AGM Battery Failure - Two Sets in Two Years

[MarcoPoloMark]

Great problem. Lots of well intended solutions. The one thing that prematurely kills AGMs is overcharging, so you most likely have an overcharging issue. If you charge from multiple sources at the same time (engine, solar, wind...), you definitely have an overcharging issue.

Each charger is volt/amp/pulse limited to avoid bubbling the electrolyte, which in AGMs is very very precious. Absorbed Glass Mat batteries are also called "starved electrolyte" batteries. Really, they cannot spare one drop of precious water. The charging profile is very precisely calculated, sometimes with precisely timed pulses of electricity filling the last 80% to 100% charge (absorption).

Bulk = full on charging; give it everything you got.

Absorption = very precise charging profile (0.10 volt or amp for 0.10 second too long, and you've started ripping the water molecules into oxygen and hydrogen gas, which Sealed Lead Acid batteries (like AGMs) then have to burp out or crack open from the building gas pressure).

Float = constant 13.62 volt (temp adjusted) "charging" at very low amperage to preclude the lead from slowly and permanently bonding with the sulfur from the battery acid (slowly growing lead sulfide crystals [PbSO4] which turns the battery acid [sulfuric acid H2SO4] into water and excess hydrogen [H2] which also makes the battery burp potentially explosive gas).

If two precisely controlled charging systems are simultaneously charging, their overlapping "absorption" charge profiles are usually ignorant of each other, thereby bubbling the electrolyte and likely burping out precious hydrogen and oxygen.

In other words, I'm guessing your loss of capacity is due to severely dehydrated AGMs caused by simultaneous charging from multiple uncoordinated sources. Or maybe just one source is using the wrong absorption charge profile.

An unrecommended proof is to inject a CC or four of distilled water into each damaged sealed battery cell, seal the injection site, charge singly with a proven AGM charger, then capacity test again. It would be easy to inadvertently unbalance the cells by hydrating them unequally, though the cells are likely already unbalanced from the destructive overcharging. Regardless, recovered capacity would definitively prove absorption phase overcharging.

Another proof would be to install new (or mostly healthy) AGMs and fully charge normally with the existing boat system(s), all-the-while listening to the AGMs burp valve(s). These valves are usually designed to release excess gas once the internal pressure is about 2psi higher than the external air pressure. If the AGMs are burping or brapping or farting or raspberrying, you are slowly killing them again.

Best wishes,
MarcoPoloMark

[alaite]

We had exactly the same problem in our sail training boat here in Finland. Our 4 pieces of 12V/160Ah Mastervolt AGM batteries died in about one year. At home I tested one of them by charging it with CTEK charger specially programmed for AGM batteries. The charger showed that the charging went fine up to the float charge. I loaded the battery at 10A current and the voltage dropped in one hour down to 10-11V. So my understanding is that there must be some serious quality problems with Mastervolt AGM batteries.

[alaite]

As a test I injected one battery with distilled water to each cell of the Mastervolt AGM battery and even charged it by using a lab power supply unit up to 15V and saw the battery gassing. When I discharged at 10A for 1 hour the battery voltage dropped between 10-11V. I still believe that there is a serious quality problem with these AGM batteries. The previous battery bank was Exide closed cells in the same environment. They lasted for 5 years and we used the same charging units during those years.

[sailjumanji]

Originally Posted by CarlF

My understanding is that most AGM’s don’t mind a long absorption (e.g, at 14.4v). If it’s OK with MasterVolt, I would just increase the minimum absorption time on your solar controller to 4 or even 6 hours.

The standard absorption time is maximum 4 hours. It is settable, but with a MasterBuss system. I think we are heading in that direction.

[sailjumanji]

Originally Posted by MarcoPoloMark

Great problem. Lots of well intended solutions. The one thing that prematurely kills AGMs is overcharging, so you most likely have an overcharging issue. If you charge from multiple sources at the same time (engine, solar, wind...), you definitely have an overcharging issue.

