DIY battery box best practices (true DIY not bought box)

[Hedges]

Which is why my DIY chest freezer battery box is destined for the retaining wall, away from the house.
Same place my UL listed LG RESU-10H is located (a model with known history.)

Dead Wrong or Dead Right?

Or follow my lead and use lead batteries. They're safer.

 

[Aridom82]

My future DIY battery box. Plan to put 60 kWh in it, BMS outside the climate-controlled environment.
This will go outside, so roof with drip lip to keep water out.
Refrigeration unit keeps it cool, battery heaters keep it warm.

View attachment 317762

(file photo)

Hmm inside one of those big refrigerators. That is interesting. Do you think the bottom can hold the weight?

 

[Hedges]

I think bottom has foam, weight would be spread across that and should be OK.
The step that covers compressor might need help.
I will need to have supports between layers, forget if steel cans around my 4s modules have flange at top that can transfer weight.
Lack of access will be inconvenient.

I'd rather have door on side and pull-out shelves for 200 lbs. each 48V battery.

Five of them (60kW) to fit in this chest freezer, with a few inches foam extension to lid on top.
My old upright fridge may have more volume but didn't really measure to hold more of my size modules.

Ideal would be a rack and insulation around outside. I don't find slides to hold that much weight.

I'm considering smaller chest freezer I've got for my sister's 24kWh 16s2p battery.

I got an AC heat/cold electronic thermostat to regulate temperature.
Will replace (failed) freezer thermostat with refrigerator temperature as backup cold limit, and stick on snap-action switches as backup hot limit.

 

[Zwy]

I disagree with the lack of a spring... at least with the larger 314ah cell... I haven't read the datasheet of the 280ah cells or smaller... they do list the pressure and compression amounts in the datasheet. But the thing is as the cells expand and press outwards the amount of tension will increase with a static tension like you are doing. With springs that start out soft and grow stiffer as they compress ... so when the cells are at minimum expansion there isnt as much force to counter it... as the cells expand the springs press back harder....

As you read the datasheet, did it mention springs?

Now we aren't talking about much movement at all.. Just 1mm at most per cell and the poron takes up the vast majority of that...but you want the poron to not be totally flattened so the spring let's things give a bit more without letting anything be loose... and if you have a static load so the cells expand and bottom out against the plates I would imagine that could cause damage due to overpressure...

I don't use springs. It's not in the datasheet fixture. If you looked at any of the drawings for fixtures on the datasheets, the cell is clamped in a fixture with bolts holding the two plates together.

What I do not get is why you say the springs will allow the centers deform over time...

The pack length will increase over time and the springs will compress. It won't be the cells compress and the springs lengthen.

And for the record either method is probably enough so that the age of the cells will be more detrimental than the difference in the two methods.....

Springs are a waste of time and not what the datasheets show.

And all of this has me getting out my old college book on statics and dynamics to check the math... yes that dusty 30 year old book has some good stuff in it still.


@Zwy what torque do you use and do you start while the SOC is 5d0% or some other value?

I don't use torque when using foam. Watch the video.

 

[Zwy]

they also require not to exceed 10,000N force (39psi) to avoid cell damage. I understand it the cell squeezed between 2 hard surfaces that won't move is capable of developing self destructing expansion force during charge.

Foam here provides another advantage- the box can be potentially made without compression rods and simply squeezed before assembly so the Poron spacers would serve as those springs.

You want the rods. If you ever have to take the pack apart after a long period of time, you will find the cells aren't easily removed. If you place a FR4 sheet in between the cell and foam on the one side, then it might be easier. I do stick the Poron onto one side of the cell to prevent any shift in foam position.

 

[Zwy]

thanks for sharing the video - enjoyed it.

Watch the series on building the shop bank. I built that after my house bank.

 

[GuyG]

Is there a best practices guide when building your own DIY battery box.

The best best practices guide is to start with this forum and then just build something. Don’t get discouraged by all the options; read this thread and then just do the thing. If it’s not perfect the first time around, do it better the next time.

