FilipK wrote: ↑Feb 23 2020 3:31am
It's the last bit that worries me - by taking just the cell volume (ignoring any spacing between them), the 72V 100Ah pack like the one I am hoping for would be 29 litres (volume) for the EIG Cells, but only half that for 18650 (15 litres) and if I went for 21700 cells then this drops to less than 10 litres. Spaces comes at a premium because I have to contain everything underneath the bodywork.
it does seem large, but since it is including the cell protection casings and the mounting hardware / busbars / etc., it's not really that big.
if you use other (cylindrical) cells, you're going to need a case around them for stiffening to keep spotwelds from breaking, and to protect against shorts across the cell faces, etc. you have to allow for padding or insulation between the main and balance / sense wires and the cells *and* the casing, so they don't get pinched and short or break, as well. then space for whatever bms you use, and allow for cooling of the bms, so space for airflow too. (you can also use a bms on the eig pack, but it's not really optional on 18650/etc packs because there is more cell variation than with the large-cell ev packs).
the one advantage to cylindrical cells is that since each cell contains it's own compression mechanism, they can be distributed in groups wherever they will fit in a vehicle or bike. a large-format-cell pack is harder (or impossible) to do that with, since the cells are supposed to be compressed (though i don't do this with mine and don't have cell problems because of it, it's supposed to be done and should increase capability/lifespan when it is--some cells can actually be destroyed by not doing it; they puff up and fail).
there are other easy-to-assemble large-format cells, too, including used nissan leaf cell modules out of totalled cars, etc. i don't have direct experience with any other than the eig cells, though.
Why am I not seeing more 18650 packs in here
in where?
on es in general, id' say almost all of the packs used are 18650 of one type or another, though some are generic cheap pouch cells (including rc lipo). only a few are ev-grade large-format cells of whatever type.
- they seem fairly cost effective, do they require cooling to work properly?
cost of cells themselves may be higher than you're expecting, from sellers that actually test their cells and that sell them in relatively matched sets. most of the cheap places may not even sell you stuff that's from the same batch of cells (and may even just have random recycled garbage cells that have been rewrapped as "better" cells).
like any other cell type, it depends on which cell you use and what it is capable of relative to your demands on it.
for instance, i have an 18650 pack here that is "rated" for something like 30a or more, but it's only 4p, so each cell has to handle 7.5a+ at those currents. i forget which cell model is in it, but if it's used anywhere near that rate, the pack gets hot--much hotter in the center than the outside, and the outside, even in cold weather, is hot to the touch. it also unbalances very quickly used at those rates. iirc, the cells are actually being used within their "spec sheet limits", but this is still very hard on them.
the eig cells don't even get noticeably warm at currents near "spec sheet limits", most likely those limits are very conservative compared to the 18650 spec sheets' limits. (meaning, the 18650 makers are specing them for the worst-case usage, while the others are spec'd for the average usage). i'm not using htem near those limits very long, so it's possible that if i was, they'd get warmer than they do--but the 18650 pack i have heats up very very rapidly.
there are several people here on es that do battery testing, and there's dedicated threads for a number of 18650 cell types if you want to poke around to see what they can actually do.
as far as cost effective....are you including the costs of the interconnect materials, of the spare cells in case you get duds, of the significant extra labor you'll spend building them vs large-format cells, the spot welder and power source for that, time and materials spent learning how to use it and make good welds every time without damaging your cells, etc?
(or the cost of the no-weld interconnect system, and time spent configuring and assembling that, and troubleshooting and testing all the connections for equal pressure and resistance, etc)
and including the time spent on testing each cell to ensure all the ones you use are well-matched, or at least that the mismatches are distributed among the groups evenly, and the cost of the test equipment to do this?
I'd venture it takes me, slowly, around an hour, maybe two with any complications, to build a 14s2p eig cell pack. (more if i were to build an outer casing for it, depending on what i started with; i used a 50-cal ammocan for the 14s1p version, which took about 20 minutes to prep and install the pack into). if i had to cut busbars for it myself id' probably have to add a few hours for that, but mine came with them (and they can be designed and ordered from anywhere that does laser or waterjet cutting easily enough).
I'd imagine a good BMS plus a smart charger and the issues with balancing that you mention are not a concern.
that depends.
many of the 18650 cells, if used at the rates we use on evs, especially if used for their entire capacity range, become mismatched in characteristics much more rapidly than the large-format ev cells. so they then *require* balancing, and longer and longer balancing times, as they age (while the large-format ev cells may not do so for many more cycles).
if cells require balancing, it is primarily because they're losing capacity at different rates, and the rate of change increases over time--more rapidly with some types of cells than others. this means you get less capacity from the pack as a whole, because once any cell in the pack is lower than others, it'll hit bottom sooner--and as soon as that cell hits bottom, the entire pack discharge must be stopped, so the capacity of that lowest cell (or the group it is in) *is* your pack capacity. meaning, if there's one cell that has only 80% of the capacity of the rest of them, then in a 1p pack that would mean you can only use 80% of your entire battery capacity (or else you damage that cell even more by discharging it further and further below it's actual capacity, and eventually make it a fire risk). (the math gets more complex the more parallel cells, but also the more likely that you will have multiple low-capacity cells, or broken interconnects within a group that actually take cells completely out of the pack, lowering that group's capacity by an entire cell (or more).
if cells don't require balancing, it's becuase they're all the same capacity, so you're not "leaving it on the table" for the cells that don't have a problem, which is what happens in the above situation.
if your pack is large enough that you don't use its' entire capacity anyway, and say use only the middle 60-80% of it, locking the rest out by setting up a bms to only charge to 80-90% of full, and to shutoff discharge when it's still got 10-20% left, and also don't use the pack near it's current limits, the cells (of any type) will last longer (probably by quite a lot).
but then you're carrying around significant "extra" weight and volume....
that's how some of the evs get by with using 18650 and other small-format cells, and generally also what they do with large-format cells. they can guarantee a certain capacity and pack lifespan by locking out some of the capacity and capability, using a pack that's capable of much more means that even when it's old it'll still be able to do *at least* what they said it could. (sometimes by "unlocking" some of that previously untapped ability, but mostly by simply not using them very hard, so they don't age as fast as they would if used at or near their limits).
anyway....there are advantages and disadvantages to every way of making battery packs. if there were one with no disadvantages, then most packs would be made that way.
