Heavily boiled FLA battery

[foo1bar]

During normal operation, the flowing currents through each string would be more complex (dynamically changing in and out) but you can't have large variations in individual string currents as long as all the batteries are in a healthy condition and, obviously, they are of the same type.

What happens when battery V1 is 5 degrees warmer than battery V4 because V4 is next to the wall and V1 isn't?
What happens when you add resistance between R1 and R2, between R2 and R3, between R3 and R4, between R4 and load?
And same for the bottom - what happens when you change that ideal wire into a more realistic model with resistances?
The currents for the different batteries will be different - probably only slightly different when the batteries are new.
But as the batteries are cycled and are continuously exposed to different environments how well matched will they be after a year or 2 years?
And will those differences self-regulate and you have a feedback loop that keeps the batteries relatively matched?
Or will the feedback loop cause them to diverge more and more until one or more batteries are no longer "healthy"?

The last two questions actually aren't rhetorical - I honestly don't know what happens when you have lead-acid batteries in this type of configuration. But I'd be very concerned about the latter question having an answer of "yes".

[santaclaws]

What happens when battery V1 is 5 degrees warmer than battery V4 because V4 is next to the wall and V1 isn't?

R1 is decreasing, self-ballancing stills remain.

What happens when you add resistance between R1 and R2, between R2 and R3, between R3 and R4, between R4 and load?
And same for the bottom - what happens when you change that ideal wire into a more realistic model with resistances?

What's the point with those resistances between R1 and R2 and so on?? R1-R4 are the string equivalent resistances. They include (in that equivalent circuit) the wiring resitances, too (which are of fixed values and identical for all the strings, btw). The (Vbank) represents the busbar (the single point where you connect all the strings) hence there's no other R1-to-R2 resistances in this circuit.

Btw, the model above is realistic - that means it includes in those Rx the whole resistance path of a string (the battery electric and ionic resistances and the wiring resistance).

The currents for the different batteries will be different - probably only slightly different when the batteries are new.
But as the batteries are cycled and are continuously exposed to different environments how well matched will they be after a year or 2 years?
And will those differences self-regulate and you have a feedback loop that keeps the batteries relatively matched?
Or will the feedback loop cause them to diverge more and more until one or more batteries are no longer "healthy"?

The only "feedback" needed is an equalizing charge.

I honestly don't know what happens when you have lead-acid batteries in this type of configuration.

Nothing happens. They just work as intended. FWIW, I have such a configuration for about 2 years and a half now and my only problem occured these days, due to the cheap & junky batteries I'm using (maybe they reached the end of life already).


PS: It's funny that you keep asking "what happens"! Why don't you tell me what's happening? Hint: make use of Kirchhoff's laws.

[Sunking]

That's all you have to say in your defense?! So I'm an electronics engineer, too.. what's the point??

You do not know power and batteries or the physics. I am not denying physics, you are. Or I should say you do not know the physics of a battery. I set on IEEE Battery Committe and have worked with every major battery manufacture for 40 years and writing standards.

Your assumptions are wrong and you are missing some very important facts jflorey tried pointing out that you did not catch. Your first mistake is you are assuming the wire and connection resistances are equal, they are not. You also have completely skipped over the most important characteristic of any battery, the internal resistance which varies by a magnitude of 3. You have completely ignored that, or IMO do not know.

When you take that into consideration the whole picture changes. It means every string charges and discharges at a different rate. One string with the lowest overall resistance does most of the work charging and discharging. That string will be the first to fail and fail prematurely.

The other point you are missing is it takes 12 to 24 hours for batteries to Equalize. That is fine in a Emergency Standby system that rarely ever gets used and spends it life on FLOAT. Not the case with Solar where you cycle every stinking day. With each day the batteries become more and more unbalanced. All of that is solved very easily with one series string.

You maybe an electronics engineer, but you are not a power engineer with 40 years experience working with batteries and the manufactures. You can parallel Pb batteries and it works great. The catch is it is done with single 2-volt cells. That is the big Reason EV manufactures do it that way. Otherwise they would go bankrupt from warranty claims from failed batteries. It also allows the owner to get home if a cell fails.

[Mike90250]

That's it, when the batteries are used in a S+P configuration, there will be a self-ballancing mechanism (ohm low) as to get equal internal resistance for each string (which roughly means equal SOC).

can you show me any Pb-H2So4 battery mfg site that proclaims their batteries in parallel are self-balancing ? Or shows how to wire it up that way ?
There are WAY TOO MANY variables internally to the battery and external, for this to be remotely possible AFIK.

[santaclaws]

Your first mistake is you are assuming the wire and connection resistances are equal, they are not. You also have completely skipped over the most important characteristic of any battery, the internal resistance which varies by a magnitude of 3. You have completely ignored that, or IMO do not know.

