Parallel cells - what are the pitfalls as the number of cells in parallel gets deeper ?

[740GLE]

Is there even a good supply of LiFePo cylindrical cells? Sure NMC and other chemistries are more prevalent in 18650 27100 sizes, I just haven’t seen too many vendors carrying Lifepo.

I think Battery hook up had their 100ah kits, but handful of reviews has dismal capacity tests over some cycles and I think they pulled it from their site.

Seems like the go to option for DIY is prismatic when dealing with lifepo.

 

[OzSolar]

As pointed out, if a large 280Ah cell goes bad it will take out the entire battery. If that same 280Ah cell is in a 2P pair and one cell goes bad, the battery still maintains its voltage overall, but with less amps. Or does the one cell going bad take both cells out?

I will offer that if you lose one cell in 16S1P battery you can remove it from the pack, adjust the voltage parameters on the necessary devices accordingly and still have a functioning system with 15/16th the capacity of the original pack.

 

[A.Justice]

.... Cons of going 96v

1/ Not much gear out there for a 96v nominal system.

Except for the motor, which is THE piece of gear for you. You can always get a step-down converter for other devices.

2/ What is out there is more expensive.

True. You might save some $ on wiring.

3/ Safety concerns - DC arching at that voltage seems to be an issue that needs to be carefully considered.

Also true, but the risk can be mitigated by using proper safety procedures.

4/ Loss of solar on the boat - this is a big one. Cant get enough panels in a series string to get the voltage high enough to charge a 96v bank.

I have some 60v nominal panels, I have heard of ones at 80+ volts. That would either be 2 or 3 in series, you might have to get some special panels though.

5/BMS's become harder to find for this voltage.

They sure do.

I am not tied to 96v at all. I would prefer to stay at 48v.

There are 2 things that are critical in this changeover to electric propulsion - the motor running as cool as possible and a tight battery setup. 96v seems to tick both boxes resoundingly.

I agree 100%. I would build my system around my main load, and what it takes to run that load properly and efficiently.

 

[A.Justice]

Is there even a good supply of LiFePo cylindrical cells? Sure NMC and other chemistries are more prevalent in 18650 27100 sizes, I just haven’t seen too many vendors carrying Lifepo.

I think Battery hook up had their 100ah kits, but handful of reviews has dismal capacity tests over some cycles and I think they pulled it from their site.

Seems like the go to option for DIY is prismatic when dealing with lifepo.

A123 is the only manufacturer I know of. Not sure who makes the BH cells.

 

[Texas-Mark]

To me a 25P is a bridge too far. There have to be higher odds of a cell going bad on you when you times your chances by 25. At least it would seem so

One will never get a high capacity pack without paralleling. And I don't think the odds of a failure really go up that much. And in some cases, I think the odds go down. I have a 10P setup, but don't do a heavy draw on it. So each cell is only supplying 1/10 of the total draw which is better for the cells (reducing the odds of a failure).

 

[Steve_S]

Suggested Short Watch

 

[Ohms_Cousin]

Suggested Short Watch

So BB use small cylindrical cells in parallel. Never heard of a BB battery going bad. Not saying it doesn't happen, just that it seems any issues with BB are far and few between. However they have real engineering behind them. It does show however that many small cells in parallel does work.

 

[Ohms_Cousin]

One will never get a high capacity pack without paralleling. And I don't think the odds of a failure really go up that much. And in some cases, I think the odds go down. I have a 10P setup, but don't do a heavy draw on it. So each cell is only supplying 1/10 of the total draw which is better for the cells (reducing the odds of a failure).

Seems there is no fixed answer on this. As you say, you are running 10P with out issues. Others warn off paralleling cells, yet others feel its fine.

I think the bottom line is, as you state - to get a high capacity pack means that paralleling is how it done.

 

[toms]

Seems there is no fixed answer on this. As you say, you are running 10P with out issues. Others warn off paralleling cells, yet others feel its fine.

I think the bottom line is, as you state - to get a high capacity pack means that paralleling is how it done.

You really need to look closely at the level of experience of the people giving advice.

Sometimes it’s not “how much time have you spent reading stuff and watching youtubes”, rather “how many systems have you commissioned and over what time frame”?

 

[Ohms_Cousin]

rather “how many systems have you commissioned and over what time frame”?

and THAT is the crux of it.

Thing is plenty of people have working systems using either method - a straight series config, some with a straight series config with multiple batteries in parallel, some with cells paralleled first then put in series and then systems with cells paralleled fist then put into series and then paralleled batteries in the bank.

So to get capacity, it really comes down to cells paralleled first then put into series and then having multiple batteries in a bank OR making a battery out of single cells connected in series and then having paralleled batteries in the bank. How else cn one get to a large capacity if around 200Ah cells are the limit for boats?

Pros and cons on each approach. If possible I would stick with cells connected in series and a BMS monitoring every cell which I can see, but that just does not give the capacity needed -unless I go to a large number of series connected batteries connected in parallel to a bank. I am shooting for a 50kWh battery bank x 2. I would like to have 2 completely different battery systems for redundancy. Its all very well wanting but doing is another thing altogether. Space and weight concerns rear thier ugly head here.

