Battery failure?

[Ian_Edwards]

On our last cruise the starter battery failed, it's a sealed, valve regulated, 110 amp hour, wet lead acid battery, in its 6th season.
We first noticed the problem when the gas alarm went off, showing both LPG and CO alarms, we couldn't smell LPG and we were sailing, no engine, no gas cooker or hob lit.
When I lifted the floor boards to have at the LPG sensor, the reason for the alarm became clear, the started battery was very hot, and acid had escaped from the battery vent.
The only source of charging in use was the 200 watts of solar panel, pushing out about 17 to 18 amps with about 13.3 volts at the batteries (hot sunny day in NW Scotland), via a Morning Star MPPT controller.
The wiring for the batteries is very simple, all the charging sources are brought to a common point and connected to the starter battery, a voltage controlled relay will then parallels up the domestic (500 amp hours) and started bank when the voltage gets high enough.
The only thing I can think is that the battery failed with an internal short circuit.
Is this likely?

 

[Hadenough]

6th season! Borrowed time for that type of battery. Sealed, valve regulated? Unless you prise the top off and top up the water now and again there is no way a 'sealed' battery will replenish itself, especially with the charging regimes we throw at them with solar / mains etc.

 

[coopec]

"we couldn't smell LPG"

You may not have been able to smell LPG but did you light a match to check?

Clive

 

[macd]

200 watts of solar panel, pushing out about 17 to 18 amps with about 13.3 volts at the batteries (hot sunny day in NW Scotland)

They're some panels. We don't get that specific output at solar noon in the Med.

 

[Momac]

As said 5 or 6 years is an okay lifespan for an ordinary battery.

 

[lw395]

As said 5 or 6 years is an okay lifespan for an ordinary battery.

But failing in a dangerous way is not an Ok end of life.

I would guess the battery was overcharged, eventually resulting in one cell shorting, so the remaining cells get seriously cooked.

For comparison, my neighbour's car battery needed replacement this year, 12 years old and 125,000 miles.

 

[Graham376]

They're some panels. We don't get that specific output at solar noon in the Med.

I suspect his regulator is telling rather large lies, around half quoted amps would be nearer.

 

[Ian_Edwards]

They're some panels. We don't get that specific output at solar noon in the Med.

These are brand, good quality, new, back connected panels, I've checked with a good quality multimeter and I get 9 to 10 amps with a bright overcast sky, no direct sun, and 17 to 18 amps if full midday sun.

With 200 watts at 13.3 volts you'd expect just over 15 amps, with a good MPPT controller, so what I'm getting is about 15% more than the spec', however, you have to remember that this is the far NW of Scotland, with very little if any air pollution, and the 200 watt spec' is defined by some standard irradiation of watts per square meter, which may or may not happen in the real world.

 

[lw395]

These are brand, good quality, new, back connected panels, I've checked with a good quality multimeter and I get 9 to 10 amps with a bright overcast sky, no direct sun, and 17 to 18 amps if full midday sun.

With 200 watts at 13.3 volts you'd expect just over 15 amps, with a good MPPT controller, so what I'm getting is about 15% more than the spec', however, you have to remember that this is the far NW of Scotland, with very little if any air pollution, and the 200 watt spec' is defined by some standard irradiation of watts per square meter, which may or may not happen in the real world.

Panels presumably aimed straight at the sun?
15% over nominal 'spec' is typical for a new panel, dropping to 100% of nominal after a couple of years and about 80% after 15 years.
In standard conditions.

 

[Momac]

But failing in a dangerous way is not an Ok end of life.

For comparison, my neighbour's car battery needed replacement this year, 12 years old and 125,000 miles.

It did not harm anyone.
A car generally keeps the battery fully charged and if used daily is much kinder to a battery than irregular use on most leisure boats.

 

[Ian_Edwards]

But failing in a dangerous way is not an Ok end of life.

That was my feeling, and why I asked the question, it's lucky that I lifted the floor boards, not to check the batteries but to check the gas sensors, the battery was hot enough to boil the acid, and left long enough could have caused a fire.

I don't think it was cause by the regulator, the batteries were only connected to the solar cells at the time. 17 amps charging 610 (500 domestic +110 starter) amps hours of battery connected in parallel isn't really enough to make a battery that hot, and the 4 remaining domestic batteries are OK. I've also carefully monitored the charging voltages since the failure, from the solar, alternator and mains charger, and it all looks OK.

I wasn't really surprised that it failed after 6 year, a bit disappointed, it had been well looked after, never discharged more than it takes to crank the Yanmar, and kept on float over the winter. In a similar way to lw395, we had a Ford Focus, which was originally my wifes car, which we gave to our eldest daughter, was sold after 9 years with the original battery.

 

[syc1]

I'm no sparky so don't pretend to fully understand your system, but post 8 seems to be saying that these are brand new recently fitted panels, is there any chance that the something happened to disturb the existing set up whilst they were being fitted?
Law of sod and all that.

 

[lw395]

The problem is, once the start battery fails with a shorted cell, it is charged not just by the solar panels, but also by the house bank.
If the start battery is getting hot, its gassing voltage is falling.
Once it gets hot, it all starts to run away until either some wires burn out or the battery potentially explodes all the hydrogen it has electrolysed.
I have seen the aftermath of a small battery explosion, I would not want a bigger version of that on a yacht.

I suspect this will become more common as more people fit more solar power.

