12V AGM batteries in series- voltage issues

lazza · September 2015

HI Forum

We recently switched from normal FLA 12V batteries in smaller (cheaper) systems to the equivalent 12V batteries in AGM (http://www.damiasolar.com/productos/bateria_solar/bateria-solar-ciclo-profundo-u-power-agm-250ah_da0495_36). This was principally to avoid the problems we've been having with normal 12V FLA, ie clients not checking water levels, large voltage drops with high power equipment like pumps etc

However, i've noticed that when put in series, either for 24V or 48V systems- as soon as the voltage reaches over say 27V in 24V systems, the difference in voltage been the batteries can be quite concerning. For example- at 27V or one battery may read 13.3V and the other 13.7V! They supposedly come equally charged from the factory.

The trouble is most of these systems have no way of correctly this imbalance once established. We have been putting in "battery balancers" (http://www.victronenergy.com/battery-monitors/battery-balancer) to try to correct with limited results.

We did have similar problems with FLA 12V batteries but it was MUCH less pronounced. Is there any easy explanation/solution for this??

Many thanks
Larry

vtmaps · September 2015

Perhaps you need to equalize them (without the battery balancer). I'm not sure what the safe EQ voltage is for your batteries... for some AGM batteries an EQ is just an extra long absorb. What you are attempting to do with an EQ is overcharge most of the cells while waiting for the slowest charging cell to catch up.

There are other possibilities... perhaps the mismatched voltages you are seeing are caused by a poor connection or crimp somewhere in the battery cables.

Another possibility is that what you are seeing is normal variability for that particular lot of batteries. Half the cells are above average and half are below average. The life and capacity of the battery is determined by its weakest cell. As long as your particular weakest cell meets warranty, the whole battery will meet warranty.

--vtMaps

softdown · September 2015

What size cable are you using for your battery to battery connections?

BB. · September 2015

Clip a12 volt light bulb (like brake light bulb) and clip it across the battery with the high voltage.

This will discharge the high battery and let some current bypass through the lamp to help bring up the low battery.

-Bill

waynefromnscanada · September 2015

BB. wrote: »

Clip a12 volt light bulb (like brake light bulb) and clip it across the battery with the high voltage.

This will discharge the high battery and let some current bypass through the lamp to help bring up the low battery.

-Bill

I'm doing this now with one of my series 2V L-16 batteries with high SG.

lazza · September 2015

HI

Thanks Bill, that never occurred to me. I'll give it a go.

As for equalization. The trouble is that these systems dont have an equalization option in the charger. What I could do is increase the absorption voltage to say 30V for 24V system for a while and then bring it back down again after a few days. My concern is that the difference in voltages between the 2 batteries may get so great that it damages the higher battery. For example if an AGM reaches 15.5V.. that's bad isnt it??

waynefromnscanada · September 2015

lazza wrote: »

As for equalization. For example if an AGM reaches 15.5V.. that's bad isn't it??

My first impression - - OUCH! Way to destroy an AGM !
But I'll leave the final word to BB.

BB. · September 2015

I am not a battery engineer--I would try to contact the mfg. and get their take on the issue. AGMs are supposed to have low self discharge current and be (on average) much more balanced between cells+batteries than flooded cell.

If they started out the same voltage and are drifting apart--I would suspect a bad cell/battery.

But if they started out different, then they may not have been equally charged to start with--And since you really cannot high voltage equalize most AGM batteries, you are stuck with trying to bring them back into balance--Separate charger, or using the lamp discharge/bypass trick.

Certainly, the longer they operate this way--The faster one of them is going to die... Taking a guess, you should be seeing a maximum of 0.015 to 0.030 volts difference between cells (6 cells * 0.03 per cell = 0.18 volts on a 12 volt battery). At least that is what I would expect from flooded cell chemistry and the suggestion to equalize when you have an SG difference of 0.015 to 0.030 units difference between high/low cells) (the basic SG to voltage equation assumes that change in SG unit reading = Voltage Change of cell).

-Bill

zoneblue · September 2015

You are probably over reacting. Any series battery comprised of cells will have cell voltage variance. Inside the 12v battery there are 6 cells, and they too will have a cell variance. Theres not a huge amount of production consistency (or binning) in the lead acid industry, because its an on old technology that traditionally was pretty tolerant of charging conditions.

So long as the cells are all from the same batch, purchased at the same time, with no prior use history, then youve done all you reasonably can. You can get BMS systems that balance indivdual cells, but it would be a waste of time on a 12V battery, as you cant get at the cells.