I’m in the prototype-building stage of converting a small fleet of golf/utility carts. Couldn’t find an off-the-shelf product in the right form factor that met our needs at a halfway reasonable price so am building them. Using 230 Ah cells, a 300 Amp JK BMS, T-class fuses.

Welded up some 14-gauge steel boxes with 3/4” square tubing welded lengthwise into bottom for additional strength/rigidity. Am using 1/2 inch aluminum as a compression plate to secure the cells. That plate is secured by 1/4” threaded rod covered in clear tubing, and a small angle bracket into the bottom of the box. No springs, but I put 3mm Poron sheets between the cells, and epoxy sheets to line the bottom of the box. Tapped the 1/2” aluminum plate to accept #10 machine screws to mount the BMS, which stands off the plate by virtue of short pieces of 1/4 white pex. Will use hydraulic punch to knockout holes for bulkhead fittings where the conductors will exit the box. Welded a bit of angle at the corners for hold-down bolts, and cut a sheet of thick commercial kitchen rubber mat to put under the box as a vibration damper. Haven’t decided on covers yet, but will use plexiglass if I’m feeling rich and stupid when the time comes to decide. Am also thinking about building a prototype box entirely out of 1/2” aluminum, as if I don’t have enough things to keep me busy.

Don’t fret about all the options. Just pick some ideas you like and go for it. The only real requirement is to be safe.

 

[robbob2112]

I don't use torque when using foam. Watch the video.

I will watch the whole thing when I have the time -- I've watched a number of your videos and they are always great at showing what you are doing --- but most of the ones I have watched leave out or just glide over the "why" of what you are doing --- I'll watch this whole video and come back with questions about it.


That is the single cell test fixture...i.e. put the cell in that fixture, torque down to a compression force across the face 300kgf and measure thickness to see if it passes...


There is a swell force fixture on the next page of the pdf ...they don't describe it well... I am assuming this is 4 smooth rods to slide on fixed at the ends... then a screw in the middle that pushes the left center plate and then the plate against the cell -- for even distribution of pressure I assume -- I wish they described or showed the mechanism between -- I am assuming it is some kind of flat disc or bearing to allow it to spin freely and the two of them to even out the forces -- but I can't tell for sure

It goes on to say

Customer shall fully consider the influence of the cell swelling force when designing the module. The product generates expansion force during use, and the expansion force is about 60000 N when the cell capacity attenuates to 60% under thetest conditions of 15 mm steel plate + 0.0 mm GAP (the space for cells to expand). Customers shall consider there liability of structural strength in the product design process, and it is suggested to reserve 2.0 mm ~ 2.5mm expansion space while grouping the cells.

But is that suggesting reserve 2.0mm~2.5mm per cell or for any stack --- I would interpret that to mean per cell since all the cells will swell and in a string that would add up -- 16cells - 32mm -- 1.26 in -- which is taken up by the poron --

And - no springs are not mentioned -- at any point in the datasheet - only that the pressure on the cell needs to fall within a range of 3000N~7000N and never exceed 10000N ---


For both of the test fixtures they say to pre-load the plates with 300kgf of force -- Which would seem to say the cells need to be pre-loaded





And, this is from higher up in the datasheet -- -- The 300kgf is equal to 2941N -- so that tracks --





Which means for 6 rods the torque applied to zinc plated steel all-thread with zinc nuts and lubricated with nickel anti-seize would be 4.95 in-lb or for 4 rods 7.43in-lbs

 

[Tulex]

Just how many prebuilt batteries have this?

 

[Calvin98]

There are rules and regulations for solar panels and mounting systems.
Not for homemade batteries.

Definitely
Which is why a lot of us build our own.

If I had unlimited funds, I wouldn't be building my own anything.
I would pay someone else to do it.
Diy is all about saving money.