So the next step down the ladder in terms of risk would be a battery made from cells connected in series with its own BMS and then having multiple of these batteries in parallel - lots of BMS's needed for this approach, plus main fuse for each battery and probably a contactor for each battery as well. If one goes for the cheaper Chinese BMS that seems to be the flavour of the month it can help with keeping costs down to some degree - but then how good and reliable is that BMS? Reading some of the comments on another thread, the seller does not back up the product. So seems a bit suspect.

If one goes for a high end BMS you can be assured it will work well but it will be very expensive for all those BMS's. Then is still the main fuse and contactor per battery that is paralleled in the bank.

Next step down the ladder in terms of risk is paralleled cells first then the paralleled cells connected in series to give the desired voltage. Then repeat this a number of times by connecting these batteries in parallel to make up the battery bank. The only advantage here is a lessor number of batteries in parallel in the bank because the capacity was built up somewhat by paralleling cells first and then connecting in series to the desired voltage. So lessor number of BMS's and associated wiring etc, less number of main fuses and contactors...... BUT one loses visibility to each and every cell.

I think that pretty much sums it up from what I have read about it.

Now which way to go .... that is the question and it seems there are as many opinions as there are people who comment. Which leads back to your comment above.

 

[labeeman]

I think a lot of misinformation about paralleling cells is spread by people who have never done it, because they came up with some whacky theories when they thought about doing so and were then too afraid to try.

On Nicad cells the manufacturer says do not put in parallel could be the source of misinformation.

 

[apctjb]

I have a battery bank using 32 EVE 280 cells configured 2 parallel, 16 series with a single BMS protecting the bank. Its has worked fine until recently when I had one cell run away triggering the BMS to shut down the system. I am now planning to reconfigure this system into two separate 16S batteries in parallel, each with a separate BMS to provide greater visibility of individual cell voltage and greater system redundancy. (Failure of one cells will not bring down the entire system)

Not suggesting that paralleling cells is good or bad, it all depends on the specific circumstances for each system and end user.

 

[cs1234]

Individual cell monitoring/balancing is always the ideal, it's just not always the most cost effective. There will always be a tradeoff between safety/performance/cost.

 

[Steve_S]

My own experience which drove me away for Parallel Cells was a Commercial Pre-Built pack with cells in parallel that were NOT Matched, this resulted in a runaway situation resulting in almost burning through the FR4 Sheet. That pack was disaasembled and rebuilt into two, each with Active Balancing BMS and now they usually have a delta of 0.005 difference between them without issues.

 

[Texas-Mark]

My own experience which drove me away for Parallel Cells was a Commercial Pre-Built pack with cells in parallel that were NOT Matched, this resulted in a runaway situation resulting in almost burning through the FR4 Sheet.

Did it not have a BMS?

 

[Ampster]

Is there even a good supply of LiFePo cylindrical cells?

Headways are LiFePO. They are larger format than 18650. I believe 38120 and 40152.

 

[Steve_S]

Did it not have a BMS?

Yes it did, a dumb BMS.

 

[740GLE]

Headways are LiFePO. They are larger format than 18650. I believe 38120 and 40152.

I checked their site and noticed those, looks like they came up with a replacement to those 100ah bricks they previously had.

 

[Substrate]

The old addage of having "no more than 4 batteries in parallel" comes from the FLA world, where more than that becomes a maintenance chore for the average diy'er, trying to keep one battery from overcharging, and another from undercharging and sulfating. Quite the common problem for those who slap 4 together from different makes, models, capacities, and soc's, and differing jumper cable IR's together initially. Even harder with more than 4!

Cylindrical LFP's? No need to worry about compression. Go down that 500 thread rabbit hole elsewhere.

Ideally, all cylindricals should be matched for capacity AND ir. Advanced diy'ers will also perform perhaps a month-long self-discharge test, and group them accordingly. Not that they are all bad, the idea is to group them together as self-alike as much as possible before paralleling them into a pack for eventual series / parallel needs to for the overall battery. Makes the bms's job a little easier.

If one just slams 100 cells together randomly, and if they are of lesser quality, then yes, some cells can have their poor performance "hidden" by the surrounding paralleled cells. This means lot's of bms activity / balancing.

Other than overall capacity and IR grouping, it would be rare for a manufacturer to also do a self-discharge test grouping before assembly. Hence, go with reputable suppliers where the chances of cell deviations are small, and assembly is precise.

 

[jah]

Is there even a good supply of LiFePo cylindrical cells? Sure NMC and other chemistries are more prevalent in 18650 27100 sizes, I just haven’t seen too many vendors carrying Lifepo.

I think Battery hook up had their 100ah kits, but handful of reviews has dismal capacity tests over some cycles and I think they pulled it from their site.

Seems like the go to option for DIY is prismatic when dealing with lifepo.

Take a look to this post Thread 'Cylindrical cells 100AH' https://diysolarforum.com/threads/cylindrical-cells-100ah.22958/