Note that the voltage difference will be least at rest, and greatest during absorb. Its one of the downsides to AGM, different to traditional FLA tolerance of overcharge, in that AGMs cant have water added.

Note also that the longer the string of cells, the worse this problem gets. With 2 12V batterys i really wouldnt lose any sleep over it.

PNjunction · September 2015

lazza wrote: »

However, i've noticed that when put in series, either for 24V or 48V systems- as soon as the voltage reaches over say 27V in 24V systems, the difference in voltage been the batteries can be quite concerning. For example- at 27V or one battery may read 13.3V and the other 13.7V! They supposedly come equally charged from the factory.

The question is, are they roughly from the same manufacturing date? Are some much older / newer than the others? What was the open-cell voltage of each when you received them?

I notice you have only solar. Do you have *any* charger, even just a 12v one to allow each battery to receive a truly full initial charge, or are you only relying on the distributor? The first charges on an agm are especially important, and if you can do it to each 12v battery individually, so much the better at the outset.

Perhaps a 30A Samlex model or similar would be nice:
http://samlexamerica.com/products/Pr...t.aspx?cid=S14

I use the SEC 1215ul model myself for smaller batteries, and like how you can set the dip switches to immediately put it into just a power-supply / float of 13.5v, or it can be set to do a 2 or even 3 stage charge. YOU have to supply your own clamp wiring. Great docs!

Heck, if you are near fully charged anyway, but just want to make sure that each one has gone through the same amount of absorb, then even the 1215 might do.

You may want to just put them on a very long float initially, or perhaps if you have never had a charger on them, use something like this on each one in the 3-stage routine, and try again.

lazza · September 2015

HI

Thanks for the replies. Yes the batteries are all of the exactly the same batch-all series of batteries have equal cable runs and i'm confident with the crimping and connectors. Assuming that they all come equally charged, we havent been testing the voltages beforehand- we will from now on. It seems that there are two options:

1. We charge each battery individually before installation to ensure full charge of each. This is somewhat of a pain.
2, We use the current bypass mechanism suggested by Bill to drain the higher voltage battery a bit in situ before finalising installation.

How many decimals of difference per battery would you consider normal at an absoprtion voltage of say 28.4V??

cheers
Larry

BB. · September 2015

Since you are a professional organization--The Light Bulb is probably not going to fly very well (other than as an experiment back at the shop).

One way to go would be to get a few DC to DC converters. Here is one (family has 24:12, 36:12, and 48:12 volt) DC converters available. Get the 24:12 volt model and clip it across two 12 volt batteries. The converter works bi-directionally, it will take excess voltage (energy) from the high battery and pipe it to the low battery. The DC converter just tries to make the two voltages equal. 20 Amp capacity:

http://www.solarconverters.com/index...58-eq-12-24-20

Would be a much more professional looking unit and since is automatic--You cannot get into trouble leaving it unattended. These devices probably are not cheap--And, in the long run, I wounder if designing/building your own may not be a cost effective alternative. Or buy 5 of them and call it a day.

How far is the voltage spread during absorb--I am not sure. We measure resting voltage to estimate battery state of charge. I would hope that matched cells (brand/model/date/temperature) would be pretty close (call it less than 0.1 volt per cell difference or 0.6 volts for a 12 volt battery bank)--But the issue is that few modern sealed batteries give you access to per cell voltage... That 0.6 volt spread could be a 0.1 volt per cell difference, or it could be a single 0.6 volts across one cell (definitely a problem cell).

A net tool I would also be tempted to buy (still tempted, but don't have any way to justify it as anything but as a toy)--A (relatively) cheap FLIR camera for a cell phone.

http://www.amazon.com/s/ref=nb_sb_ss...id%2Caps%2C353

I would use it on AGMs when charging... Remember that each cell has a small Palladium (or similar) catalyst to convert hydrogen+oxygen gasses back to water. And this generates heat. I would suspect that a decent FLIR cell phone camera would identify hot/cold cells/batteries during installation/acceptance testing.

Otherwise, just get a standard AC to DC or DC to DC battery charger at XX Volts--Use it to "charge" the low battery in the bank. Yes, not efficient, but at least automatic and a good temporary solution.

-Bill

zoneblue · September 2015

To repeat, during absorb you will always get significant differences. However you ought to have pretty consistant rest voltages to within 50mV per cell. Thats becasue after a decent float (several hours beyond absorb termination) the cells will top balance.

Essentially the higher SOC cells take more voltage. Batteries are rated usually by SOC variance and Ir difference. Its predominantly the latter that will determine which cells reach absorb first, and thus take a higher voltage.

