LiFePo4 -- Parallel or Not?

[CatNewBee]

A parallel cell setup like 2P4S is not redundant, you only increase the capacity of a single cell.

If we are talking about real redundancy, then all components must be built redund, inclusive BMS and power logic, and and the design gets much more complex.

[hzcruiser]

^^^^Agreed

[rgleason]

Originally Posted by CatNewBee

A parallel cell setup like 2P4S is not redundant, you only increase the capacity of a single cell.

If we are talking about real redundancy, then all components must be built redund, inclusive BMS and power logic, and and the design gets much more complex.

I would call the FLA/AGM Starter/Reserve as backup unless that is also made into LiFePo4 + BMS

[s/v Jedi]

Originally Posted by hzcruiser

Elegance? Hmm, ok. I would call it simplicity, and it certainly is simpler. However, many cruisers strive for redundancy which is something your approach does not provide.

Most "conventional" FLA setups also have more than one 12V battery in parallel in order to get the (usable) capacity needed, so this is nothing new (or ugly) IMHO.

No, I’m completely with you and your setup. What I don’t like is setups like 4p4s configurations to make 400Ah out of 100Ah cells in one single battery. I have two house battery banks and normally two AGM starting batteries although currently, temporary just on one starting battery.

Bringing the battery close to the consumer is something I agree with as well. Our boat came with big cables to deal with the voltage drop issues or I would have done the same.

[T1 Terry]

Originally Posted by s/v Jedi

I have never heard of it, but did not really search as it’s irrelevant to me. I much prefer the elegance of just single cells in series. I believe the drive for many cells in parallel is driven by low cost cells that only come in smaller capacities.

The cost per cell for quality manufactured gear is based on Ah capacity, 2 x 100Ah cells cost the same as 1 x 200Ah cell. If people are looking at buying cheap cells and then talk about redundancy, that just doesn't make sense. If you use quality gear then the quality of the BMS should be the part that makes a second back up system no longer necessary.
Let's face it, do you drag a spare hull along with you just in case there is a problem with the one you are using?
I understand the idea of redundancy as far as inverters etc, but the BMS should be built in a way that there are enough back ups in play to get by with only half of it working ...... that is true redundancy .....

T1 Terry

[T1 Terry]

Originally Posted by hzcruiser

Well, I can't really agree with that statement, although I know where you're coming from. Parallel batts do increase the max current draw, as every batt is able to provide, say 0.5C of its capacity safely.
Nevertheless, I agree it would be prudent to design your system so that the loss of some LFPs is not overloading the others. But that's the same for FLA batts.
In reality, the batt capacity is designed to keep the boat running for at least two days at average consumption and no charge. Hence there can't be a consumer that draws so much power that the batts would be overloaded.
As an example, my 2000W inverter is next to a 100Ah LFP. But I never really draw 2000W, it's more like a 100-600W. Another example is the windlass: there is another 100Ah LFP up front and it can handle the short startup current of 120A but while I'm weighing anchor the average draw is around 60A. If that batt would get tired, its voltage would sag and there would be more current coming from the other LFPs in the back.
This does and has happened, even more with the FLA batt that was in there originally.

So separating and paralleling the batts does not save on wiring diameter but it reduces losses, spreads the weight and provides redundancy.

You have not brought into consideration the difference between the internal resistance of lithium batteries compared to lead acid batteries. A lead acid battery will drop terminal voltage as the load is increased allowing the other batteries in parallel to add to the total load.
That doesn't happen with lithium batteries. The terminal voltage will hold up until the battery is virtually completely drained, the resistance per mtr in the paralleling cables will always create a voltage drop the higher the current draw between each battery and this just increases the load on the battery with the shortest cable run and/or heaviest cables between the load and the battery.
Theory says it doesn't happen, a 15 minute demo to a few hundred stand alone system users (caravans and motorhomes) at a get together showed just how serious the problem is when lithium batteries were involved. There were a number of drop in battery resellers there that I invited to do the same test using their batteries, I was quite happy to let them use the 4 x Victron BMV's to measure the output from each battery and the inverter and load to prove or disprove the demo I'd shown live up on the big screen ..... they didn't take me up on it

