diy solar

diy solar

This is my first Diy battery rack. Any advice on how to improve it or important things i may have missed?

Factory specs are compression against entire face, but you could possibly add a taller shim (of steel that's sufficient to not deflect) in the ends to get there?
I am going to have to think how to fix it. Each compression plate takes 20 hours to 3d print so is 120 hours of printing time i wouldnt like to repeat
 
I am going to have to think how to fix it. Each compression plate takes 20 hours to 3d print so is 120 hours of printing time i wouldnt like to repeat
So you can use the existing end plates just need a steel plate between the end batteries and the end plate so it covers that small lip on the top.
 
Just to be a party pooper, for about $110 you can purchase an all steel shelving rack with 4 shelves that can be all stacked or just 2 shelves x two. Plus they are rated to each hold over 1,000 pounds. That's a good starting point for placing batteries you are cobbling together.

Question: should one of your batteries decide to no longer enjoy working for you and plan a run away lithium fire, it doesn't appear that you have any containment around these batteries to prevent spread.

You may decide to use any existing waterline plumbing you have access to and set up a sprinkler system for emergency fire suppression. And a fire and smoke detector with central alarm monitoring so living creatures near your project do not become victims in their sleep.
 
Just to be a party pooper, for about $110 you can purchase an all steel shelving rack with 4 shelves that can be all stacked or just 2 shelves x two. Plus they are rated to each hold over 1,000 pounds. That's a good starting point for placing batteries you are cobbling together.

Question: should one of your batteries decide to no longer enjoy working for you and plan a run away lithium fire, it doesn't appear that you have any containment around these batteries to prevent spread.

You may decide to use any existing waterline plumbing you have access to and set up a sprinkler system for emergency fire suppression. And a fire and smoke detector with central alarm monitoring so living creatures near your project do not become victims in their sleep.
I have tons of steel profiles in my warehouse and i am used to just chop some and weld them together when i need something like this.
You are right about the fire risk issue. But i dont know enough yet about lithium fire risks and ways to contain it. I was thinking about some type of enclosure, i am going to have to study this as well. Thanks for your feeback.
 
Lets see.
48 cells went about 5.5k eur
3 jk bms around 300 eur
Dc breakers 240 eur
T fuses 180 eur
The 12 beefy springs costed me 100 bucks in total
Around 50 bucks of pla
And maybe 250 eur of other stuff (threaded rods, steel bars, screws, etc)

So in total about 6.6k, and for 48kwh that is 137.5 eur, lets call it 140 eur per kwh
Thank you for putting this together. Looks like the price is almost the same (around $150kwh after exchange rate conversion) with the built in floor standing batteries delivered to your door in US with warranty and support. See below for more details -

 
So you can use the existing end plates just need a steel plate between the end batteries and the end plate so it covers that small lip on the top.
The plates are 3d printed black pla with spray painted steel bars to reinforce it. I am thinking about just chopping the pla in half horizontally, add a 3d printed strip that matches in the middle to make up what is missing to cover the cell completely, and just put longer yellow painted steel bars on it. That should work i believe. Thanks!
 
I have tons of steel profiles in my warehouse and i am used to just chop some and weld them together when i need something like this.
You are right about the fire risk issue. But i dont know enough yet about lithium fire risks and ways to contain it. I was thinking about some type of enclosure, i am going to have to study this as well. Thanks for your feeback.
In California, the state has embarked along with the utility owners and investors to build massive lithium battery storage facilities. the state has an excess of daytime PV energy production, but without battery storage, needs to burn gas to generate electricity, ergo the battery storage facilities.

So I undertook a googling one day and it seems that the best way to put out lithium battery fires is containment cabinets, separating batteries from each other AND flooding them with water. NFPA 855 addresses lithium batteries and talks about residence locations. Stacks limited to 20 kW, 20 kW stacks separated by 3 feet from each other and walls, no locations in living areas, near windows or doors or in closets. If in garages, the walls should not be open studs.

The UL9540a is a battery testing standard for run away fire containment of a battery and UL9540 is a standard for communication between the battery BMS and inverter to shut down charging currents to prevent fires. When the inverter and battery bank together meet the standard, the local authority will allow a larger bank of batteries than NFPA855.

In your case, unless you can place the banks into some kind of containment cabinet, the next best thing is going to be to install a fire suppression water based system. Many are available, especially if you look for them for California due to wildfire risk here.

In my case, my batteries and inverter meet UL9540a and UL9540 as well as the batteries are in metal cabinets which are UL lister for the batteries and fire suppression and indoor and outdoor rated. Locked to prevent tampering too and ventilated, with optional heating and cooling. All designed to prevent fires.

Total cost for 41 kW of batteries and cabinets $13.5K before Federal tax credit. And the 10 year over 6000 cycle warranty.

You may be saving some money and of course, I like you love to do projects. But some things for me have a line like DIY batteries in a high wildfire risk zone and putting pilings into a river bed to build my own dock.

You may have done a great job, but you have little to no protection from an actual cell that fails in terms of containment. Your only real means of protection is visual cell inspection daily and then shut that cell down and remove from the pack. That won't be fun.

Finally, don't take all this as personal. It is really written for other forum members to gain an understanding of the advantages and problems faced if doing a DIY and think of the potential for property and life loss in case of failure.
 
Finally, don't take all this as personal. It is really written for other forum members to gain an understanding of the advantages and problems faced if doing a DIY and think of the potential for property and life loss in case of failure.
Oh not at all. I greatly appreciate feedback from people that knows more than me. Without you all i couldnt have built this.
The lithium fire risk is kind of scary. Since i see commercial lithium battery racks all over i didnt gave it a lot of thought, but chinesium is a dangerous chemical element as they say :ROFLMAO:
Thank you for bringing awareness to this issue and for your detailed explanation.
 
