0
\$\begingroup\$

I have a 48 volt, 20 kW motor that I need to supply with power. I am looking at a configuration of 16 LiFePo4 cells each in series, each 200 Ah rated at 1 C. At that rated capacity, it would seem the most power I could draw is half that, or 9.6 kW which means a lot of unused power to the engine.

If that is correct, the 200 Ah cells must be 2 C rated to make use of the 20 kW of power. Is that correct? (Note: the 20 kW is occasional short-duration peak, not the normal range. Normal use would be 10 kW - 15kW.)

Then I thought, what if I had a larger motor than 20 kW, say 40 kW? How do you feed a hungry power draw like that if only 2 C 200 Ah cells were available? If I had 32 cells, divided into two 16 bank 48 V batteries (say, bank 'a' and 'b'), and banks 'a' and 'b' were wired in parallel to each other, would that satisfy the current draw to a 40 kW motor?

\$\endgroup\$
0
\$\begingroup\$

Yes, it's perfectly acceptable to put banks of cells in parallel. You'll actually commonly see this in battery packs. If a pack has the designation "5S3P" for instance, that would mean that it has stacks of 5 cells in series, and has 3 "banks" of those in parallel.

Have you worked with Lithium-Ion batteries before? If not, I'd advise extreme caution for your use-case. You're dealing with quite a lot of potential energy, and if these batteries fail it can easily become destructive and dangerous (gouts of flame or explosions). Particularly in your case, you'll need to take precautions against thermal runaway (where if one cell fails and vents/overheats, the heat can spread to the adjacent cell, and cause it to vent/overheat, causing a chain reaction which can easily destroy an entire battery pack, and everything else that happens to be in the same room as it).

You should take care to ensure that the batteries are sufficiently protected, and that they have dedicated battery monitoring circuitry keeping them safe (preventing overcharge, overdischarge, temperature monitoring, overcurrent protection, etc.). You'll need to also ensure that you are balancing the cells that are stacked in series (most BMS solutions will include some way to do this). Moreover, you must ensure that the cells are all at identical, or nearly identical states of charge when you hook them all up in parallel. If one cell is at 3.9V, and another is at 3.4V, when you plug them in to one another an extremely large current will flow between them, until they balance with one another. This current can and will destroy the cells if it is large enough.

\$\endgroup\$
  • \$\begingroup\$ I have not worked with LiFePo4 before. I am trying to learn as much as I can in order to design, configure, and purchase this system. I will be shopping for a BMS with Lithium settings. The cells will be bottom-balanced before first use, though I'm not sure how often I'll need to re-balance. Once a year? The max discharge rating of the 200ah cells is 3C which should be fine, but I need to be comfortable that I understand the concepts and caveats, ergo my question. The application is for an auxiliary boat motor. \$\endgroup\$ – MorningCloud Oct 29 '18 at 20:09
  • \$\begingroup\$ Interesting application! I can't say from personal experience how often you will need to re-balance your cells if you're using a bottom-balanced setup. Every time I've done a project involving batteries I've opted for either passive or active balancing, and if I were in your position I think I'd have at least a passive balancing system in place. Here are some links that may prove informative: blog.epectec.com/… batteryuniversity.com/learn/article/… \$\endgroup\$ – Platytude Oct 29 '18 at 20:21
  • \$\begingroup\$ So let's assume all cells have been balanced to the same voltage level, say, 3.5 volts. In a 16S0P (is that the correct designation?) the bank would be at 56 volts (within acceptable power range of the 48v motor). Since each cell has a nominal voltage of 3.2v and now stands at 3.5v each, does that mean I've got more slightly more capacity in the bank? In other words, is the rated capacity of a cell given for the nominal voltage? If that nominal voltage is exceeded slightly there is also a slight increase in capacity? \$\endgroup\$ – MorningCloud Oct 30 '18 at 14:56
  • \$\begingroup\$ Cell capacity is rated using current, not voltage, so if you're calculating total capacity based on nominal voltage, you'll get a rough estimate at best. Look at this link: batteryuniversity.com/learn/article/… You'll see that the voltage vs. amp-hour curve is very flat for the most part, meaning that once the battery reaches about 3.3V, it's already at nearly 100% capacity. Sure, if you charge up to 3.5 or 3.7V, you'll gain some extra capacity, but it will be at diminishing returns, (ie. you might only get a couple extra % of capacity out of it). \$\endgroup\$ – Platytude Oct 30 '18 at 19:08
-1
\$\begingroup\$
  1. Yes, if you're drawing 2C then you need 2C cells.
  2. Instead of talking of C ratings, let's just say you have a cell design available that's capable of I_cell. If the motor draws 2 * I_cell, then you need a series-parallel arrangement. With LiFe cells it's better to parallel the cells first, and then make a series string of the result, but it can be done the other way.
\$\endgroup\$
  • \$\begingroup\$ Not an answer, per se., since you didn't ask this question -- 1C (or even 2C) is a much lower discharge capability than I'm used to seeing. I fly model airplanes, and it's not uncommon to see cells advertised for a 40C discharge rate, with 10C being about the minimum. While I'm not totally astonished at this (there's tradeoffs in cells that have that high of a discharge), it makes me think you may want to look around for alternatives. \$\endgroup\$ – TimWescott Oct 29 '18 at 19:34
  • \$\begingroup\$ Parallel first... hadn't thought of that but it makes sense. It is likely I'll just have a 16 cells in series to get 48v, but that isn't a hard decision yet. \$\endgroup\$ – MorningCloud Oct 29 '18 at 20:11
  • \$\begingroup\$ It's actually better to series the cells first, especially if you have a BMS. Even when you parallel the same model cell they will have slightly different characteristics, so the more you put in parallel with a series BMS the less accurately you control the series string. For a good example, look at how Tesla or the other auto manufacturers build battery packs. \$\endgroup\$ – Jack Creasey Oct 29 '18 at 23:29
  • \$\begingroup\$ @JackCreasey: I shall keep that in mind. In all of the work I've done directly with these things the answer to needing more current has always been "don't mess around, just get bigger cells" -- so my experience has been peripheral. Which, come to think of it, applies if you're working on something with 200mAh 1C cells. \$\endgroup\$ – TimWescott Oct 29 '18 at 23:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.