If two precisely controlled charging systems are simultaneously charging, their overlapping "absorption" charge profiles are usually ignorant of each other, thereby bubbling the electrolyte and likely burping out precious hydrogen and oxygen.

In other words, I'm guessing your loss of capacity is due to severely dehydrated AGMs caused by simultaneous charging from multiple uncoordinated sources. Or maybe just one source is using the wrong absorption charge profile.

Another proof would be to install new (or mostly healthy) AGMs and fully charge normally with the existing boat system(s), all-the-while listening to the AGMs burp valve(s). These valves are usually designed to release excess gas once the internal pressure is about 2psi higher than the external air pressure. If the AGMs are burping or brapping or farting or raspberrying, you are slowly killing them again.

The advantage of having everything onboard Mastervolt, is that the chargers are set up to not step on each other. When the MassCombi is on, the Solar Chargemaster immediately drops out. I've checked it, and its easy to see on the display. It reads the system voltage, and 0.1 amps. When I put amp meter on the cable, it reads 0.0 amps.

The one charger not Mastervolt though is the two Yamaha alternators. Those are regulated with rectifiers. But after very long motoring stretches - like 8 hours - I have seen them get batteries all the way up to 14.1-14.2 volts maximum, and current only around an amp. I have a concern that holding at this high voltage for long period is one culprit. That said, the start battery is still original, and gets the same voltage, and it is only slightly below par.

[Bycrick]

You’ve got measurements knee high to a tall giraffe. You’ve got enough suggestions and comments to write a book. So, back to basics. You only kill batteries this quickly for 4 reasons:

1. It’s the wrong battery for the expected service, eg using a starter battery for your house loads. Master volt knows what they think the batteries are for, so this is unlikely.
2. Lousy batteries from the beginning. I’d have discounted that except for some of the previous problems reported in this thread. I mean, two sets of bad batteries? Why you?
3. Consistent overcharging leading to loss of electrolyte.
4. Consistent undercharging leading to sulfation.

I just noticed that none of your measurents seem to track amp hours. AH measurements are perfectly accurate to the measurement capability of the meter. An AH meter that reads amps to 1% is going to count AH to 1% plus any clock error. That’s more than good enough to tell you what’s going in and out of the battery if your measuring amps at the battery.

So, if you start with a "fully charged battery" at sunset and have drawn it down to an indicated -100Ah at 0600, that’s an accurate number. Now you charge it all day, with whatever combination of sources, at whatever voltage, and with whatever current. If, at sunset, your AH number is less than zero, you know without doubt that you’re undercharging. If the AH number is way positive, you have an indication of overcharging. You won’t be able to really tell what the battery state is but you’ll know which direction to look.

[MarcoPoloMark]

Alaite,

That is pretty telling: same usage environment, vastly different longevities between Exide and Mastervolt AGMs. I guess the relevant question is: who manufactures MasterVolt batteries? In the USA, three manufacturers make almost all the batteries sold here under many many different labels: Exide, East Penn, or Johnson Controls (who alone has almost half the market).

I don't know if MasterVolt has their own dedicated lead smelting and battery manufacturing plant(s).

I assume the highest quality battery comes from well compensated, well educated, skilled workers who derive their happiness from a job well done rather than from a bottle or a syringe. Without access to that worker-level information, I place great weight on honest and unbiased end-user testimonials such as yours.

How much distilled water did you add? Could you tell if you over- or under- hydrated (which would weaken the sulfuric acid concentration- or continue starving the electolytic volume- and respectively still diminish performance)? Was there sufficient time for the water to re-hydrate the glass mats and re-energize the sulfuric acid, or did the sulfur leach out with some lead and crystallize into lead sulfites. Did you charge the re-hydrated battery to full using a conventional AGM charger prior to re-testing the Ah capacity? Did the amp/hour capacity improve with added water, or was there another mechanism that was principally responsible for diminished capacity?

Sorry for all the questions, but I'm trying to learn. I've never messed with MasterVolt batteries. When able, I often prefer FLAs to SLAs because I can add distilled water to the flooded ones. However I prefer affordability over simplicity over maintenance over disposability, so SLAs still have their place in my heart, as do LiFePO4s, NiMH, and other chemistries (check out NDB - nano diamond batteries in 2 to 10 years!)

