Flat Battery Mystery

[KAM]

I went down to the boat last week. The two 110 Ah domestic batteries were down to 12 V so I plugged in the charger. I went down a couple of days later and unplugged it but left the battery charger connected. 3 days later the batteries were down to 9.8 V. I checked and the charger was polarity was correct as it was still connected. All switches were off and main switches off. I checked the charger leads and they read open circuit with the charger unplugged. Two 20 watt solar panels are connected to the battery but not charging much due to mid winter and position. I can't work out why the batteries would go completely flat so quickly. The battery monitor is a Xantrex which draws about 100ma and has been on the boat for 12 years. The solar controller is a Vectron with trickle charge for the starter battery. I had the starter battery go suddenly flat last winter so moved the power supply for the Xantrex monitor which was wired to the starter battery over to the domestic battery as I thought the starter battery was suffering as it only gets a trickle charge from the solar panels via the Vectron controller. To summarise the installation has been on the boat for 12 years without problems. Starter battery went flat suddenly last winter. I moved the Xantrex power supply to the domestic batteries at the start of the winter. The domestic batteries have now also gone flat suddenly despite holding charge reasonably well over the winter and being recently fully charged. The batteries were 5 years old so due for replacement anyway but I'm reluctant to install new batteries until I understand the failure. Can anyone shed any light on what the problem might be.

 

[superheat6k]

I would be looking to see if something has stared draining the batteries, and for this task I would start at any bilge pump float switches.

As a simple check to see if an unwanted electrical path does exist with all services off, including your battery monitor, disconnect the +VE terminal from the battery then measure for voltage between the disconnected lead terminal end and terminal post. If there is no circuit the display should be 0 V. If there is any voltage registering then a current flow exists, indicating a problem somewhere.

 

[peteK]

I think I would disconnect every thing from the batteries and then put them on charge and moniter them over a few days to see if they are holding the charge.

 

[KAM]

OK so batteries recharged and seem to be holding voltage. I disconnected and everything seems open circuit. However when I reconnected the Victron solar controller connection to the battery there was a spark. Quite a big one. A bit of a surprise as it was raining at the time and there were only 2x 20 W panels working. All seemed normal otherwise. Output from the two 20W panels was 15v. There are two other panels connected to the controller but they are under a cover and not outputting. Could this cause a problem. Really don't understand where the spark came from. The panels seem to have their own diodes. All left disconnected to see if the batteries hold charge over several days. Anything else I can check.

 

[pvb]

OK so batteries recharged and seem to be holding voltage. I disconnected and everything seems open circuit. However when I reconnected the Victron solar controller connection to the battery there was a spark. Quite a big one. A bit of a surprise as it was raining at the time and there were only 2x 20 W panels working. All seemed normal otherwise. Output from the two 20W panels was 15v. There are two other panels connected to the controller but they are under a cover and not outputting. Could this cause a problem. Really don't understand where the spark came from. The panels seem to have their own diodes. All left disconnected to see if the batteries hold charge over several days. Anything else I can check.

When you reconnected the solar and saw the spark, did you look at the Xantrex monitor to see whether current was flowing into or out of the batteries?

 

[KAM]

It was such a dark rainy day although I could measure voltage at the panels there was no current flow into the batteries from the controller that I could see. The spark was a bit unusual. I've reconnected the charge cable from the controller before and not noticed it. It was big enough to make me jump. Can't think there are any big capacitors in the system. It may have come from the starter battery. Think I'll check the diodes tomorrow.