Large Single Battery vs Smaller Ones- Resurrecting this old topic

[Henderson]

Firstly, let me apologize for resurrecting this topic which has been discussed and beaten to death and that is the topic of building a large battery as opposed to a multiple smaller batteries and running them in parallel. In my scenario, I'm looking at 96 cells...lifepo4 with 280Ah cells which can make me a 6P16S...quite a large battery. I raised this question in this forum a few months back when I started this DIY solar journey and many persons seemed to recommend multiple smaller batteries than one large one. This got me to thinking why was that the case. Let me say early, I'm not trying to cheap out on buying multiple BMS-es. What I am trying to do is to avoid complexity in my battery build. I am also very concerned about multiple batteries in parallel and having to deal with inrush currents if I have to remove one of the batteries for some reason and when adding it back, it will potentially have a different SOC to the others which can cause the inrush.

They are BMS-es out there that can handle up to 500A max with relay with integrated active balancers to the tune of 2A to 5A and dedicated balancers up to 10A so balancing such a large bank shouldn't be too much of an issue assuming I'm starting out with all grade A cells and balanced properly to begin with.

Redundancy is not a major concern of mine as I will still maintain my connection to the grid and only use the grid as a backup. Plus, I also have a generator at the house as a last resort.

In terms of cabling for the large single battery, I'm looking at multiple cables in parallel to handle the current, including multiple parallel cables from the battery's main +ve and -ve terminals. I'm still doing some calculations on the battery's bus bar side of things but it's looking quite doable.

During my research over the last several weeks, I have not been able to find any scientific reason (outside of redundancy) why building a large battery as opposed to multiple smaller ones is a bad idea. Am I missing something? Would appreciate your comments and thanks in advance.

 

[timselectric]

In my opinion, redundancy is the main reason for multiple smaller batteries. If you don't mind losing the entire battery bank to an issue with one cell, one BMS is fine.
In my case, I'm doing multiple. But each will be capable of powering all needs. With my extra capacity, I will also have redundancy.

 

[mikefitz]

Multiple smaller batteries will have individual cell monitoring via the BMS. A large battery with several cell In parallel won't have that feature.

Mike

 

[Substrate]

You have answered your own question. You want simplicity and are not concerned about redundancy. If it fits your budget, go with the system that reduces possible points of failure.

You will always get two answers to this, so it depends on what you desire. Much of it depends on your own skills to ensure that you don't add additional POF's of your own making.

If anything, my suggestion is to make an all-purpose starter system first and see how that goes. You can always use that as a convenient addition to your big system later.

The biggest historical mistake even back in the lead-acid days was when people got in over their heads at the outset. Instead of a pair of common 6V GC-2's wired in serial for a "starter" 12v system, they got way over their heads with a huge multi-kw $$Rolls or other without knowing their own skills or enthusiasm to maintain such.

Usually ending up as the "in the corner of the garage collecting dust" system. Along with their wallet. And never again seen on any forum help.

 

[toms]

Firstly, let me apologize for resurrecting this topic which has been discussed and beaten to death and that is the topic of building a large battery as opposed to a multiple smaller batteries and running them in parallel. In my scenario, I'm looking at 96 cells...lifepo4 with 280Ah cells which can make me a 6P16S...quite a large battery. I raised this question in this forum a few months back when I started this DIY solar journey and many persons seemed to recommend multiple smaller batteries than one large one. This got me to thinking why was that the case. Let me say early, I'm not trying to cheap out on buying multiple BMS-es. What I am trying to do is to avoid complexity in my battery build. I am also very concerned about multiple batteries in parallel and having to deal with inrush currents if I have to remove one of the batteries for some reason and when adding it back, it will potentially have a different SOC to the others which can cause the inrush.

They are BMS-es out there that can handle up to 500A max with relay with integrated active balancers to the tune of 2A to 5A and dedicated balancers up to 10A so balancing such a large bank shouldn't be too much of an issue assuming I'm starting out with all grade A cells and balanced properly to begin with.

