Mixing Battery types in 1 Bank

[goboatingnow]

The issue being debated seems to be on recharge

both batteries will essentially recharge together its only in absorption mode that one battery make take longer to finish leaving the other to ( potentially ) fizz.

what happens is during bulk , the voltage will only rise as both batteries recover SOC. ( this assumes there is isn't a massive capacity differential ) This is exactly the same as when you recharge your battery with a load on the system, the system loads don't suddenly stop just because the battery is low and sucks all the current from the charger.


in practice during recharge , based on dynamic inout resistance of the battery , current can flow into or out of both batteries at various stages, both batteries will arrive at the absorption voltage together ( they have to ) essentially the batteries will load balance as they charge up. however at absorption mode things are slightly different and the battery that completes absorption mode first , will be left exposed to longer absorption cycles then necessary.

This is easily seen in modern chargers that are supporting system loads while charging. absorption goes on too long in these cases and a safety timer is usually deployed.


Again think of the dual tanks being refilled by their combined output pipe. the input flow is distributed amongst the tanks in proportion. Only in a dynamic state where large currents where impressed would the bigger tank over flow before the small tank filled.

You have to be careful to separate the boundary conditions because batteries unlike tanks are non-linear.


remember the sum of two remittances in parallel is always less then each of the constituent resistors. Some current always flows into the other resistor. but don't forget that as a battery charges its equivalent resistance rises, hence charging current is " directed " into the other loop, causing it to "catch" up.

( sorry about the simple analogies )


dave

[colemj]

Originally Posted by TeddyDiver

Mark, if the voltage is the same and resistance differs, which route the current mostly goes?

I don't fully understand your question, but i=v/r, so the current either goes up or down depending on the direction of change of r.

In charging/discharging, resistance increases/decreases as current is supplied at a constant voltage. For charging two batteries with different internal resistances, the lower R battery will take more current and its resistance will increase to the point where the combined voltage does not let it take any more current until the other battery has reached that transient point of internal resistance. At that time, the "hungry" battery can pig out on current again.

But it does not run up to full charge and sit there boiling waiting for the other to catch up.

It is a long series of micro-catchups all the way up to achieving full SOC at about the same time. Think of it similar to how a sine wave is reproduced from digital switching. Lots of little step functions that become almost indistinguishable from a continuum.

Mark

[goboatingnow]

Originally Posted by colemj

If that last paragraph is wrong about charging, then the discharging part is wrong also.

Again, I have always stated the instantaneous case of charge/discharge differences between the two batteries. I have agreed that at any given point in time, they are behaving this way, but stressed that they need to be understood in terms of an equivalent circuit. I have also described experiments that would result in this case, but in those experiments they would no longer be in an equivalent circuit.

What I don't agree on is that at the end of the load/charge period, the batteries are sitting with different SOC's. This is real-life usage - fractional C loads are applied over time and fractional C charging occurs at least up to absorption voltage.

I don't know anyone with unity C or greater charging systems in real-life cruising (we can generate 330A of charging, and that is only 0.5C of our modest bank). I also do not know anyone who generates unity C or greater loads.

In SC's example above, there is no way a charger could/would supply 500A to a combined bank. It may momentarily supply a large current to the AGM, but that battery's internal resistance would build to the point that the current acceptance would be low until the internal resistance of the FLA built up enough to allow the combined bank to increase in voltage.

Is this last paragraph wrong? Is the one above it wrong?

If they are, then my understanding is wrong. If they are not, then I am correct.

Mark

Look guys, in a steady state system with fractional charging ( and discharging) , the simple case is the batteries will charge up together and discharge together. This is assuming that the charge cycle completes and is not terminated before each battery reaches in 100% ( which will not be the same , as the absorption time will be different )

what is the case is that with different chemistries and discharge curves , at any point in the discharge curve, where you to stop and examine SOC they would not be the same. they would only be the same if they had identical discharge curves. SOC will deplete at rates that are a function of the particular batteries discharge curve. i.e. the current will aggregate up to the sum of each battery , but it does not mean that at any one time , the current is a fixed percentage, in fact scenarios can exist where one battery is charging of the discharging battery. but ultimately if allowed to discharge they all reach the bottom ( and the top) together.

so the issue is more to do with partial discharge and recharge cycles , which unfortunately are very common. In that case in practice you WILL get a lazy battery and you will get one battery that is not cycling to full ( or both actually)

of course that in itself isn't a problem, but various partial recharges kill various LA technology in different ways. AGNM is more prone, even though in this scearanio its getting a better partial charge !