T1 Terry

[hzcruiser]

Originally Posted by T1 Terry

You have not brought into consideration the difference between the internal resistance of lithium batteries compared to lead acid batteries. A lead acid battery will drop terminal voltage as the load is increased allowing the other batteries in parallel to add to the total load.
That doesn't happen with lithium batteries. The terminal voltage will hold up until the battery is virtually completely drained,

Hello Terry. I'm aware of the impedance difference between different chemistries. LFPs have a discharge curve as well even though the voltage sag is less pronounced than with FLAs. LFP then show the well known knee starting at about 80-90% DOD (depth of discharge).



From my experience with 40, 70 and 100Ah batts, even LFPs show a reduced terminal voltage at higher loads (>0.5C).

the resistance per mtr in the paralleling cables will always create a voltage drop the higher the current draw between each battery and this just increases the load on the battery with the shortest cable run and/or heaviest cables between the load and the battery.

It's correct that the voltage drop across the cable can be higher than the LFP sag for a significant load. One might argue that this makes it even more important to put the batts closer to the load and sized so that the closest batt can provide all the power for that consumer. In the rare event that one of those batts would reach 80% DOD the other batts would then be able to keep the power flowing.

[T1 Terry]

Originally Posted by hzcruiser

Hello Terry. I'm aware of the impedance difference between different chemistries. LFPs have a discharge curve as well even though the voltage sag is less pronounced than with FLAs. LFP then show the well known knee starting at about 80-90% DOD (depth of discharge).



From my experience with 40, 70 and 100Ah batts, even LFPs show a reduced terminal voltage at higher loads (>0.5C).






It's correct that the voltage drop across the cable can be higher than the LFP sag for a significant load. One might argue that this makes it even more important to put the batts closer to the load and sized so that the closest batt can provide all the power for that consumer. In the rare event that one of those batts would reach 80% DOD the other batts would then be able to keep the power flowing.

I guess there could be specific tasks that were very short term where a battery for each application yet linked via parallel cabling could put forward a case where that set up was preferable ...... but surely the cost for a quality BMS for each battery far outweighs the cost of heavier cabling from the centrally mounted battery pack to the short term high draw loads. The cable running between the batteries already needs to be substantial to avoid the voltage drop over length negating the other batteries sharing the load until the closest battery is pulled well down the death knee that ends in cell failure.

The drop in voltage under a 1CA load (for quality cells in good condition)
Shenzhen Smart Lion Power Battery Limited
As can be seen in the chart, a 1CA load on a 100Ah cell (100 amps) the voltage still holds better than 3v per cell even at the 100% discharged level, it is at the 110% discharged level, not good for long cycle life to do that to a cell and the chances of getting 4 cells in series down that low without one of the cells being that bit lower in capacity and hitting zero voltage while under load . That is usually fatal for any chemistry cell because reverse current flow is the next step if the battery is still under load and this is certainly fatal for an LFP cell.

T1 Terry

[CarlF]

T1 Terry,

Your warnings on parallel connection of drop-ins seem reasonable but I'm having trouble squaring that with the large number of owners - especially big cat owners - who are putting in house banks of 4 or 5 300AH drop-ins connected in parallel to run big inverters, induction cooktops or even air conditioning. With the battery suppliers and installers apparently not raising a concern. Some have been cruising for years without having a problem.

What am I missing? I'm still sailing with a lead mine so this is all new territory to me.

[T1 Terry]

Originally Posted by CarlF

T1 Terry,

Your warnings on parallel connection of drop-ins seem reasonable but I'm having trouble squaring that with the large number of owners - especially big cat owners - who are putting in house banks of 4 or 5 300AH drop-ins connected in parallel to run big inverters, induction cooktops or even air conditioning. With the battery suppliers and installers apparently not raising a concern. Some have been cruising for years without having a problem.

What am I missing?

