How to find a "dead" battery

[sp4rk]

Nothing at all to do with our cars but my office in the UK reports that one of our UPS's is reporting a dead battery.

With all the batteries in our cars, perhaps we can find the answer here?

We're running huge UPS's in our data floor there and the UPS says ONE battery (in one UPS) has failed. One of 32.

Problem is on inspection all battery trays read 108 volts. Yet during a self test, we're told ONE has failed.

Rather than recycle 32 batteries and buy 32 more at £15 each, is there a way to determine which one is gone / going bad? Not just the money, these batteries are 2 years old; pity to throw away 32 when only 1 needs replacing.
Voltage is good ... what do we look for?

 

[bollar]

Nothing at all to do with our cars but my office in the UK reports that one of our UPS's is reporting a dead battery.

With all the batteries in our cars, perhaps we can find the answer here?

We're running huge UPS's in our data floor there and the UPS says ONE battery (in one UPS) has failed. One of 32.

Problem is on inspection all battery trays read 108 volts. Yet during a self test, we're told ONE has failed.

Rather than recycle 32 batteries and buy 32 more at £15 each, is there a way to determine which one is gone / going bad? Not just the money, these batteries are 2 years old; pity to throw away 32 when only 1 needs replacing.
Voltage is good ... what do we look for?

May or may not work, but... If you shake each battery, you may find that one has a rattling sound. Try replacing that one. On smaller UPS systems that's an easy way to identify a likely bad battery.

 

[sp4rk]

Might mention ... relevant or ??? ... this is the batter in question ...

RITAR 12V 5.0ah (replaces all 4ah, 4.5ah) - Brand New - Cube Shape battery | eBay

 

[Ampster]

I don't know anything about your battery configuration, but I will guess that you have 32 cells in series. You could measure voltage of each cell and that may help you find the bad cell. If you can figure out the pack arrangement try each half (16 cell group) first. The weak cell should be in the half with the lowest voltage. You can continue to narrow it down by measuring each sucessive half until you find the cell.

 

[sp4rk]

Apparently, ALL cells measure the same proper voltage, hence my posting it here.

 

[ggr]

The most reliable symptom of failure is usually an increase in internal resistance in the cell. So if you can measure them while delivering power, you should find that the dead cell has less voltage at that time. This also manifests as less voltage when at least partially discharged. At least that's my understanding; I'm by no means an expert on this stuff.

 

[jerry33]

If those are the usual hot-swappable batteries, they're usually mounted in packs of three to six. Just purchase one pack and replace an arbitrary pack. If after a day, there is still a bad battery, replace the next pack with the pack you took out. Repeat until you find the pack that has the bad battery.

 

[MichaelS]

If the voltages read normal when measured, is this during standby or under active load? In other words the battery voltages must be measured while being discharged. If you measure the voltages while the UPS is idle, all you are measuring is the charging float voltage.

There could also be a problem with the monitoring circuit, and nothing wrong with the battery itself.

There really isn't enough information to properly diagnose this problem.

 

[Ampster]

Apparently, ALL cells measure the same proper voltage, hence my posting it here.

Then, the subsequent tests should help. The swap suggestion may be the best approach.