The exact same problem occurs in Li, which is why we have bottom and top balancing. ( we don't care that much in LA, because they can withstand reported slight overcharging so in effect , we always top balance LA)


As has been suggested over on the Li thread , the same thing happens when you parallel AGMs with Li. once you don't exceed any specs, both systems co-operate ( no difference than multiple charge sources in parallel "co-operate").

dave

[colemj]

Originally Posted by goboatingnow

The issue being debated seems to be on recharge

both batteries will essentially recharge together its only in absorption mode that one battery make take longer to finish leaving the other to ( potentially ) fizz.

what happens is during bulk , the voltage will only rise as both batteries recover SOC. ( this assumes there is isn't a massive capacity differential ) This is exactly the same as when you recharge your battery with a load on the system, the system loads don't suddenly stop just because the battery is low and sucks all the current from the charger.

in practice during recharge , based on dynamic inout resistance of the battery , current can flow into or out of both batteries at various stages, both batteries will arrive at the absorption voltage together ( they have to ) essentially the batteries will load balance as they charge up. however at absorption mode things are slightly different and the battery that completes absorption mode first , will be left exposed to longer absorption cycles then necessary.

This is easily seen in modern chargers that are supporting system loads while charging. absorption goes on too long in these cases and a safety timer is usually deployed.

Again think of the dual tanks being refilled by their combined output pipe. the input flow is distributed amongst the tanks in proportion. Only in a dynamic state where large currents where impressed would the bigger tank over flow before the small tank filled.

You have to be careful to separate the boundary conditions because batteries unlike tanks are non-linear.

remember the sum of two remittances in parallel is always less then each of the constituent resistors. Some current always flows into the other resistor. but don't forget that as a battery charges its equivalent resistance rises, hence charging current is " directed " into the other loop, causing it to "catch" up.

Actually, discharge has been the debate too. I don't see the practical difference between the two.

The opposite side of the debate from mine has been that during discharge, the AGM battery will run itself dead supplying 6X more current to a load a long time before the FLA depletes. It seemingly does this because of its lower internal resistance. If it can do this, then it can also run up to absorption charge well before the FLA does for the same reason. No?

The fizzing part of absorption is highly dependent on a lot of things. I don't think one should be counting on this happening. We have been discussing equal sized batteries with identical absorption voltage points. I have conceded that batteries with large capacity and charging point differences should not be combined.

Given that, what do you think the chances are of fizzing the above AGM before the above FLA (assuming a good charger)? How long do you estimate an AGM would be at full charge waiting for a FLA to catch up?

Given that we have also discussed real-life charging to less than 100%, what are the chances of fizzing ever?

If this bank were in real-life usage, being discharged to 50% and recharged to 80-90% (low fractional C for both), how much would you estimate the lifespan of the bank would be shortened over having all AGM? How about all FLA?

Mark

[colemj]

Originally Posted by goboatingnow

Look guys, in a steady state system with fractional charging ( and discharging) , the simple case is the batteries will charge up together and discharge together. This is assuming that the charge cycle completes and is not terminated before each battery reaches in 100% ( which will not be the same , as the absorption time will be different )

Yes, I have been saying this for several pages now.

Originally Posted by goboatingnow

what is the case is that with different chemistries and discharge curves , at any point in the discharge curve, where you to stop and examine SOC they would not be the same. they would only be the same if they had identical discharge curves. SOC will deplete at rates that are a function of the particular batteries discharge curve. i.e. the current will aggregate up to the sum of each battery , but it does not mean that at any one time , the current is a fixed percentage, in fact scenarios can exist where one battery is charging of the discharging battery. but ultimately if allowed to discharge they all reach the bottom ( and the top) together.

so the issue is more to do with partial discharge and recharge cycles , which unfortunately are very common. In that case in practice you WILL get a lazy battery and you will get one battery that is not cycling to full ( or both actually)

of course that in itself isn't a problem, but various partial recharges kill various LA technology in different ways. AGNM is more prone, even though in this scearanio its getting a better partial charge !

I think you just said that the SOC of the two batteries combined in a resting bank will only be the same if they are completely discharged or completely charged? In other words, they are only the same nominal SOC at 0% and 100%? Let's say you do a partial discharge of this bank to 12.4V during the day and it is now time to recharge. What is the relative SOC of each battery at this point? Surely, the SOC of each battery is close to the same even before complete depletion?