What is the maximum continuous draw from one of the 300Ah drop in batteries? Are we talking 12v or 24v? Generally what you will find with a drop in high capacity battery, the cells are linked in parallel under the plastic top. Say the 300Ah capacity battery is built from good quality cells in parallel/series inside the plastic box. A 300Ah cell @ a 1CA discharge rate is 300 amps, there would not be an issue in such a case. The problem comes when 3 x 100Ah drop in batteries are linked together in parallel and the same 300 amp load is applied. Now the first battery in the parallel string will see a 3CA load, looking at the chart I put up a few posts back, @ a 3CA load the voltage will still be above 3v per cell @ 50% SOC before the voltage dropped enough to equal the voltage drop in the next length of cable and the second battery in the parallel string starts to add assistance. By the time the 3rd battery in the parallel string comes into play, the first battery is seriously depleted, the second battery is 50% depleted and the third battery is only just starting to share the load. Recharging is the reverse, the first battery gets the full current until the voltage is high enough to over come the voltage (now) increase due to cable resistance for the current to start to flow to the second battery, then the same deal before the third battery gets charged .... if of course the charge controller didn't sense the high voltage at the first battery and drop the charge voltage to a float setting or actually shut down the charging.
The voltage will equalise across the three batteries, but that does not mean the state of charge equalised across the three batteries.
Gradually the third battery gets less and less % SOC and later and later before it has a high enough terminal voltage under load to share the work with the other batteries. How seriously will the first battery be discharged before the third battery reaches 100% discharged?

Even 0.5v difference under load means the third battery is still at 13v when the first battery is at 11.5v .....

It doesn't take a lot to understand why this sort of set up is prone to failure

T1 Terry

[s/v Jedi]

Originally Posted by T1 Terry

What is the maximum continuous draw from one of the 300Ah drop in batteries? Are we talking 12v or 24v? Generally what you will find with a drop in high capacity battery, the cells are linked in parallel under the plastic top. Say the 300Ah capacity battery is built from good quality cells in parallel/series inside the plastic box. A 300Ah cell @ a 1CA discharge rate is 300 amps, there would not be an issue in such a case. The problem comes when 3 x 100Ah drop in batteries are linked together in parallel and the same 300 amp load is applied. Now the first battery in the parallel string will see a 3CA load, looking at the chart I put up a few posts back, @ a 3CA load the voltage will still be above 3v per cell @ 50% SOC before the voltage dropped enough to equal the voltage drop in the next length of cable and the second battery in the parallel string starts to add assistance. By the time the 3rd battery in the parallel string comes into play, the first battery is seriously depleted, the second battery is 50% depleted and the third battery is only just starting to share the load. Recharging is the reverse, the first battery gets the full current until the voltage is high enough to over come the voltage (now) increase due to cable resistance for the current to start to flow to the second battery, then the same deal before the third battery gets charged .... if of course the charge controller didn't sense the high voltage at the first battery and drop the charge voltage to a float setting or actually shut down the charging.
The voltage will equalise across the three batteries, but that does not mean the state of charge equalised across the three batteries.
Gradually the third battery gets less and less % SOC and later and later before it has a high enough terminal voltage under load to share the work with the other batteries. How seriously will the first battery be discharged before the third battery reaches 100% discharged?

Even 0.5v difference under load means the third battery is still at 13v when the first battery is at 11.5v .....

It doesn't take a lot to understand why this sort of set up is prone to failure

T1 Terry

No, this is BS which can not happen when the wiring is done right. When you wire three drop-in batteries in parallel, connect the positive busbar cable to the first battery and the negative busbar cable to the last battery of the string, then the 300A load splits over the three batteries and each only sees 100A.

[evm1024]

Originally Posted by s/v Jedi

No, this is BS which can not happen when the wiring is done right. When you wire three drop-in batteries in parallel, connect the positive busbar cable to the first battery and the negative busbar cable to the last battery of the string, then the 300A load splits over the three batteries and each only sees 100A.

I agree, BS it is. Ohm's and Kirchhoff's laws say it is BS.

