that might call for as much as 30-100 kwh of capacity and might want to draw N x 1C (~173 amps) for 2-4 ESCs running as many motors. The motors will draw ~120A each at their maximum continuous rating, but you're talking about a big enough boat that to legitimately use the ~250A burst rating they're capable of, if you have the cells to feed them below those cells' 172A continuous rating. And you'd like four motors rather than two, ideally, so 480A continuous from the whole battery bank is the goal, which you need a minimum of 3P to achieve within the CATL cell specs.
The 'dumb, straightforward, and technically functional' solution of one pack per BMS per ESC, with everything isolated has all sorts of pros/cons. I'd much rather have a singular bank with robust safeties and the ability for it to function with partial failures. Just as an example, it would be much better in every practical scenario I can think of to be able to run all of four motors on three of four 24s strings if one cell in one of the four packs failed, than it would be to be stuck with three of four motors each on its own isolated 24s string/BMS.
I also definitely don't want to be running all 2-4 motors on a single string until it's empty then swapping to the next one, just because of how much it would limit maximum power to be in that kind of a manual 'limp mode' all the time. I could see that getting you back to land if you had any number of different specific potential failures, but it seems like a terrible way to *plan* to use enough cells that you could otherwise cruise at full speed.
Anyway, so, that's why I want to do what I'd want to do, here's the actual question about how I'd want to do it (in two parts):
If I wanted to do 24s4p with Dilithium, and I got one 'Controller' and three 'Satellites,' would it be as simple as connecting the main/satellite strings in parallel instead of series, putting some appropriately sized fuses in strategic places for each string, and assigning some values/ordering the Dilithium components with the correct initial parameters? This seems worth the cost increase versus using cheap BMSs and 'human eyeball' safety monitoring when the packs need to be in parallel if it would work. I also like the idea of being able to choose how many strings get loaded onto the boat for a given trip, if I can disconnect and reconnect satellites as wanted (and occasionally pull all the cells and shuffle/pick and replace as needed).
If that main goal (a 24s4p pack) won't work with Dilithium, how about an Orion II BMS with the biggest, 180-cell (or smaller, for a smaller overall pack) total capacity? Will that run 24xN series strings in parallel like I want? For the one that does 96 cells, it's about the same price as Dilithium but seems worse-adapted to being modular/expandable like I'd want if it was an option, but maybe the Orion dissertation on parallel topology was just too dense for me and explained exactly how to do that? Even for the version that uses the biggest bargain surplus pile of CATL cells I've seen gets you a ~92kwh 72v battery bank for ~$5400, which seems super compelling if the Orion II would work for this.
Let's say you can buy a big pile of 173ah CATL LiFePO4 cells at an amazing bargain. Let's say 72v nominal (76.8v, if you actually multiply the 3.2 V nominal by 24, though) is perfect for the waterproof motors and ESCs you want to use. Let's say you can make a somewhat arbitrary number of 24s strings, like as many as 6 or 7 out of the salvaged battery backup unit you see for sale.
I've been trying to figure out the best way to put those cells to use for a boat that might call for as much as 30-100 kwh of capacity and might want to draw N x 1C (~173 amps) for 2-4 ESCs running as many motors. The motors will draw ~120A each at their maximum continuous rating, but you're talking about a big enough boat that to legitimately use the ~250A burst rating they're capable of, if you have the cells to feed them below those cells' 172A continuous rating. And you'd like four motors rather than two, ideally, so 480A continuous from the whole battery bank is the goal, which you need a minimum of 3P to achieve within the CATL cell specs.
The 'dumb, straightforward, and technically functional' solution of one pack per BMS per ESC, with everything isolated has all sorts of pros/cons. I'd much rather have a singular bank with robust safeties and the ability for it to function with partial failures. Just as an example, it would be much better in every practical scenario I can think of to be able to run all of four motors on three of four 24s strings if one cell in one of the four packs failed, than it would be to be stuck with three of four motors each on its own isolated 24s string/BMS.
I also definitely don't want to be running all 2-4 motors on a single string until it's empty then swapping to the next one, just because of how much it would limit maximum power to be in that kind of a manual 'limp mode' all the time. I could see that getting you back to land if you had any number of different specific potential failures, but it seems like a terrible way to *plan* to use enough cells that you could otherwise cruise at full speed.
Anyway, so, that's why I want to do what I'd want to do, here's the actual question about how I'd want to do it (in two parts):
If I wanted to do 24s4p with Dilithium, and I got one 'Controller' and three 'Satellites,' would it be as simple as connecting the main/satellite strings in parallel instead of series, putting some appropriately sized fuses in strategic places for each string, and assigning some values/ordering the Dilithium components with the correct initial parameters? This seems worth the cost increase versus using cheap BMSs and 'human eyeball' safety monitoring when the packs need to be in parallel if it would work. I also like the idea of being able to choose how many strings get loaded onto the boat for a given trip, if I can disconnect and reconnect satellites as wanted (and occasionally pull all the cells and shuffle/pick and replace as needed).
If that main goal (a 24s4p pack) won't work with Dilithium, how about an Orion II BMS with the biggest, 180-cell (or smaller, for a smaller overall pack) total capacity? Will that run 24xN series strings in parallel like I want? For the one that does 96 cells, it's about the same price as Dilithium but seems worse-adapted to being modular/expandable like I'd want if it was an option, but maybe the Orion dissertation on parallel topology was just too dense for me and explained exactly how to do that? Even for the version that uses the biggest bargain surplus pile of CATL cells I've seen gets you a ~92kwh 72v battery bank for ~$5400, which seems super compelling if the Orion II would work for this.