Oh and the end caps, don't seem to cover the whole face of the cell, the top edge is uncompressed.
That's fine. Cell jellyroll does not extend into that area. Most expansion happens in middle side. I would insert 1/2" (12.7mm) thick aluminum or 1/4" (6.4mm) steel plate on each side of cell stack in between compression corner clamp and the cell to spread that compression force toward cell middle. Springs can be stretched to accommodate 28mm worst case increase in total cell stack width (with insulator or rubber spacer material between metal plate and cell surface).

clamp.jpg
 
That's fine. Cell jellyroll does not extend into that area. Most expansion happens in middle side. I would insert 1/2" (12.7mm) thick aluminum or 1/4" (6.4mm) steel plate on each side of cell stack in between compression corner clamp and the cell to spread that compression force toward cell middle. Springs can be stretched to accommodate 28mm worst case increase in total cell stack width (with insulator or rubber spacer material between metal plate and cell surface).
The idea was to 3d print a template that can hold in place painted steel bars that i have in stock (instead of laser cutting a whole plate of steel or aluminum).
See the image below. They distribute the pressure very well. But i cant print something bigger than 200mm so i couldnt cover the 204mm cell. I hope that is ok.
pressplate.jpg
 
Those yellow plates look thin and I don't think they will be rigid enough to transfer force to the middle. You could add another plate horizontally between plastic and vertical plates to act as spacer and to slightly bend yellow vertical plates to make them act as springs. Red arrows show force directions. Edit: This may not be a good idea because concentrating total compression force on small center area could over compress that part of cell jellyroll and damage the cell. Compression needs to be uniform across entire cell jellyroll.
 
Last edited:
Or 3D print a hump on that PLA template in the middle to force yellow plates into spring mode to transfer force to the middle. You could also bend them slightly into U shape.
 
Or 3D print a hump on that PLA template in the middle to force yellow plates into spring mode to transfer force to the middle.
I see. I can add a small square of some thin rubbery material between the cell and pla template so i increase the pressure at the center. That will be quick and easy.
Thanks for your feedback.
 
Nice build.
I think that you have done a great well planned project.
I don't see any issues that are major enough to mention.
Since you already plan to protect from inquisitive fingers.
 
Have a read through this thread - it is super long of course with oodles of speculation based of little information. Somewhere around the middle someone posted an automated extenguisher system that looked interesting. With LFP there is no need to drown them like Lipo

And this thread
We concluded that his shelves were sagging because they were inadequate. Your won't have that problem.

They may give your some ideas of things to add for safety

One thing I would do is add class T or NH1 fuses as close as possible to the battery - looks like there is just enough room above the breakers you are using. I say this because breakers can be slow to trip and thermal sensitive. The second link didn't have any breaker trips if I remember correctly.

For containment - short of an external building - A layer of firerock and/or sufficient durok. Firerock is gypsum board that has been treated to have a 4 hour burn through time. Durock is cement board. When I build I intend to line the battery room with these.
 
Have a read through this thread - it is super long of course with oodles of speculation based of little information. Somewhere around the middle someone posted an automated extenguisher system that looked interesting. With LFP there is no need to drown them like Lipo

And this thread
We concluded that his shelves were sagging because they were inadequate. Your won't have that problem.

They may give your some ideas of things to add for safety

One thing I would do is add class T or NH1 fuses as close as possible to the battery - looks like there is just enough room above the breakers you are using. I say this because breakers can be slow to trip and thermal sensitive. The second link didn't have any breaker trips if I remember correctly.

For containment - short of an external building - A layer of firerock and/or sufficient durok. Firerock is gypsum board that has been treated to have a 4 hour burn through time. Durock is cement board. When I build I intend to line the battery room with these.
I have put class T fuses directly on the dc breakers of each battery, they are small so is difficult to see them. I am afraid the wire will come out loose if they pop so i have to fix that i guess.
The Lifepo4 fire i dont understand yet. Is a reaction with air or a runaway reaction of the chemicals within the battery itself?
If is a reaction with oxigen in the air a small enclosure that will starve the oxigen could be an option.
 
I have put class T fuses directly on the dc breakers of each battery, they are small so is difficult to see them. I am afraid the wire will come out loose if they pop so i have to fix that i guess.
The Lifepo4 fire i dont understand yet. Is a reaction with air or a runaway reaction of the chemicals within the battery itself?
If is a reaction with oxigen in the air a small enclosure that will starve the oxigen could be an option.

LFP batteries offgas hydrogen when they vent. No spark no fire. They do produce a small amount of oxygen as the electrolytes break down, but the amout is tiny compared to the oxygen in air.

In the first I listed I think if the correct fuse was used or it had been contained long enough for the vents to burn out the house would have survived.

In the second it was in a detached garage and dispite catastrophic damage to the battery banks the structure is ok and the OP thinks he can salvage most of the cells in the other banks. I am not sure of that since the extinguishing agent fire departments use for this is basically a wet sodium cloride slurry. Not toxic, but very corrosive. It works by smothering the fire.

The extinguisher mentioned in the top one was a potasium and something mix that sucked the heat out of the surrounding area and smothers it.

Incidentally the 1 ground rule means your main panel in the houe uses ground from the power shed.
 
I see. I can add a small square of some thin rubbery material between the cell and pla template so i increase the pressure at the center. That will be quick and easy.
Thanks for your feedback.
Thinking more about this. It may not be a good idea because concentrating total compression force on small center area could over compress that part of cell jellyroll and damage the cell. Compression needs to be uniform across entire cell jellyroll. To avoid modifying your 3D printed end plate fixture you could replace those yellow metal plates with plates of same height and width but 3x - 4x thicker to resist bending under load and to provide even compression force distribution across cell side surface.
 
Back
Top