MarcoPoloMark

[MarcoPoloMark]

There may be a voltage drop at the battery isolator/combiner between the house bank and the starting battery. This may lead to the house bank over volting and the starting bank (of one) perfect volting. Reversing the two output cables may switch the longevity problem to the starting battery and keep the house bank healthy.

[sailjumanji]

Originally Posted by Bycrick

You’ve got measurements knee high to a tall giraffe. You’ve got enough suggestions and comments to write a book. So, back to basics. You only kill batteries this quickly for 4 reasons:

1. It’s the wrong battery for the expected service, eg using a starter battery for your house loads. Master volt knows what they think the batteries are for, so this is unlikely.
2. Lousy batteries from the beginning. I’d have discounted that except for some of the previous problems reported in this thread. I mean, two sets of bad batteries? Why you?
3. Consistent overcharging leading to loss of electrolyte.
4. Consistent undercharging leading to sulfation.

I just noticed that none of your measurents seem to track amp hours. AH measurements are perfectly accurate to the measurement capability of the meter. An AH meter that reads amps to 1% is going to count AH to 1% plus any clock error. That’s more than good enough to tell you what’s going in and out of the battery if your measuring amps at the battery.

So, if you start with a "fully charged battery" at sunset and have drawn it down to an indicated -100Ah at 0600, that’s an accurate number. Now you charge it all day, with whatever combination of sources, at whatever voltage, and with whatever current. If, at sunset, your AH number is less than zero, you know without doubt that you’re undercharging. If the AH number is way positive, you have an indication of overcharging. You won’t be able to really tell what the battery state is but you’ll know which direction to look.

Actually, look at the graphs again. All of them show cumulative amp hours. Since we measure so often, we just integrate the amps at time to come up with cumulative. We've ignored the AH counter, because now we are testing one battery at a time, and using the larger house cables - and shunt - whether testing the start or single house battery.

I know how to use an AH counter, and the DCSM does this for both banks. It requires a calibrated shunt on each, which we have. An input is also capacity - which we have now as well since we capacity tested all batteries. It also requires a Peukerts factor, which I know how to calc on good batteries, but not on these.

That said, what you are getting to is "did we monitor that when we thought the batteries were working right?" Answer is yes. And it looked to count down to a reasonable to slightly less-than-actual-used AH number, and then counted right back up and would just about click to capacity at the same time it went to float. That was all when we were cruising. Everything looked fine. There was no indication that anything was failing by that observation.

It's even working pretty well now. But here is how this whole "we gotta problem" started. Look at this pic, taken on our last attempt to overnight. We are down 17 AH from our expected capacity of 450 AH, for our one year old batteries. The voltage is 11.6 volts, and current load only 3 amps.

Honestly, we have probably beat this horse to death. I've got a lot of the input I hoped to get. I have some calls with Seawind and Mastervolt New Zealand scheduled for early next week. I got a sneak preview that they are going to propose some configuration changes - take out the VSR so that the motors only charge the start battery, possible issue with the solar controller dropping into float early due to some erroneous sense that the battery is fully charged (and the add-on depletion effect of doing that multiple days in a row). Also add a MasterBuss to facilitate some changes to standard programming, as well as better monitoring capability.

The heat affecting batteries, we cannot do much about other than more space between when the new ones are installed.

As someone says, some of the equipment testing we are doing is hard to tell whether its an artifact of the bad batteries, or equipment being off a bit. Will make configuration changes, put in the new batteries, and then retest.

[mcon12000]

Reading threads like this one from start to finish makes me really happy I'm taking the ABYC certification course right now. I'm far from an expert but I'm reading as much good information here as bad albeit all of it with good intentions.

I gather from the back and forth that you have confirmed that there are no negative wires on the battery side of the shunt? It seems like this has been beat to death but I have to ask because a small load you're unaware of will throw off all your calculations and leave the bank discharged beyond what might be ideal at 40% in this case shortening the service life dramatically.