Partial charging, like is common, creates lazy batteries that do not cycle to full regardless of type and mixture. By definition of "partial charge" BOTH batteries in a mixed bank will not be full. Unless you are defining "partial charge" to be 95><100%.

BTW, earlier you said it would be the FLA that suffered...

Originally Posted by goboatingnow

The exact same problem occurs in Li, which is why we have bottom and top balancing. ( we don't care that much in LA, because they can withstand reported slight overcharging so in effect , we always top balance LA)

I don't think the Li chemistry is a good analogy here. First, it is the same chemistry. Second, you are concerned with the physical phenomenon of voltage knees - which do not even exist with LA. Third, that balancing done on both ends is to prevent one battery from catastrophically (and instantly) damaging the other battery during charge or discharge. This is not even close to the case in the AGM/FLA example.

Mark

[hellosailor]

Sailorchic-
"AGM has a peak charge rate of C*5. So a 100 amp FLA can take 33 amps without damage. A 100 amp AGM can take a 500 amp charge current with no overheating or out gassing."
Where do you get that from?
I think it was JCI that told me wet leads could safely be charged at C/5 while AGMs could take C/4 but nowhere anything near C*5. That would be one hell of a fast charge and justification for a whole different breed of charging system.

And in any case, a "safe" charging current also depends very much on the voltage being applied at the time, not just the current itself.

[TeddyDiver]

Not all the FLA's being the same nor AGM's either so that using different brand or age but same type batteries will cause same problems sooner or later prematurely killing the bank. I'd be cautious even connecting paralled any batteries even identical ones if there's any reasonable way to avoid it.

Recharge:
To be realistic lets assume using 80 to 200 A current?

[donradcliffe]

I wrote a little program to calculate the discharge from a 200 amp-hr battery bank where 100 amp-hrs is FLA and 100 Amp-hrs is AGM. The program uses the internal resistance provided by Sailorchic, and a linear model of internal voltage vs SOC. I started with full batteries, and held the total bank discharge rate steady at about 13 amps. The results seem pretty reasonable--the AGM starts by taking nearly all the load, but by the end of the second hour, the FLA is taking over half the load. At the end of the 10th hour, the terminal voltage is down to 12.01 and the total amp-hrs taken from the bank is 149, with 67 coming from the FLA and 82 coming from the AGM--not ideal, but not the end of the world. Next time I have some spare time, I'll do the charging scenario.

hour amps-FLA amps AGM
0 -2.5 15.2
1 5.1 8.6
2 6.8 6.7
3 7.3 7.6
4 7.5 5.8
5 7.1 6.3
6 7.2 6.8
7 7.2 5.8
8 7.0 6.4
9 7.1 6.8
10 7.1 5.9

[TeddyDiver]

Originally Posted by colemj

I think you just said that the SOC of the two batteries combined in a resting bank will only be the same if they are completely discharged or completely charged? In other words, they are only the same nominal SOC at 0% and 100%? Let's say you do a partial discharge of this bank to 12.4V during the day and it is now time to recharge. What is the relative SOC of each battery at this point? Surely, the SOC of each battery is close to the same even before complete depletion?

Nope, the battery with lower resistance will deplete faster. When the loading grows even more and when lighterly loaded they might balance some higher SOC depleting to lower SOC.

[sailorchic34]

Originally Posted by hellosailor

Sailorchic-
"AGM has a peak charge rate of C*5. So a 100 amp FLA can take 33 amps without damage. A 100 amp AGM can take a 500 amp charge current with no overheating or out gassing."
Where do you get that from?
I think it was JCI that told me wet leads could safely be charged at C/5 while AGMs could take C/4 but nowhere anything near C*5. That would be one hell of a fast charge and justification for a whole different breed of charging system.

And in any case, a "safe" charging current also depends very much on the voltage being applied at the time, not just the current itself.

I got the 500 amp charge rate from lifelines site. another site referenced a max of c/3 for LA

[sailorchic34]

In my mind, during charging, the agm must be taking more amps the the fla, even in bulk. As agm's tend to not like overcharging, I see an earily death of the agm in a mixed battery bank. I think that I can agree that at 100% and 0% the soc will be the same. But as we try to use no more then 50% soc under normal conditions, the agm will arrive there first.

