Strapping/bundling LFP batteries?

[tanglewood]

I'm interested in somehow bundling together several batteries into a single physical unit. In particular, I'm looking at four CALB 180Ah LFP cells that will be wired in parallel, and pre-wired with one or more temp sensors, and a voltage sense point. All that will come to a single connector for wiring to the BMS, but allowing for handling, installing, and removing the battery block as a single plug-in unit. There is no electrical magic in it - I'm just exploring construction/packaging approaches.


One thought was to make a shallow tray to place the cells in, then use poly banding to strap it all together and secure it as a single unit. I could even incorporate some sort of handle to make it easier to pick them up and move them, much like the cord handles on LA batteries.


I have seen the sandwich packaging use on other Li chemistries, but I don't think support like that is required for LFP. So I'm hoping for something simpler, but maybe that's as simple as it gets.


I would welcome any ideas, or pointers to approaches others have taken. I'm guessing there are no manufactured trays or boxes for these cells, or perhaps I haven't looked hard enough.


Thanks in advance

[john61ct]

Standard practice is rigid aluminum plate at the ends, and strong rods+bolts or straps holding the pack in compression.

Without that the sides may bulge later on, stressing the rigid intra connections or the adjacent components, container integrity etc.

[nebster]

I don't have any great insight for you, but I have a few related observations and opinions:

1. LFP cells now come in plastic-shell (like your 180Ah CALBs) and aluminum-shell. The latter are no longer pressed together at all, but rather are held at the edges via some kind of insulated plastic "frame" material, and then typically dropped into an enclosure custom-sized. These two cell types are pretty different in how they need support, and I predict the latter will become more common over time (because they are substantially denser and also because they almost certainly cool off better).

2. Given that there is a whole class of LFP cells that use no compression support at all, my theory is that the compression recommendation of yesteryear is basically outdated advice. If someone is at the point where compression has a chance at stopping a bulge and therefore holds the electrodes in a mechnically more stable orientation during a stress event, they've probably already messed up electrically by allowing the cells to get out of range in the first place. At that point you have questionable cells that you probably don't want to count on in the future.

3. Regarding your actual question... especially if you wish to be able to pick up the block of cells by its enclosure, I would go to great lengths to make sure that there is no mechanical loading of the terminals and bus bars*with whatever casing you concoct. It's not clear to me that you could do that without a pretty beefy design. Indeed, what you end up building might look similar to the compression-style enclosures that I just finished saying I don't think are necessary.

My preference, space permitting, would be to just put them all in a single, large, simple enclosure that fits them very snugly, and not plan on being able to pull them in and out in blocks. It takes up less space that way and may make the wiring cleaner, too. You can still disconnect them if you need to access the cells individually. (But I bet you won't ever need to!)

[Montanan]

Originally Posted by nebster

I don't have any great insight for you, but I have a few related observations and opinions:


2. Given that there is a whole class of LFP cells that use no compression support at all, my theory is that the compression recommendation of yesteryear is basically outdated advice. If someone is at the point where compression has a chance at stopping a bulge and therefore holds the electrodes in a mechnically more stable orientation during a stress event, they've probably already messed up electrically by allowing the cells to get out of range in the first place. At that point you have questionable cells that you probably don't want to count on in the future.

Bulging batteries means an inappropriate buildup of gas inside. The gases are produced due to electrochemical oxidation of the electrolyte. Such oxidation occurs usually due to overcharging of the battery due to a faulty battery, or faulty charging electronics. Any swollen batteries should be replaced and the cause of such discerned and resolved so as to not happen yet again.

[nebster]

Originally Posted by Montanan

Bulging batteries means an inappropriate buildup of gas inside. The gases are produced due to electrochemical oxidation of the electrolyte. Such oxidation occurs usually due to overcharging of the battery due to a faulty battery, or faulty charging electronics. Any swollen batteries should be replaced and the cause of such discerned and resolved so as to not happen yet again.

I completely agree. I haven't seen any evidence of prismatic LFP cells swelling under normal, modern charge management.

Oops, correction: looks like the cobalt chemistries do exhibit something like 0.1% dimensional swelling at sub-C rates and normal temperatures, see http://www-personal.umich.edu/~siege...p_Swelling.pdf. That probably means the iron phosphate chemistries do, too. There is enough room inside the prismatics to accommodate this, for sure. I don't think we can see that from the outside of the box.

When placed under abnormal conditions, of course, the chemistry can produce lots of gas then a lot of swelling. It may be possible to continue to use a swollen cell for a while, but it will be structurally compromised inside and more likely to spontaneously fail. And perhaps in an exciting way.

[john61ct]

My understanding is that with many even top quality makers, case swelling occurs in normal usage, not just as a result of abuse or accidents.

I really like the idea of 4S / 12V packs easily transported as needed.

Depending on your cell count vs AH capacity, and target voltage, wiring the bank that way may not be optimal electrically, affect longevity a bit or require closer monitoring.

But for some of my use cases, I think getting greater usage each year out of my investment overrides those potential issues.