Redundancy is not a major concern of mine as I will still maintain my connection to the grid and only use the grid as a backup. Plus, I also have a generator at the house as a last resort.

In terms of cabling for the large single battery, I'm looking at multiple cables in parallel to handle the current, including multiple parallel cables from the battery's main +ve and -ve terminals. I'm still doing some calculations on the battery's bus bar side of things but it's looking quite doable.

During my research over the last several weeks, I have not been able to find any scientific reason (outside of redundancy) why building a large battery as opposed to multiple smaller ones is a bad idea. Am I missing something? Would appreciate your comments and thanks in advance.

Using smaller cells in parallel to obtain the required kwh battery pack has proven to be very reliable and long lasting.

There is a massive misunderstanding about how parallel cells interact, i see it often here on this forum where people have no idea on how parallel cells work in practice.

 

[Substrate]

Very true. But quite often individual skills are not taken into account, pushing a shopping-cart lego-land plug-n-play concept that totally overlooks things like points-of-failure.

Hence, the two views will never seem to meet.

 

[Henderson]

Thanks all. Really appreciate your input on this.

 

[Any name you wish]

This is great info. Is there a capacity you would use as cutoff for multiple batteries or is it managed by voltage of those batteries? Say one large 48V and any less is multiple batteries? I have a 100Ah one and want more capacity but cash is limited to small bonus money from work so I have about $400 for the past 6 months and don't see myself getting 1.5k any time soon.

 

[Henderson]

Using smaller cells in parallel to obtain the required kwh battery pack has proven to be very reliable and long lasting.

There is a massive misunderstanding about how parallel cells interact, i see it often here on this forum where people have no idea on how parallel cells work in practice.

@toms Can you elaborate a bit more on the point you're trying to make? Thanks.

 

[timselectric]

@toms Can you elaborate a bit more on the point you're trying to make? Thanks.

They are in the one BMS camp. (They said "cells" , not batteries)

 

[Any name you wish]

They are in the one BMS camp. (They said "cells" , not batteries)

So you recommend a single BMS with multiple parallel cells then?

 

[timselectric]

So you recommend a single BMS with multiple parallel cells then?

No, I'm in the redundancy camp.
I was just clarifying their response for you.

 

[toms]

@toms Can you elaborate a bit more on the point you're trying to make? Thanks.

I still see many people believing you can have different voltages across a parallel pair.

With multiple parallel batteries you inevitably end up with balance issues.

If you need a 200ah cell, two 100ah cells gets you there.

If you need redundancy, build two independent systems, if you want the extra complexity of parallel batteries so you can have half the battery capacity in the event of a cell failure then go parallel batteries.

Overwhelmingly with cheaper components the inverter or BMS are the first to fail - it’s rare for a cell to fail without first dropping capacity.

 

[Henderson]

I still see many people believing you can have different voltages across a parallel pair.

With multiple parallel batteries you inevitably end up with balance issues.

If you need a 200ah cell, two 100ah cells gets you there.

If you need redundancy, build two independent systems, if you want the extra complexity of parallel batteries so you can have half the battery capacity in the event of a cell failure then go parallel batteries.

Overwhelmingly with cheaper components the inverter or BMS are the first to fail - it’s rare for a cell to fail without first dropping capacity.

Gotcha, thanks....makes sense!

 

[Warpspeed]

You also need to think about what happens with parallel connected cells, if one cell fails shorted.

One individual cell may die fairly gracefully, but if there are several other good parallel cells driving power directly into the shorted cell, things may become a bit more lively.

I have had three Winston cells fail shorted through dendrite formation over four years. Perfect one day, dead short the next day with just a slight swelling of the battery case.

There may not be enough stored energy contained in one single cell to raise the internal mass to a dangerous temperature or pressure very quickly with this type of internal short, which may be gradual in forming. But not so sure how the same kind of failure might manifest itself with multiple cells all sourcing high additional fault current.