Remember that the internal resistance on the agm will always be much lower then fla. The interactions are a bit more complicated, but in my simple mind, the agm is going to be flogged in a mixed bank.

With all the batteries being the same, well mostly the same you don't have the imbalances that a mixed bank would have. Similar to mixing old and new batteries. The older batteries have higher internal resistance and tend to cause the new batteries to age faster.

[colemj]

Originally Posted by donradcliffe

I wrote a little program to calculate the discharge from a 200 amp-hr battery bank where 100 amp-hrs is FLA and 100 Amp-hrs is AGM. The program uses the internal resistance provided by Sailorchic, and a linear model of internal voltage vs SOC. I started with full batteries, and held the total bank discharge rate steady at about 13 amps. The results seem pretty reasonable--the AGM starts by taking nearly all the load, but by the end of the second hour, the FLA is taking over half the load. At the end of the 10th hour, the terminal voltage is down to 12.01 and the total amp-hrs taken from the bank is 149, with 67 coming from the FLA and 82 coming from the AGM--not ideal, but not the end of the world. Next time I have some spare time, I'll do the charging scenario.

hour amps-FLA amps AGM
0 -2.5 15.2
1 5.1 8.6
2 6.8 6.7
3 7.3 7.6
4 7.5 5.8
5 7.1 6.3
6 7.2 6.8
7 7.2 5.8
8 7.0 6.4
9 7.1 6.8
10 7.1 5.9

So to summarize, there is theoretically at most a 5% difference between the two SOC's when the bank itself is at a reasonably low SOC?

BTW, I haven't been able to confirm SC's internal resistance numbers. They are remarkably hard to find, and few manufacturers publish them. I have found one reference in a physics paper for FLA of 20mOhm, which she also lists. For AGM's, I have found 5mOhm from PowerSonic and Century and 9mOhm from Eaton. Lifeline just gives a blanket statement that their AGM's are 5x less internal resistance than FLA's without giving any of the numbers supporting that statement.

So it looks like the real numbers to use are somewhere between 5X and 2X difference between the two - 4X seems like the best conclusion. Does this change your analysis much?

Mark

[hellosailor]

Ah, LIFELINE.

Bear in mind that you cannot compare a Lifeline AGM to ANY other AGM battery. Lifeline does not make "AGM" batteries, they make military.aviation AGM batteries.

These are designed to withstand the pressure change involved with repeatedly going to 30-40,000 feet and back to sea level again, without bursting or venting the case. A conventional AGM battery, or any "sea level-to-Denver" battery would be ruined by one cycle with that big a pressure difference.

Which is probably why Lifeline AGMs can and should be equalized--which would, again, cause all conventional AGM batteries to be ruined.

Lifeline is the unicorn ranch.

[colemj]

Originally Posted by TeddyDiver

Nope, the battery with lower resistance will deplete faster. When the loading grows even more and when lighterly loaded they might balance some higher SOC depleting to lower SOC.

You will excuse me if I wait for Dave to answer the question I asked him about what he meant to say in his statement.

I am having a hard time comprehending the rest of your statement (after the first sentence). Perhaps you could rewrite it more clearly?

Mark

[colemj]

Originally Posted by sailorchic34

In my mind, during charging, the agm must be taking more amps the the fla, even in bulk. As agm's tend to not like overcharging, I see an earily death of the agm in a mixed battery bank. I think that I can agree that at 100% and 0% the soc will be the same. But as we try to use no more then 50% soc under normal conditions, the agm will arrive there first.

Don's discharge analysis shows the AGM's supplying considerably more amps for the first hour, so I would expect that they would also consume more amps during charging - for the first hour or so of bulk, after which they behave similarly to the discharge data.

Don's analysis also showed that the AGM did not arrive at 50% SOC first. Well, OK, it arrived there at the same time, only at 5% less SOC. However, Don did not show what happens 10 minutes after that discharge was removed. I bet that 5% disappears…

Since the discharge analysis did not show that the AGM's arrived at a low SOC way before the FLA, I suspect this will also be true for arrival at high SOC.

There is little chance that an AGM can become "overcharged" (plates shedding). What we are really talking here is boiling or gassing of the electrolyte to the extent it causes the regulation valve to open and the electrolyte be lost.

The question is: will the AGM have to sit at 14.4V for such a long time waiting for the FLA to stop taking current that it will boil and gas?

My bet is that it does not.

Mark