Perhaps it's on its way to being shorted altogether.  

Have you checked the specific gravity?  Is the battery "bulgy?"

Physically, the resistance is a function of the spacing of the plates and the conductivity of the solution.  If the solution is more or less normal, then the only other possibility would seem to be that the spacing has been reduced.

TTFN

What IRstuff said.

You could be slowly migrating metal from + to - and shortening the distance between the electrodes. There could in fact be an electromigration-formed shorting path which is so small in cross-section that you still maintain full voltage. If this is the case, you should expect the voltage to eventually start to drop. Your leakage path (if you have one) could be at the micro-amp level.

Have you put a load on the batterry? My guess is this cell will not do very well when a load is placed on it. It sounds like you are developing a short across the plates.

If you have access to an ESD gun, you might try pulsing it across the battery terminals a few times. If the short is microscopic, this might obliterate it.

Hi,

Many thanks for the input,

I recently performed a 5min discharge test to the battery string, the suspect battery held up OK with no real drop off in voltage in comparison to the other cells?

On checking the cell resistance a couple of days later, it had returned to around the expected value (4 ohms). It has been flagged up for closer monitoring in the future. Possibly the discharge altered the solution / formation of the battery in some way and gradually allowed it to return to normal?

We had several fixed cell monitoring systems installed about a year ago to around 1600 - 6 & 12 volt batteries. In the past low resistances would of gone un-noticed, but now we have to investigate all alarms that arise. The problem is we have to set our own alarm threasholds, these may be too low or high as we have no recommmended values to go by. Only percentage values relating to new batteries are available and these degrade over the life of the battery. Getting the levels right is going to be a matter of experience along with over reactions to wrongly set alarm levels.

Regards

My guess would be that you had a conductive filament form which was disrupted under load.

Are you really sure you're measuring 4 ohms?

If it's a UPS application based on 192 cells then I'm guessing your looking at around 100AH cells - as any lower and you would have probably stuck 12v blocks in. Typically Yuasa, Hawker, Exide, Chloride, etc, start doing big boys batteries for real UPSs. If so internal cell resistance should be in milli-ohms - around 25mOhms when checked with a 1000Hz bridge would be very typical for a 100AH block.

The lower the internal cell resistance the better. Batteries can be regarded as u/s as their internal cell resistance goes much above 25% of initial new. If you really have 4 ohms in a 6V block then you're not going to get an awful lot of power out of the battery so something is iffy in the measuring gear!