On our system, i might see 25mV variance in rest voltage, but during absorb this might go as high as 0.2V. Per cell.

If you have larger numbers of these batteries you might try binning them yourself. That is measure SOC and Ir and pair them to their best mate.

Also be sure to give the batteries some plate formation cycles, before doing any of this.

PNjunction · September 2015

lazza wrote: »

1. We charge each battery individually before installation to ensure full charge of each. This is somewhat of a pain.

It may be a pain, but especially since we don't have access to individual cells, that first charge is extremely important. The more cells you have in series initially, it makes an easy hiding-place for a low-cell to never truly finish an absorb, while his surrounding buddies gas more than normal.

Sure you can hammer the whole higher-voltage bank in hopes of reaching that hidden cell, but it can take it's toll on the healthy ones. Essentially, a balance may be achieved solely by quickly aging the good ones.

On a smaller scale, I demonstrated this to myself when charging a purposely unbalanced 24v bank of 2v Cyclons.

I'm not sure if your agm's came with any "commissioning" instructions. For the higher quality agm's, some have instructions for either doing a long laborious float, OR a shorter period a extended-absorb, say keeping 14.4v for 12 hours or more initially - something most automatic chargers won't actually do.

I note that these are recommended for seasonal / weekend use, and may in fact just be conventional agm's more suited for ups duty than daily deep-cycling - even if only to 50% DOD. I have no idea the quality of the manufacturing, where a cell could indeed be very low compared to the others out of the box - although you would never know it. I would still suggest a good charge initially to help ensure that there are no hidden low cells.

2, We use the current bypass mechanism suggested by Bill to drain the higher voltage battery a bit in situ before finalising installation.

That won't cure a hidden low cell. My 2-cents (you know what that's worth!

says that you are starting out on the wrong foot by not ensuring each has completed a good charge individually at least ONCE.

lazza · September 2015

Thanks for the replies.

We assumed that the supplier would have given each battery a perfectly equal charge- not really our job as we are buying them new and in "perfect" condition. However, if you think that is the best option for us, we'll set up a system- i'll also check with the supplier exactly what charge they are giving the batteries before shipping.

I'm slightly concerned that as these AGMs are much cheaper than AGM used to be, although still dearer than FLA, that maybe the quality isnt what it should be. Have AGM prices dropped recently too in the United States? Is this a trend?

PNjunction · September 2015

Heh, most suppliers here will tell you anything to make a sale.

I also don't trust them to do a GOOD job of it. One battery shop will only charge to 13.8v and no higher - basically never going into absorb, but it makes the voltmeters happy on the sales desk just before you get there.

Tell you what, how about something simpler like these to make it a whole lot easier - perhaps the CTEK MXS 10 or perhaps the MXS 25? Just push the button to "snowflake" mode, and let it go to town:

http://www.ctek.com/es/es/chargers/12v

These should be pretty easily available in Spain...

waynefromnscanada · September 2015

I was pleased when I bought my last battery bank. They checked the EQ of each cell right at the factory door, in front of me, before loading on the truck.
Rightly or wrongly I assumed that proved each cell good. They all read 1.265 on the electronic EQ reader. I asked if they sold those there, he replied no, but they could order one in for me if I had $2500.00. LOL
So now I have a question for those who know more than I - - - - does 1.265 show a good cell on a brand new battery, or does it only show it was newly filled with electrolyte of that SG? BTW, I also had my FLUKE meter with me and checked the voltage of each 2 volt battery before loading. I no longer remember the exact voltage, but it indicated fully charged. Put perhaps they had only just come off a brief charge.

inetdog · September 2015

You are correct that an SG that matches the manufacturer's recommendation will not necessarily tell you anything about the capacity or condition of the battery.
It does tell you that it was recently fully charged OR had high SG solution instead of distilled water added to bring up the SG.
If the battery started life with proper electrolyte SG and had only water added over its life, then sulfation would decrease the measured SG at full charge.

But in any case, unless you know more about the history of the battery a good SG does not tell you anything about the remaining capacity. Only a proper constant current load test (not easy for DIY) can tell you that.

zoneblue · September 2015

All AGMs are not really created equal. How long a VRLA battery lasts depends in practice on two main things. How well is the battery sealed? Loss of pressure means loss of water. And how decent a catalyst they fitted? The catylyst reduces the pressure on the seal by assisting the oxygen and hydrogen to recombine.

Both of these cost money to do right, and at one end of the spectrum you have, no real seal or catalyst , maybe 2-3 year life, with a high pressure seal and palladium catylyst at 7-10 years.

12V AGM batteries in series- voltage issues

Comments

Categories