What do you think those R1-R4 are representing? As I've explained in the post above, R1-R4 include the total string resistance: wiring resistances, cells electric resistances and cells ionic resistances.

When you take that into consideration the whole picture changes. It means every string charges and discharges at a different rate. One string with the lowest overall resistance does most of the work charging and discharging.

If by "most of the work" you mean 2-3%, you're right. If you work in the battery industry, please tell me what's the dispersion rate of the internal resistance for a specific model?.. is it 2-3% or so?

The other point you are missing is it takes 12 to 24 hours for batteries to Equalize.

The equalizing charge is necessary for any configuration.

You maybe an electronics engineer, but you are not a power engineer with 40 years experience working with batteries and the manufactures.

It happens to know "electronics engineers" with lifetime experience who never used a soldering iron(!) nor they know how to apply the Kirchhoff's laws. For that matter, I'm still waiting for you to include more scientific facts in your explanation rather than "practical"/experimental facts. I've made a diagram (which is for real) thus anyone could do the math. Now show me the math that proves that one string could go wild (up or down). We're not talking about insufficient charging power and such, which will affect any battery on any configuration.

[santaclaws]

can you show me any Pb-H2So4 battery mfg site that proclaims their batteries in parallel are self-balancing ? Or shows how to wire it up that way ?
There are WAY TOO MANY variables internally to the battery and external, for this to be remotely possible AFIK.

We have that schematic. We have those equivalent string resistances and the Kirchhoff's laws. What in the world of physics could keep those strings unballanced?

[Mike90250]

How could one keep those strings unballanced

still no battery mfg recommendation or any professional recommendation.

It's the batteries !

The wire is a fixed resistance. The crimp lug is a resistance (fixed if it's a good crimp). The torqued connection joints, with silver grease are a fixed resistance. The batteries are the wild card. They change daily, with age and temperature. 0.00002 ohms is not much, till you start cranking 80 amps, then you get weird current imbalances in the battery network and the best battery suffers till it's the 2nd best, and so on and so on.

But hey.. you are a clever monkey and have it all figured out, so we are just going to let you go to battery hell on your own, But we have tried to warn you.

Maybe you have not been to the smartguage site - try it http://www.smartgauge.co.uk/batt_con.html

[Sunking]

Mike he cannot do math or apply Kirchhoff Law correctly. He has not figured out battery Internal resistance magnitude varies by a factor of 3, and that it takes batteries 12 to 24 hours to charge and reach equilibrium. Not possible with daily cycle on solar charger. Let him keep destroying batteries. No better teacher than loosing lots of money.

He thinks there can only be a 2 or 3% difference which is a fools game when it can be as high as 3:1 magnitude. That is what happens when 1 string is .005 Ohms and another is .015 Ohms. If he knew the math and applied it correctly with all the other variables, he would quickly learn a load of 100 amps splits to 66 and 34 amps.

Some day he will pull his head out of his arse, listen to the pros, and end up buying the proper size single string battery which is less expensive than multiple strings, and will last 4 to 7 years if he buys quality. Hell he does not even understand the danger of high current yet. I bet he does not even EQ his batteries every couple of weeks on a generator or commercial AC. I do not think he knows solar cannot EQ a battery.

[jflorey2]

During normal operation, the flowing currents through each string would be more complex (dynamically changing in and out) but you can't have large variations in individual string currents as long as all the batteries are in a healthy condition and, obviously, they are of the same type.

Sure you can - and your diagram shows why.

Let's say for some reason string 1 has a slightly lower state of charge than the others. It can be for any reason - manufacturing tolerances, dirt on the battery case, someone at the factory dropped a wrench then yanked it off before any permanent damage was done. It's at 99% charge while the rest are at 100%. That's practically identical - right?

Now you discharge. You get down to 49% (string 1) and 50% (string 2 3 4.) No worries.

Now you charge at 100 amps. Since string 1 is at a lower state of charge, R1 is larger than R2 R3 R4. So string 1 gets charged at 22 amps, strings 2 3 and 4 get charged at 26 amps. You charge to what you think is 100%.

Now - nighttime! Charging stops. String 1 is charged to 95% (because it got less current) and the rest of the strings are charged to 100%.

Now the nighttime discharge starts. String 1 gets discharged down to 45% and the rest discharge to 50%.

Now daytime. Charging starts. R1 is even larger. String 1 gets even less charge current.

And the imbalance continues to grow.

And that is why you don't parallel lots of strings in cyclic applications.

[santaclaws]

The math is even simpler: no string could ever discharge lower than the bank voltage.

When it reaches this threshold, it stops supplying current to the load and eventually get charged by the rest of the bank.