When a load is presented to 1, 2, 3 or N batteries the current supplied from each battery is proportioned.

In an ideal case where all 3 batteries are equal in all respects (as is their wiring to the summing node) the current supplied will be equal (1/3 each).

Minor variations in battery SOC and such and actual wiring resistance will alter the exact current from each battery. but that will come to an equilibrium is relative short order.

Given the normal wiring to each battery (being nearly equal) it is not possible to have the voltage at one battery be much different from any other battery in the array.

[CatNewBee]

Originally Posted by evm1024

I agree, BS it is. Ohm's and Kirchhoff's laws say it is BS.

When a load is presented to 1, 2, 3 or N batteries the current supplied from each battery is proportioned.

In an ideal case where all 3 batteries are equal in all respects (as is their wiring to the summing node) the current supplied will be equal (1/3 each).

Minor variations in battery SOC and such and actual wiring resistance will alter the exact current from each battery. but that will come to an equilibrium is relative short order.

Given the normal wiring to each battery (being nearly equal) it is not possible to have the voltage at one battery be much different from any other battery in the array.

True for anything parallel, wrong for serial things. In serial circuits the current is the same and the voltage may vary.

that said, 3 parallel batteries (built each of 4 cells in series) will have the same pack voltage, but the cells may have very high differences in the strings. If only one cell is foul, one pack will be charged less, have a different resistance and may not take 1/3 of the current either way. The cells of the whole string may be damaged, not only the one with the issues, because the other cells may run off when the string is not isolated / disconnected from the other packs.

[T1 Terry]

Originally Posted by s/v Jedi

No, this is BS which can not happen when the wiring is done right. When you wire three drop-in batteries in parallel, connect the positive busbar cable to the first battery and the negative busbar cable to the last battery of the string, then the 300A load splits over the three batteries and each only sees 100A.

The next demo I do I'll video it and put up a You Tube link, that should dispel the theory V reality problem. In the mean time, even linking lead acid batteries that way has issues, there is website in the UK that posted a whole lot of info about this yrs back, I'll try to find the website and put up a link. I put it up on a forum over here and a Uni professor and an EE disputed it claiming it didn't match computer modelling. When I posted up the photos from multiple shunt readings where the shunt was between each battery confirming that it did actually happen with lead acid batteries, they claimed the testing was faulty ..... yet having the access to all the university equipment and labs, he never came back with an opposing finding. Khirchoff's 1st law, the current law relates to junctions, the voltage law relates the total closed loop will always equal zero. A refresher here might help https://isaacphysics.org/concepts/cp_kirchhoffs_laws
Ohm's explains exactly how this voltage difference you can't seem to reason through actually occurs, put the numbers in and you will see the result is the same in theory as it is in reality ..... but you must put all the figures in or the theoretical result will be incorrect.
Just as a starter, what is the voltage difference at the battery terminal between 1 amp charging and 1 amp discharging? Once you can get your head around how that happens you are on the way towards understanding why batteries in parallel will not load share if there is any cable between them because of the additional resistance added to the internal resistance of each battery.

T1 Terry

[evm1024]

So basically you are saying that I cannot parallel 3, 100AH batteries and pull out 300 amps

Is that what you are saying - That the current draw is limited to the 1 C rate of the individual batteries?

And that this is a function of the cells or the BMS in the dropin?

Say the 300Ah capacity battery is built from good quality cells in parallel/series inside the plastic box. A 300Ah cell @ a 1CA discharge rate is 300 amps, there would not be an issue in such a case. The problem comes when 3 x 100Ah drop in batteries are linked together in parallel and the same 300 amp load is applied. Now the first battery in the parallel string will see a 3CA load, looking at the chart I put up a few posts back, @ a 3CA load the voltage will still be above 3v per cell @ 50% SOC before the voltage dropped enough to equal the voltage drop in the next length of cable and the second battery in the parallel string starts to add assistance. By the time the 3rd battery in the parallel string comes into play, the first battery is seriously depleted, the second battery is 50% depleted and the third battery is only just starting to share the load.