# Li-ion battery cell balancing statistically unnecessary when many cells are in parallel?

Lithium ion batteries packs need to be balanced because of a small random voltage difference for each section in series, that may cause the voltage levels in a pack to diverge over time, right? So if you had a large number of cells (<40) in each series section wouldn't it be unnecessary to have balancing circuitry as the voltage difference of each series section would be extremely small due to the large sample?

• The way I am thinking: you have a bunch of parallel cells in a section. Several sections are then placed in series. Balancing is not required because the average voltage of each section is always very similar due to the large amount of cells in parallel. Aug 16, 2018 at 15:55

When several lithium cells are connected in series, it is the variation between series sections that requires balancing to be used. The problem with leakage and charge efficiency is that differences in these have a cummulative effect, and battery imbalance grows with each charge/discharge cycle.

I understand that you think that making each series section out of many cells in parallel would tend to even out the variation between those series sections.

However, 'tend to even out', and 'reduce to such a low level that it's not an issue during the expected lifetime of the pack' are very different things.

In the worst case, each series section will be as unbalanced as each individual cell. So there may be no improvement.

In the 'typical' case, you could statistically expect sqrt(number_of_cells) less variation in the parallel ensemble than for individual cells. For 40 cells, that's about 1/6th. So if you relied on this improved balance, you could typically cycle your cells 6 times more often before they became as unbalanced as a single series string. Doesn't really sound good enough improvement to me.

In the matched case, where you measure all the cells for capacity, internal leakage, charge and discharge efficiency, over temperature don't forget, and select the cells so each series section matches on all parameters, unfortunately those parameters change with age. Maybe you'll get 10 or 20 cycles out of the pack before differential parameter drift causes them to become unbalanced.

If you have series lithium cells, actively balance the charge between them.

Some people wonder why we need to balance lithiums, when we never needed to balance lead or nickel chemistries. The answer is that lead and nickel can be overcharged safely (as long as certain limits are observed) which naturally balances them. Unfortunately, there are no conditions that will allow the safe overcharging of lithium chemistries, overcharge rapidly destroys the cell.

• Thanks. By the way, i was thinking more in terms of: reduce to such a low level that it's not an issue during the expected lifetime of the pack Aug 16, 2018 at 17:28
• I know you were thinking that, to avoid needing balancing. It reminds me of the joke about the watch with a lifetime guarantee. Just before it breaks, a sharp spring comes out and slashes your wrist. Unbalance in a lithium battery ends its life in a very short time. Even the best matching you can do between cells will result in only a few 10s of cycles, when you can expect 100s of cycles from an actively balanced battery. When you don't use active balancing, the cell stays balanced for life, its life. Aug 16, 2018 at 17:53

Your title mentions cells in parallel yet the text talks about cells in series, that's confusing. I just answer both cases:

In case you mean cells in parallel:

No, cell balancing does not work like that. What matters is the difference in voltage between individual cells.

Using a large number of cells does not guarantee that the cell with the lowest voltage will have a higher voltage, closer to the average. So when that cell is connected it will still be charged (with a large, possibly damaging current) by the other cells.

Any time two (or more) cells are about to be connected in parallel their voltage must be similar to prevent large current to flow. This is done by balancing. This applies to any number of cells.

In case you mean cells in series:

No, cell balancing does not work like that. What matters is the difference in capacitance of the individual cells.

When one cell is almost empty but the other cells are not then you cannot keep using the battery pack as that empty cell will then be reverse charged and it will be damaged forever. So discharging has to stop.

The same is true for charging, when one of the cells in the series is nearly full the charging current through it has to be lowered to not overcharge it.

Using a large number of cells does not guarantee that the cell with the lowest capacity will not be over charged/discharged.

Any time two (or more) cells are about to be connected in series the voltages of all cells need to be monitored to prevent over charging/discharging.

• The question is ambiguous -- the title says "many cells in parallel", but the body says "large number of cells in each series section", which may or may not mean the same thing. I think the first version of your answer was closer to the mark. Aug 16, 2018 at 14:50
• Indeed, only noticed that now. Aug 16, 2018 at 14:53
• I don't think you're right in the second sentence: "any time two (or more) cells are about to be connected in parallel their voltage must be similar to prevent large current to flow. This is done by balancing. This applies to any number of cells." I have never seen cells in parallel that have balancing circuitry. I understand that when initially connected in parallel the cells would need to have equal voltages though. Aug 16, 2018 at 15:20