Help with DIY off grid battery bank upgrade to used Lithium packs

jonoslade

I am helping my parents upgrade from their current Lead Acid battery bank and need some help designing a cost-effective but safe system. They currently have about 3.5kw of solar, an Outback 80amp charge controller, a Xantrax 5500 inverter, and 8 Trojan L16s.

The new bank will consist of 10 used LG Chem 48volt 2.2kw 14s battery packs from Batteryhookup that will be mounted in a server rack and will each have a 100amp BMS.

My questions:

1. What additional safety elements i.e. fuses and or shunts (and what size/type) should I be looking to install to improve overall safety. 

2. I am thinking about using a smashed 3/4 inch copper pipe as my bus bar to connect the negative side is that feasible and safe to do? I'd prefer not to spend several hundred on a 5' 100% copper bar.


3. What else am I overlooking?

Thanks

mcgivor

Welcome to the forum @jonoslade

Can you provide details of the modules, from the 14S topography they are not LiFePo4 (LFP), do you have the specifications for the BMS used? There are numerous chemistries each having differing characteristics, advantages, disadvantages, energy density and most importantly safety, the chemistry is generally selected for the particular application. 

1. The use of over current protection, fuses or breakers depends on the BMS specifications wether they have intrinsic safety or not. Shunts are generally used for battery monitoring to determine the capacity for graphic displays or automatic generator control, in an offgrid application they tend to drift due to the dynamic changes occurring continuously, that's my observation at least. 

2. Buy flat stock copper rather than copper pipe, you are after all paying for the copper value and pipe may actually be more expensive for the same cross sectional area and will be easier to use as well as being more professional looking. A 12 foot length of 1/8 inch by 1 inch copper flat stock is < $84 https://www.metalsdepot.com/brass-products/copper-flat-bar

What is the cost of the used modules, just curious to compare to new top of the line LFP prysmatic cells,  which would be way simpler to arrange for a workable bank needing a single BMS.

The other thing to consider are the settings for the charge controller, but more on that later.

jonoslade

@mcgivorThank you for the reply I'm reaching a bit with my experience in doing this build so any advice is certainly appreciated! I've done several small RV builds but nothing of this size/complexity.

The modules are Lithium Polymer Flat Pack I believe. After further research, each module already has a 300amp fuse build in.
Individual Specs and notes:

--used condition. tested and balanced.

-- Original capacity: 3.2kWh, 62 Ah

-- Currently, batteries range from 60-85% of original capacity.  Typically, 2.2kWh, 43Ah

-- weight: 83LBS

-- Internally they are 2x 7s 24V  cells in series giving 14S 48V on external terminals

Price was $202 each

The BMSs were sourced here: https://www.aliexpress.com/item/32997552090.html?spm=a2g0s.9042311.0.0.40694c4ddzaYMt

They are 7S to 16S 100A.

1 and 2. The thoughts on the shunt and copper bus came from some reading I've done here from @john p which was really interesting. https://forum.solar-electric.com/discussion/6495/battery-interconnections/p2

I did find these bus bars on super clearance and am now looking at using them instead of the copper pipe or flat stock (although your source for the copper flat stock is way cheaper than anything I came across): Blue Sea Systems 1993, PowerBar BusBar with Cover, 12 Gang, 5/16"-18 Studs, 1,000 Amp

mcgivor

The battery modules are what I suspected, LIPO are among the more unstable potentially dangerous chemistries, this is important to understand, whlist it may be possible to limit the charging voltage to reduce the risk, they are not something I would recommend on an inside installation.

Knowing the makeup of the battery itself would be helpful, some are made up of tens or hundreds of small cells in parallel to form a cell block which are then series connected to achieve the nominal voltage. With multiple cells within a cell block it is common to have failures of individual cells, which ultimately reduces the overall capacity of the module or battery, hence the reduction of advertised reduction of original capacity.

Personally I wouldn't consider used batteries, especially LIPO, simply because there are better, safer options, available  with testing records for capacity and sequential serial numbers albeit at a slightly higher price. Considering the used modules and multiple BMS's required.the price would be around 30% more to have new verified cells with a more stable chemistry.

Please don't take this as discouragement but rather insight, keeping things simple with large prysmatic LiFePo4  cells reduces the overall complexity and increases safety with a single BMS. Much of my apprehension revolves around the known instability of LIPO cells, particularly with  someone unfamiliar with the potential danger,  hope you understand my concerns. 

jonoslade

Thanks for the insights @mcgivor ! Your concerns are certainly noted and taken to heart. This isn't something I want to take lightly but I hope that I can still move forward with the purchased equipment and make it as safe as possible. Please talk me out of it if you think the plan below is not viable or too dangerous (not in comparison to more stable chemistries but just in general).

My hope is with a proper fusing and BMS setup I can make these modules work. The power usage is quite low for the house and the battery/solar shed is not inside the living space although it is attached so a catastrophic failure and fire would be a big issue. I've watched the LIPO pack cell failure videos and that definitely scares the crap out of me. 

The modules are made up of 2 24v packs with 14 flat LIPO cells. They have a 300amp ANL fuse between the two 24v packs. I am currently planning to attach a BMS to each module. Then run 2/0 cable from each terminal to the Blue Sea bus bars.
From the positive side, I'll have a terminal fuse 200amps (is current plan) > victron BMV712 > 250amp breaker (existing install) > inverter (all with 4/0 cable). 

jonoslade

See if these image links for the modules work:

https://ibb.co/BNWNtPk

https://ibb.co/vVpmCfD

mcgivor

What's being proposed may well be possible though I don't think it would be considered the best path to take as previously stated, 16 new 200Ah grade A CALB  LiFePo4 cells with a BMS would be approximately the sane cost of the modules you referring to and $500 less if you include the 10 BMS needed and shipping of the cells. That would provide a 6.7 kWh per day of usable daily energy by operating between 20% and 90% state of charge, so I have to ask........

Have you purchased the modules already or is this something you're considering? What daily power needs?

Dave Angelini

I hope he has not bought them ! Not much of a help to use these for your parents. Buy them something new that is not dangerous.

My history with LG Chem is not a good one. We failed the LG Chem battery field test with Schneider Solar at my business. They could not take the charge and discharge rates of large inverter/chargers. If you research them you will see failures all over the world. LG then came out with a 400vdc battery that has been better but definitely not ready for offgrid. They also are NMC based and a much higher danger than LFP.