It all depends. Even if you buy top-of-the-line everything, the manufacturer building the part could go bankrupt. They could also put certain features behind pay walls. They could discontinue the item and no longer provide any sort of support. The warranty may be rather short. They could send you a firmware to disable something or everything. A particular component may no longer be made years from now. Phone support may be terrible or nonexistent. Having a technician come out to repair or replace something maybe not possible. Etc. etc.

At least if you DIY, you have control over everything you’re doing and hopefully have backups and know how to replace stuff and fix stuff if it ever breaks.

For me, I have SMA Sunny Island inverters. They come with a 10 year warranty. And the warranty doesn’t start until you actually start using the item. So you can leave it sit around for many years and the clock on the warranty doesn’t start until you turn it on. Those inverters are super robust and could easily last 20 years. Even so. I still have several as extras as spares to use in the future in addition to the four running the primary system and the other two that are installed as a back up system ready to go at a moments notice with an automatic transfer switched should anything happen to one of the first four inverters.. I also have plenty of spare 12 V CATL batteries that can be replaced if any one 12 V pack in my 48 V systems goes bad. Worst case scenario I buy four raw cells and make a custom 12 V pack. Plus extra JKBMSS if any BMS fails. And if JK stops making the BMS, it could be replaced by a new version relatively easily. Most of the other parts are just DIY and you do what you want. Should never be a problem of not being able to buy a DC breaker or a class T fuse.

 

[robbob2112]

Just how many prebuilt batteries have this?

Well, great question and the answer is none.... but that could be for economics or any number of reasons...how many vendors sold 300ah batteries with 280ah cells?

 

[Calvin98]

Just how many prebuilt batteries have this?

The 12 V CATL packs that Hedges and I have have this I think. They’re encased in a riveted enclosure. The manual doesn’t state how many how much force is used in the compression, but they are surrounded by the steel case on five sides. Everything is completely riveted together. No springs. And these were made for Ev’s. Probably busses.

 

[timselectric]

It all depends. Even if you buy top-of-the-line everything, the manufacturer building the part could go bankrupt. They could also put certain features behind pay walls. They could discontinue the item and no longer provide any sort of support. The warranty may be rather short. They could send you a firmware to disable something or everything. A particular component may no longer be made years from now. Phone support may be terrible or nonexistent. Having a technician come out to repair or replace something maybe not possible. Etc. etc.

At least if you DIY, you have control over everything you’re doing and hopefully have backups and know how to replace stuff and fix stuff if it ever breaks.

For me, I have SMA Sunny Island inverters. They come with a 10 year warranty. And the warranty doesn’t start until you actually start using the item. So you can leave it sit around for many years and the clock on the warranty doesn’t start until you turn it on. Those inverters are super robust and could easily last 20 years. Even so. I still have 9 extras as spares to use in the future in addition to the four running the primary system and the other two that are installed as a back up system ready to go at a moments notice with an automatic transfer switched should anything happen to one of the first four inverters.. I also have plenty of spare 12 V CATL batteries that can be replaced if any one 12 V pack in my 48 V systems goes bad. Worst case scenario I buy four raw cells and make a custom 12 V pack. Plus extra JKBMSS if any BMS fails. And if JK stops making the BMS, it could be replaced by a new version relatively easily. Most of the other parts are just DIY and you do what you want. Should never be a problem of not being able to buy a DC breaker or a class T fuse.

I got lost.
I'm not sure if you are agreeing with me or not.

 

[Calvin98]

I got lost.
I'm not sure if you are agreeing with me or not.

Sorry. I guess I was trying to say that even if you buy a top-of-the-line turn key system, that’s not a guarantee of anything. Even if you hire a professional licensed electrician to install things, that’s also not a guarantee that things will work, or that they will be done correctly. My solar installer took a lot of shortcuts that were all inspected and approved. He was licensed by the state and certified to do solar installations. I ended up redoing quite a bit of what he did and it is much safer. I was agreeing to a point. Probably like with solar panels and inverters, it’s still early days for batteries. The battery landscape will probably look very different 10 years from now. With DIY, you can choose the size of the cells you want, you can choose the voltage of the pack, you can choose the BMS the inverter everything.