Replacing one dead cell is a nuisance.
Having to replace several others as well, because one killed all its its partners, can become expensive.

 

[timselectric]

Yup
Anytime you put batteries/cells in parallel. There should be OCP between them. Because, a short draws all available current.
If that doesn't make sense. Maybe the firefighters can explain it better, when they arrive.

 

[OffGridInTheCity]

I don't post this to argue for any one position - but rather to share why I personally have multiple batteries in parallel for my powerwall. I'm 18650 but this would apply to groups of LifePo4 DIY as well. A key issue is the size of the end-result Powerwall.

There are several reasons.....
1) Over a Long Time: I started with Battery #1 (in pic below) and over the last 3 years have continued to build / expand, battery after battery - hooking in parallel as I went. Could not have created the overall powewall in any other way.
2) Maintenance: By having individual batteries I can take 1 out of service and work on it while leaving the rest of the system functional. Hasn't happened yet but eventually some of these packs will fail as the cells degrade.
3) Physical size/weight: Each battery weights ~150lbs. 8 * 150lbs = 1,200lbs. Too big to 'deal with' as one large battery.



4) BMS: Batrium makes it easy to expand but at the same time have a constant infrastructure without changing things. Each blue bar is a pack being monitored. When I add a new 14s battery, all I have to do is add the 14 new longmons to the monitoring network and tell Batrium the total number. 8 batteries x 14 packs(mons) = 112 packs. Batrium will let me go up to 249 packs on a single system.



Operationally, I haven't had any issues with charge/discharge of paralleled batteries. A key thing in my design is well matched battery capacities/capabilities coupled with low stress current per battery (and per cell). The lower the stress, the less problem one has with current flow thru the battery bank. This is a happy byproduct of over-sizing to extend cell lifetimes.

 

[toms]

Yup
Anytime you put batteries/cells in parallel. There should be OCP between them. Because, a short draws all available current.
If that doesn't make sense. Maybe the firefighters can explain it better, when they arrive.

I guess i better unravel my prismatic cells and fuse between each layer then!

Dendrite shorts are common in older or mistreated cells, i have seen scans of several cells that have failed in this manner. What happens is that the remainder of the cell (which is you guessed it in parallel with the section of the cell with a dendrite puncture) discharges high current through the dendrite and blows it open circuit.

That section of the cell no longer carries any current.

Please explain how the failure mechanism is different between a dendrite failure in a single 200ah cell, and a dendrite failure in a 100ah cell connected in parallel to another 100ah cell?

 

[timselectric]

I guess i better unravel my prismatic cells and fuse between each layer then!

Dendrite shorts are common in older or mistreated cells, i have seen scans of several cells that have failed in this manner. What happens is that the remainder of the cell (which is you guessed it in parallel with the section of the cell with a dendrite puncture) discharges high current through the dendrite and blows it open circuit.

That section of the cell no longer carries any current.

Please explain how the failure mechanism is different between a dendrite failure in a single 200ah cell, and a dendrite failure in a 100ah cell connected in parallel to another 100ah cell?

If you are saying that it's impossible for a LiFePo4 cell to develop a dead short, that's wonderful news.
Otherwise, I would recommend protecting against what could be a very bad day.

 

[Warpspeed]

What happens is that the remainder of the cell (which is you guessed it in parallel with the section of the cell with a dendrite puncture) discharges high current through the dendrite and blows it open circuit.

That section of the cell no longer carries any current.

That is not what happened here. The dendrite caused a short circuit that totally discharged the whole cell. Nothing was ever blown open circuit.

If that was the failure mechanism, as individual cell pockets went open circuit, we would see only a reduction in cell capacity.
What I have here, are three totally dead shorted cells, with zero measurable voltage and only very slight barely visible swelling.

It appears that the whisker, or whatever it is, has enough resistance to rapidly drain current, and grow without blowing like a fuse. It just eventually drains the whole cell right down to zero.
What happens if you directly parallel that up other healthy cells I have no idea, but the results can never be good.