Of course that affects the design of their containing box, besides robust strength protecting the cells in transit, I can't give up much space while they're in use either.

A design based on aluminum plates seems like a good way forward.

Removable insulation also becomes an issue for freezing conditions.

[tanglewood]

Thanks for the suggestions so far.


There is nothing in the CALB manual about compression support, so I take it as a yesteryear requirement too.


My current bank is side-by-side cells. It's dense, but all the voltage and temp sense wiring is kind of a mess. And it's actually quite difficult to get a grip on a single cell and lift it out.


A big part of my motivation is to clean up the wiring, which really means compartmentalizing it more. I'm also concerned about access to individual cells if something needs to be replaced. With a bunch of cells stacked on a shelf, it will be hard to access cells at the back, and much riskier bolting and unbolting straps. Separating and removing 7 cells to access the 8th at the back would be difficult and prone to unauthorized sparks and welding.


The idea of a block of cells with fewer connection points (pos, neg, and sensor plug) seems much more maintainable. I could pull a block, or maybe two for access, and then break down the problem block on the bench in a more controlled environment.


Plus I need to be able to secure everything for rough seas, and a secure block, than can then be secured itself, as opposed to securing individual cells, seems like it might be easier to manage. The bank I have now is stationary so none of this is a concern, but it will be as I replicate it for my boat.

[nebster]

Originally Posted by tanglewood

Plus I need to be able to secure everything for rough seas, and a secure block, than can then be secured itself, as opposed to securing individual cells, seems like it might be easier to manage. The bank I have now is stationary so none of this is a concern, but it will be as I replicate it for my boat.

I see. I think I'm remembering the photos of your practice bank, rather than the one you are designing. In it, it seemed like you had great access to the cells.

If you have a tight space, I can see it being worth modularizing things a bit.

[tanglewood]

Originally Posted by nebster

I see. I think I'm remembering the photos of your practice bank, rather than the one you are designing. In it, it seemed like you had great access to the cells.

If you have a tight space, I can see it being worth modularizing things a bit.

That's right, my "practice" bank, which will power my house, by the way, is a top loaded battery box with good access. And I have the cells stacked immediately adjacent to each other, but not bound together in any way.


Over the past 24hrs I have been exploring ways to do the same thing on the boat, creating a large foot print, top loaded battery box rather than stacking them on shelves in a smaller foot print. It's looking promising, and would allow me to access any cell just as easily as another.

[nebster]

Originally Posted by tanglewood

That's right, my "practice" bank, which will power my house, by the way, is a top loaded battery box with good access. And I have the cells stacked immediately adjacent to each other, but not bound together in any way.

Oh, just your house, eh? Hehe.

[Checkswrecks]

Originally Posted by john61ct

My understanding is that with many even top quality makers, case swelling occurs in normal usage, not just as a result of abuse or accidents.
...
Of course that affects the design of their containing box, besides robust strength protecting the cells in transit, I can't give up much space while they're in use either.

Originally Posted by tanglewood

Thanks for the suggestions so far.
There is nothing in the CALB manual about compression support, so I take it as a yesteryear requirement too.
...
My current bank is side-by-side cells. It's dense, but all the voltage and temp sense wiring is kind of a mess. And it's actually quite difficult to get a grip on a single cell and lift it out.

The CALB user manual is super-minimal in what they say, so take it as a bare minimum. https://www.google.com/search?q=180A...=1818&bih=1122

Been looking at a lot of failed Li-ion batteries of various chemistries for a number of years and can personally tell you that lithium cells absolutely will swell and move during the charge and discharge cycles. It's pretty amazing to make a video and then watch it sped-up. Underwriters Labs (UL) did a lot of testing of this with the 787 batteries and with newer smaller cells and has seen it repeatedly. You do not want to stack the cells hard against each other, especially if flat-sided.

Another benefit of having at least a couple/few mm between cells is the diminished heat transfer, both when there are high discharges and in case one has a failure. High discharges create heat and it is common with starting the engine or use of big lights. With LiFePo the problem is not that the battery will fail right away, it is that the cooler cells on the ends will age very differently so the pack will go unbalanced. Creating even heat distribution is the reason the e-car manufacturers have coolant, not to actually cool the battery. It's one reason some of the CALB batteries have ridges on the outside of the casing.

A common cheap and easy way to get the spacing for flat or prismatic cells is to simply insert phenolic, plastic, or aluminum plates between them, cutting out the middle of each rectangular spacer. It's a DIY version of the frame that you mentioned. This way the cells support each other at the edges where stiffest and the middles never touch.

When the cells swell and move, bussing the connectors by solid bars puts a severe strain on the seal at the base of the connector, leading to leakage of flammable electrolyte and really fast cell failure. Use bent or woven cables to connect to the cells. Connecting these directly to the next cell would not be good:


I know the upper clips are what CALB provides, but use a higher arch or go to woven to get a lot more strain relief:


Monitor the temps at the hottest locations for the individual cells, not just one or two places in the pack. By the time a remote thermocouple detects heat any but an adjacent cell, the damage has spread. With all monitored so that hitting a threshold disconnects the battery, you should have a chance to save all but one cell. Note that even the minimal CALB instruction calls for monitoring of both individual cell T and V.