 

[timselectric]

Sorry. I guess I was trying to say that even if you buy a top-of-the-line turn key system, that’s not a guarantee of anything. Even if you hire a professional licensed electrician to install things, that’s also not a guarantee that things will work, or that they will be done correctly. My solar installer took a lot of shortcuts that were all inspected and approved. He was licensed by the state and certified to do solar installations. I ended up redoing quite a bit of what he did and it is much safer. I was agreeing to a point. Probably like with solar panels and inverters, it’s still early days for batteries. The battery landscape will probably look very different 10 years from now. With DIY, you can choose the size of the cells you want, you can choose the voltage of the pack, you can choose the BMS the inverter everything.

Gotcha
But if my funds were unlimited, it wouldn't matter.
I would just pay someone else to replace everything.
The moral of that post is never give me unlimited funding. lol

 

[robbob2112]

Interesting readings....

Bolt Torque Calculator: Loads & Preloads Data and Calculator

Calculate required bolt torque.

www.engineeringtoolbox.com

LFP Cell Clamping Pressure - Battery Design

The LFP cell clamping pressure is important as it keeps the active layers in contact. However, we have to remember that LFP cells expand over their lifetime by 10 to 20% in thickness.

www.batterydesign.net

https://www.tapesolutions.saint-gobain.com/na/en/technical-articles/innovative-compression-pads-maximum-ev-battery-cell-performance

 

[Brucey]

Interesting readings....

Bolt Torque Calculator: Loads & Preloads Data and Calculator

Calculate required bolt torque.

www.engineeringtoolbox.com

LFP Cell Clamping Pressure - Battery Design

The LFP cell clamping pressure is important as it keeps the active layers in contact. However, we have to remember that LFP cells expand over their lifetime by 10 to 20% in thickness.

www.batterydesign.net

https://www.tapesolutions.saint-gobain.com/na/en/technical-articles/innovative-compression-pads-maximum-ev-battery-cell-performance

Yes that's the curve these numbers for the poron 4701-40 represent, if one is aiming for an ideal initial psi of 5 to 13. Compress 25% and you are good to go. From my recollection when you go to the harder -50 and -60 range theres just not a large range of safe psi like the -40 does.

 

[fatjay]

Just how many prebuilt batteries have this?

Prebuilt batteries only have to last until the warranty runs out. The compression is not so much a safety issue, but more so getting your rated cycles. If pre-builts are warrantied for 3-10 years, that's less than half the rated cycles. I would say with a degree of confidence that properly compressed packs created by diy users could reasonably last up to twice as long as pre-builts with similar components, the difference being compression vs non-compression.

 

[Hedges]

How many cycles to EV busses expect?

My 228 Ah CATL 4s module data sheet only claims something like 3000. That's the pop-riveted steel box Calvin98 posted above.
Maybe the 6000 to 8000 cycle claims rely on different (unobtainable?) compression.

I would think that spring loaded maintaining same pressure regardless of displacement would be optimum, also being most forgiving and compliant.

 

[Calvin98]

From the manual.


Still seems pretty good to me. Greater than or equal to 3500 cycles at one C/one C, 100% DOD, 25° C, 300 KGF pressure, 80% capacity retention.

So they do list 300 kg compression.

But wouldn’t spring compression preclude the use of welded bus bars?

 

[Hedges]

Right. Unless the busbars were shaped to allow enough compliance.
We see a small kink in many commercial busbars. Hard to believe that allows much compliance.
Not sure how much terminal in case can displace vs. jelly roll inside.

Preventing bubbles from separating layers could be a reason. Some here have suggested compression was only required initially.
We're probably fine with fixed steel enclosure. Possibly an expanding constant pressure assembly would give longer life.
Now that we're going to oversize banks (because they became affordable), will rarely see anywhere near 100% DoD.