A starter battery does not need daily solar charging. At most it wants occassional float charging.

 

[VicMallows]

A starter battery does not need daily solar charging. At most it wants occassional float charging.

Precisely. If OP insists on charging the starter battery from the PV, then the VSR should be transferring the charge to the house bank once the starter battery voltage has risen.

Another problem can be that once all batteries are fully charged, if there is very little drain on the batteries (eg if the boat is not being used) the default program of most controllers can be far too aggressive, typically maintaining the 'boost' phase for 2 hours or so, before reducing to a relatively high float voltage of 13.8v. The problem is that the battery voltage usually falls enough during the night for the controller to resume a new cycle the next day.

I set up a user-defined charge profile with much lower boost voltage/time, and a long-term float of 13.5v. I switch to this when leaving the boat unused for any length of time.

 

[PaulRainbow]

It is not at all uncommon for a cell to fail in a battery, leaving you with a 10v battery. Whatever is charging the batteries will see that as a flat battery and try to charge it up, resulting in a lot of heat and gassing, usually setting the CO alarms off. Very unlikely that it will result in an explosion if left unattended, there should be no ignition source to ignite the hydrogen. The danger comes when someone investigates what is going on and causes a spark, then it goes boom! If you find your batteries boiling or gassing, close the covers and remove the charging source, for as great a distance as possible.

 

[Topcat47]

For what it's worth, I have two small solar panels on my boat. They are primarily those you'd buy to keep a battery charged when leaving your car for a couple of weeks while going on holiday. From memeory, I think they are 2.5amp. Inside my coachroof window they are sufficient to maintain and trickle charge both my house and starter batteries whenever I leave the boat. I separated the batteries many years ago and they run via independent key switches. My house battery is 4 years old and my starter 10. Of course I do not have a substantial house bank nor do I have a large demand for electricity as I only run instruments and lights, but the starter battery is kept charged by the engine running and I only need a top up. Perhaps something similar would be safer, only using the larger bank of panels when running your high demand items? Just a thought.

 

[Ian_Edwards]

Thanks to all who have contributed, I now have clearer idea of the likely cause of the overheating.

It seems that the starter battery, a sealed flooded battery failed, with on cell short going short circuit, and looking like an undercharged battery (10volts 'ish)

The voltage controlled relay, which connects the house bank in parallel with the starter was enabled when the solar cells kicked in and raised the charging voltage above the VCR's threshold. This allowed the fully charge domestic bank (500 amp hours) to supply charge to the starter battery, causing it to overheat and gas, which set off the gas alarm. Both the starter battery and the domestic batteries are fused, but the fuse rating is high, 500 amps, I think. The VCR is rated at 120 amps continuous, and I'll check the resistance of that when o go back to the boat.

The battery wasn't in thermal runaway, when I disconnected it, it cooled down, although that took quite a while to happen.

The starter battery failure was probably caused by a combination of age, and an aggressive charging regime programmed into the MPPT controller.

I've checked the controller and the data sheet, it is set to AGM, the main house bank is AGM and the charging regime is:

Bulk voltage	whatever the panels will provide
Absorption voltage	14.3 v
Float voltage	13.7v
Equalisation voltage	14.5v
Equalisation duration	3 hrs
Equalisation calendar	28 days

All these are programmable.

I'm replacing the starter battery with an AGM, with a CCA which is well over that required to start the Yanmar.

The way the boat is used, I'm on board most of the time from mid April to the end of September, I go home every couple of weeks to clear the mail and do the inevitable admin. I'm seldon away for more than few day, 1 week max. I sometimes leave the freezer running whilst I'm away and the 200 watt solar keeps up with that OK. The boat also has an Empir Bus system (CAN Bus) which controls everything, this system has a quiescent power requirement to keep the CAN Bus running, even if everything is switched off, so the domestic bank is gently cycling every day, as the solar kicks in in the morning and replaces the amp hours used by the CAN Bus.

 

[lw395]

To fully see the regime the start battery is subjected to, you need to know how the controller switches between charging phases.
Very often every time a cyclic load comes on (e.g. fridge), the charger goes back to absorption for an hour or more.
So both banks may spend the majority of daylight hours at 14.3V.

Personally, I'd switch off the VSR unless the engine was running, apart from perhaps a couple of hours a week if you go a long time without running the motor.

 

[PaulRainbow]

I had a similar thing to the OP last year, two out of three domestic batteries failed at the same time, both with shorted cells (what's that chances of that :ambivalence. Needless to say, they got boiled by the 260w of solar power. So in this case having the VSR disconnected from the engine battery would have made no difference.

However, for anyone or any particular installation that doesn't want the engine battery subjected to the domestic charging regime, don't fit a VSR. Alternator output can nicely be split by using a Victron Argofet (or another make of FET based splitter). Mains and solar charging can be restricted to the domestic bank. If there is a reason why, despite not wanting the engine battery to be subjected to the domestic charging regime, you still want it subjected to a maintenance charge you could fit a 10w solar panel to it, you could install a battery to battery charger (bit expensive) or in true PBO fashion you could install a DC-DC converter set to output 13.4v (or whatever best suits your battery), you could also ensure that the DC-DC converter had a current limit.

 

[lw395]

I had a similar thing to the OP last year, two out of three domestic batteries failed at the same time, both with shorted cells (what's that chances of that :ambivalence......

You buy two identical things at the same time, treat them exactly the same, no surprise if they react exactly the same.