Another downside to the CALB cells I've encountered is they are not aged or weathered by the factory. Plan to disconnect the bus bars, check voltages, and if needed re-stack the cells periodically. After the initial month or so (hard to say with not knowing your use) Nebster checking it quarterly sounds right.

Or so I've heard.

[KTP]

So four CALB 100AH cells are $129 each, or $516

Straps, case, BMS, maybe add another $200?

Trojan Trillium 92AH drop in battery is $850, IP67 rated, numerous over charging and discharging, temperature, short circuit protections, canbus communications for detailed stats on battery.

I mean wouldn't it be now like trying to make your own diesel engine from steel tubing and soda cans instead of buying a Yanmar for the same price?

[tanglewood]

Originally Posted by Checkswrecks

The CALB user manual is super-minimal in what they say, so take it as a bare minimum. https://www.google.com/search?q=180A...=1818&bih=1122

Been looking at a lot of failed Li-ion batteries of various chemistries for a number of years and can personally tell you that lithium cells absolutely will swell and move during the charge and discharge cycles. It's pretty amazing to make a video and then watch it sped-up. Underwriters Labs (UL) did a lot of testing of this with the 787 batteries and with newer smaller cells and has seen it repeatedly. You do not want to stack the cells hard against each other, especially if flat-sided.

Another benefit of having at least a couple/few mm between cells is the diminished heat transfer, both when there are high discharges and in case one has a failure. High discharges create heat and it is common with starting the engine or use of big lights. With LiFePo the problem is not that the battery will fail right away, it is that the cooler cells on the ends will age very differently so the pack will go unbalanced. Creating even heat distribution is the reason the e-car manufacturers have coolant, not to actually cool the battery. It's one reason some of the CALB batteries have ridges on the outside of the casing.

A common cheap and easy way to get the spacing for flat or prismatic cells is to simply insert phenolic, plastic, or aluminum plates between them, cutting out the middle of each rectangular spacer. It's a DIY version of the frame that you mentioned. This way the cells support each other at the edges where stiffest and the middles never touch.

When the cells swell and move, bussing the connectors by solid bars puts a severe strain on the seal at the base of the connector, leading to leakage of flammable electrolyte and really fast cell failure. Use bent or woven cables to connect to the cells. Connecting these directly to the next cell would not be good:


I know the upper clips are what CALB provides, but use a higher arch or go to woven to get a lot more strain relief:


Monitor the temps at the hottest locations for the individual cells, not just one or two places in the pack. By the time a remote thermocouple detects heat any but an adjacent cell, the damage has spread. With all monitored so that hitting a threshold disconnects the battery, you should have a chance to save all but one cell. Note that even the minimal CALB instruction calls for monitoring of both individual cell T and V.

Another downside to the CALB cells I've encountered is they are not aged or weathered by the factory. Plan to disconnect the bus bars, check voltages, and if needed re-stack the cells periodically. After the initial month or so (hard to say with not knowing your use) Nebster checking it quarterly sounds right.

Or so I've heard.

Thanks. Any source suggestions for those woven battery straps? I'm currently using the CALB copper straps (top in your picture). They are stacked CU sheets, not solid CU, so flex pretty well. What I don't like is that they force adjacent spacing of the cells with no gap. A source fro longer straps would give me more flexibility (no pub intended).


I appreciate the heating and cooling issues, but the feedback I'm getting is that operating in the .5C and lower range creates negligible heating. So I haven't been too worried about it, but better cooling would be, well, better.

[nebster]

Originally Posted by tanglewood

I appreciate the heating and cooling issues, but the feedback I'm getting is that operating in the .5C and lower range creates negligible heating. So I haven't been too worried about it, but better cooling would be, well, better.

I hope I haven't been one to contribute to that understanding. I have a data point for you:

I took a string of 16 100Ah aluminum skin cells, arranged in an 8x2 array that ends up occupying almost exactly one cubic foot, with the 3/32" spacing between each imposed by the dividers, in a large open room stable at roughly 75F, sitting idle for several days and so also at 75F. This string started at 37% SOC and charged at 0.3C (~1600W) for 65 minutes. At the end of the charge, the SOC was 69%.

At this point, the batteries themselves measured roughly 89F on the outside edges and 91F in the center of the cube. The terminals were at ~105F.

All of that heating does not include very much of the additional ambient rise from the ohmic losses in any wiring and fuses that might apply in a small space. So I think, depending on the enclosure's insulation and venting properties, heating should definitely be a consideration. (Maybe not for your situation specifically, but just speaking generally.)

[nebster]

Originally Posted by john61ct

My understanding is that with many even top quality makers, case swelling occurs in normal usage, not just as a result of abuse or accidents.

Oh? Can you share a reference?