 

[max2k]

Still seems pretty good to me. Greater than or equal to 3500 cycles at one C/one C, 100% DOD, 25° C, 300 KGF pressure, 80% capacity retention.

So they do list 300 kg compression.

But wouldn’t spring compression preclude the use of welded bus bars?

IMO spring compression requires ability for bottom of the cells to be moving back and forth during charge/discharge cycles which is not easy to arrange: the cell at stationary wall would need to move entire string by 1-2 mm during charge and this distance accumulates towards the cell closest to the spring. Those cells would need to move at least 8mm with every cycle for 8S string. If they don't then other cells in the string would get over compressed.

 

[max2k]

You want the rods. If you ever have to take the pack apart after a long period of time, you will find the cells aren't easily removed. If you place a FR4 sheet in between the cell and foam on the one side, then it might be easier. I do stick the Poron onto one side of the cell to prevent any shift in foam position.

I hear you, thank you for sharing the experience. I'll put the same sandwich from Poron + thin FR4 sheet between cells, it makes sense. You're right, it doesn't really matter if Poron gets completely glued to one of the cells in a pair, it would actually prevent its shifting around.

Along the same lines I intend to use flexible bus bars between cells in the string so those 1-2mm movements would not create any tension on the cell's terminals. It appears to me the cells work fine in general provided they are evenly compressed and their terminals are not under any 'shifting' load during cycle.

 

[Calvin98]

Right. Unless the busbars were shaped to allow enough compliance.
We see a small kink in many commercial busbars. Hard to believe that allows much compliance.
Not sure how much terminal in case can displace vs. jelly roll inside.

Preventing bubbles from separating layers could be a reason. Some here have suggested compression was only required initially.
We're probably fine with fixed steel enclosure. Possibly an expanding constant pressure assembly would give longer life.
Now that we're going to oversize banks (because they became affordable), will rarely see anywhere near 100% DoD.

Right. Plus the CATL cells we have do have expansion joints between the cells, even with all the compression in the box. At least for me, I’m barely doing 0.1C charge and discharge and never even got close to a 0% SOC. Maybe 20% DOD average so far. Let’s hope for long lives for these cells. A bus is probably gonna put these batteries through their paces a whole lot move than home use. All that starting acceleration at max current and stopping regeneration also at high currents, etc.

 

[robbob2112]

IMO spring compression requires ability for bottom of the cells to be moving back and forth during charge/discharge cycles which is not easy to arrange: the cell at stationary wall would need to move entire string by 1-2 mm during charge and this distance accumulates towards the cell closest to the spring. Those cells would need to move at least 8mm with every cycle for 8S string. If they don't then other cells in the string would get over compressed.

This is why I plan to sit them on the slick side of ABS plastic -- So the cells can move back and forth as needed.

As for bus bars - anything but the flexible braided type would seem to apply torque as the cells expand and contract - and also as they expand over the lifetime of the cells during usage... That was in one of the papers I listed above -- as the cells age they expand and don't fully contract -- which would seem to mean in a fixed/non-spring restraint setup the actual pressure on the cells would go up over time -- and that would mean releasing them and resetting to wider every year or 3 --- or would it? Maybe all this is just gasping at gnats and the difference is between 15 years and 15.2 years... or some small amount and it doesn't matter ---

I was looking online and for around $100 on aliexpress I can get a puck shaped load cell with display - then could build the test fixture like they have it -- the one without a good description --- It is plates held parallel with tubes that ride the rods then a load cell in the middle so it can measure the compression as things expand and contract --- There are just load cells that could be interfaced with a RPi by using an A-to-D converter board -- for long term tracking --

https://www.aliexpress.us/item/3256807890640290.html?spm=a2g0o.productlist.main.7.4a7b7438nke53N

the 300kg version would seem to be the correct range for 3000N-7000N -- Might not be perfect for the price, but for monitoring the delta from charge and discharge would be interesting ...