That's it, you cannot generate an outgoing current from a lower EMF (voltage) source (Ohm law).

[foo1bar]

he (Vbank) represents the busbar (the single point where you connect all the strings) hence there's no other R1-to-R2 resistances in this circuit.

That busbar has resistance
So there is resistance between each connection point to the batteries *as well as* resistance within each string.

Btw, the model above is realistic - that means it includes in those Rx the whole resistance path of a string (the battery electric and ionic resistances and the wiring resistance).

It doesn't include resistances for the connections between the strings.

Also because those resistances and voltages are varying over time the diagram could be misleading to people who look at it.
I assume you know they are, I know they are, but I can definitely see someone not realizing it.

PS: It's funny that you keep asking "what happens"! Why don't you tell me what's happening? Hint: make use of Kirchhoff's laws.

Well what happens when you add resistances between the strings is you get different voltages across that wire that you represented as a node.
Different voltages means charging/discharging is not equal across them. And the deeper you cycle the battery the quicker it will wear out. So after 2 years you might have one battery that's at 50% of it's life while another is at 30%.

I believe temperature variations could give a similar result - that V1 could be charging/discharging more than V4 and therefore would wear out faster.


Kirchoff's laws aren't the only thing here - you also have to know what happens in the physical aspects of the battery.
Lets say R2 is larger just because of variation in temperature, variation in manufacturing of the battery or whatever.
When charging if V2 is equal to V1, less curent will be put into battery V2.
So for that moment V2 gets less charge, but you say "Hey, V1 will get more charge, it'll become higher than V2 and then more current will go to V2"
And if that's what happens it would be a feedback loop that should self-regulate the batteries.
BUT lets say V2 has this happen constantly for a year?
Are there mechanisms in the physical part of the battery that could result in R2 increasing over time MORE than R1 increases?
(ex. could more sulfation occur on one battery and that increased sulfation causes it to have even more sulfation, and so on?)

I understand you may be reluctant to listen to a poster who is arrogant and antagonistic and IMO generally an *ss.
But I think that you and I don't have the background in battery chemistry/behavior that others do and therefore should try to understand the viewpoint of people who do have that background.
And in this case I think the "parallel batteries should be avoided" advice/viewpoint has reasons that justify it.
And you are even providing anecdotal evidence that a situation where a battery pack made up of one string probably would have lasted 5 years instead was done as parallel ones and has major problems (1/4 failed and shortened life on other 3/4) after 2 years

[santaclaws]

If you connect all the strings in a single point (as I do) the busbar resistance has no influence. Its resistance simply adds to load (or charger) resistance.

Now I have to ask, once again: does everyone agree that (during discharging), no string EMF voltage would go bellow the bank voltage level?

Now if you take care of not allowing the bank voltage to go bellow 24V, by example, you could be sure that no string will go bellow that level, EVER.

PS: You're perfectly right, those string equivalent resistances are dynamically changing but, at the end of the day (so to speak), no string will be discharged bellow bank voltage, no matter what.

By the way, it's not only the internal resistance that is changing but the EMF voltage, too.

[santaclaws]

Btw, the load current (during discharging) will be mostly supplied by the strings with the higher SOC (which have higher EMF voltages and lower internal resistances), not by the "abused" (aka lower SOC, lower EMF voltage and higher internal resistance) string.

(that's part of the self-ballancing process, too)

[Sunking]

I suppose you need to talk to about 64 engineers that make up IEEE Battery committee. It is composed by 1/2 from battery manufactures engineers and scientist and educate them. I am sure it will be a fun experience for 64 people. Not so fun you.

[foo1bar]

Btw, the load current (during discharging) will be mostly supplied by the strings with the higher SOC (which have higher EMF voltages and lower internal resistances), not by the "abused" (aka lower SOC, lower EMF voltage and higher internal resistance) string.

(that's part of the self-ballancing process, too)

*IF* the batteries are behaving in an ideal manner and your wiring is perfectly matched the self-balancing occurs and the one with the lower SoC provides less power.

But in real life you're going to have sulfate crystals forming more on some than others.
And you'll have one string has a better connection (lower resistance) than another (both external wiring AND internal to the packaged battery)
And you'll have temperature differentials.

So it will be quite possible for one battery to go through deeper cycling than others.

at the end of the day (so to speak), no string will be discharged bellow bank voltage, no matter what.

Not sure what point you're trying to make.
But it would be possible for one string's battery to be lower if you disconnect them all - you have proof of it in the first post of this thread.
While connected the whole string was obviously the same voltage as any other route between those same two points in the circuit.
But once it was disconnected and current no longer flowing through the wires and internal resistance it was a lower voltage.