The tin missing from the buss bar shouldn't be a factor in the short term but if you should replace it while you're pulling your hair out. It was likely an issue prior to you cleaning the area up and it could cause a problem right around the time you forget about it.

I'm not familiar with the charging system you have but it sounds like my victron setup and the comments about multiple charge sources being an issue are misplaced with the newer systems. The exception being the engine alternators so this has me thinking...

1. Your shore charger is intelligent and safe

2. Your MPPT chargers are intelligent and safe

So in theory the house bank is ideally serviced by the shore/generator and the MPPT but the alternators do not have comparable sensing available. Since I'm not sure you can easily add an externally regulated alternator to your outboards can I suggest adding two Victron 30amp DC to DC Chargers to your engine side of the system?

It would eliminate the last possible problem on the charging side and provide a perfect charge while motoring. They solve many problems associated with engine charging of start and house banks. So alternator would charge a start only battery and the DC to DC charger charges the house.

Mine is setup the opposite with an externally regulated alternator charging the house bank directly and the start battery (different chemistry) is charged via the DC charger.

You seem to be all over this so I'm just throwing things out there as I've found even useless suggestions spawn great ideas..

I'm following and if I have anything to offer I'll jump in again.

[Bycrick]

I missed the amp hours in the graphs. My fault. I’m also reasonably sure that you know how to measure amp hours, after all the measurements. I was just wondering whether or not you were getting swamped with so much information that it was hiding the problem. So, from what you see, is it a problem with over-charging or under-charging? We can all find out why later.

[Bycrick]

Catching up on the thread. The loss of tinning on the bus bar is because it wasn’t tight and you got moisture under the terminal. Clean it up with very fine sandpaper and make sure the terminal is tight. Tinning contributes nothing to conductivity. But the fact that it wasn’t tight might mean something for the problem.

The only suggestion I’d make at this point is ONLY CHANGE ONE THING AT A TIME. Don’t put in new batteries and change the wiring. If you change multiple things and it "fixes" it, you’ll never know what was wrong.

[sailjumanji]

Originally Posted by Bycrick

Catching up on the thread. The loss of tinning on the bus bar is because it wasn’t tight and you got moisture under the terminal. Clean it up with very fine sandpaper and make sure the terminal is tight. Tinning contributes nothing to conductivity. But the fact that it wasn’t tight might mean something for the problem.

The only suggestion I’d make at this point is ONLY CHANGE ONE THING AT A TIME. Don’t put in new batteries and change the wiring. If you change multiple things and it "fixes" it, you’ll never know what was wrong.

That's tough. I need to put in the new batteries to test some things, but I dont want to at the peril of the new batteries.

Regarding the buss bar, the terminals were tight, when removed for the first time. And they are tight now. ;-) It is something I will be watching though, to see if corrosion reoccurs.

[Bycrick]

Maybe the nut was tight, but the terminal wasn’t "tight" against the bus bar. If it were, you wouldn’t have gotten enough moisture under ir to corrode the tinning. That’s why they don’t want you to use Nyloc nuts foe electrical connections: you can’t tell whether the torque I’d from the nut or because the joint is tight. A damaged lug, where the end isn’t flat can do this also.

I can see why you’d be nervous with new batteries. But "overcharging" is something easy to see: either the voltage is too high, which you should be able to catch by setting the high-voltage alarms to lower values, temporarily, or you’re undercharging which should show up in the amphour test I talked about. It’s unlikely to damage the batteries in a couple of days unless something is wildly out of line. And you’ve done enough measurements to know how to catch that.

[Bycrick]

Another little gotcha on tight terminal nuts. While the bus bar is tinned copper, I think the studs/nuts are stainless. The Blue Sea eBay site didn’t say. But I had one fellow who’d very carefully cleaned the bus bar studs and nuts with solvent. So clean, in fact, that two of the nuts galled on tightening. The nut wouldn’t come off and he broke the studs. But the terminal itself would move.