Battery problems, really stuck!

[crisjones]

I think you have misunderstood when people say the SG does not vary - it certainly does. What they are saying is that the actual SG reading does not vary when the battery is under load or being charged - this is where the simple voltage reading can fail misserably, voltage is only a guide and only when the battery has been rested.

The SG value is a function of the chemical process going on inside the battery - basically as a battery is discharged the sulphuric acid in the electrolyte reacts with the lead plates to form lead sulphate, so the sulphuric acid percentage reduces as the battery is discharged and hence the SG also reduces. Charging is pretty much a reversal of this process with some inefficiencies.

Since SG is a measure of the chemistry inside the battery it is pretty much unaffected by amps being removed or amps being added so you can actually test the SG while the battery is being charged or discharged or at rest - the result will be pretty much the same for a given state of charge.

So if you start with an SG of 1.28 at 20 deg C then the batteries are fully charged - make sure all cells read the same within a couple of digits (1.26 to 1.30 max). At this SG reading the batteries are pretty much fully charged regardless of voltage reading, amps in or amps out. At this point you should reset your battery monitor to 100% charge and zero AH used.

After 4 or 6 hours of discharge with your voltage getting down to 12.2 or 12.3 V then you should again test the SG - you do not need to switch everything off or even to let the batteries rest but take sensible safety precautions. You do not need to test every cell if they were all very close to begin with, one cell per battery will do.
This SG value will give you an accurate state of charge so if it is 1.17 - 1.18 then the batteries are at 50%, if the SG is 1.23 - 1.24 then the batteries are at 75% charge regardless of the voltage reading.
Then you can compare the actual state of charge with your battery monitor, they should agree quite closely if you started fully charged. If not then maybe some load or charge currents are not being measured by the shunt or the monitor is not calibrated to the shunt correctly.
The SG reading can almost always be relied on to give an accurate state of charge value - main exception being if the electrolyte gets stratified that can happen in very tall batteries, also if you have just added water the reading will be incorrect since the water will not be fully mixed. However if you start with correctly topped up, fully charged batteries after an overnight charge then the readings will be valid.
SG is really the only way to accurately measure State of Charge - voltage is little more than a guess.

We have Rolls batteries, now nearly ten years old and still working very well with at least 90% of their original capacity still available when tested over a 10 hour load test. We often see the battery voltage on our Victron monitor at 12.1 or 12.2 even though the discharge percentage in AH is only 30 or 35%. I have come to the conclusion that the Rolls batteries seem to read a little lower than most others although there is no real reason for this. I rely on the state of charge percentage shown by the battery monitor and always avoid going below 50% - this has obviously worked well since the batteries are still going strong.

Finally just one other point for wet lead acid batteries - if you are not having to add any water then the batteries are not getting fully charged. Obviously you should not be adding lots of water, that probably indicates over charging.

 

[john_morris_uk]

The SG does vary, but not in an 'on load-off load' fast dynamic way. It varies slowly as the battery is discharged.

The SG is a very good way of determining the state of a battery. It doesn't lie or mislead you.

Personally, my suspicion is that one or more cells have gone down in one or more of your batteries and despite your excellent charging system, the other batteries have been pulled down and the bank has lost capacity. From what you are saying its acting like a battery bank of small capacity.

All should be revealed by taking SG measurements of every cell at various states of charge and discharge.

 

[macd]

Pete (prv) and macd have now thrown me saying that the sg won't vary and you've lost me now. The Rolls site seems to say take sg readings yet you both say that sg won't vary. Macd also says take sg readings when the volts read the magic 12.3v (after leaving for 30 mins or so).
I'm sorry, I'm stuck again

Mike: others have offered further explanation of SG readings.
To clarify: voltage readings are volatile. They vary considerably with load (just check the voltage while using the windlass or starter motor, and you'll see a huge drop).

SG readings change -- of course they do, or there would be no point taking them. But they change slowly and progressively. What they give you, as others have said, is an accurate indication of the chemical state of the battery. (Essentially it's a measure of the number of electrically-charged ions which are available to migrate from plate to plate in the battery, which is the means by which it stores and releases electricity.) SG will change only slowly under load (or under charging). You can rely on this.

Checking batteries at part charge (you seem to have misunderstood my last post):
if you wish to do this by volts, then the batteries need to rest;
if you do this by SG, then there's no need for resting: just do it. The SG is the SG. (You may find that by the time you've run the hydrometer through 24 cells, your voltmeter is showing rather more than it did when you started...because they've had a little rest )

So, next time you get a reading of 12.3V (when you think it should be more), check the SG. It will confirm or disprove the state of charge suggested by the voltage reading. I presume you have access to a table of SG readings versus state of charge? There's no reason you can't do this test on the hook or at a berth: just switch lots of things on, let the voltage drop to 12.3, and test the SG.

Others have suggested that one battery may have lost a cell. SG readings of all cells will cofirm whether this is so. If you wish to double check by voltage readings, disconnect all batteries, let them rest for a good few hours (12 or more is best), then measure the voltage of each battery. They should all be the same to within about 0.05V or less. If one is down by 2V or so, it's lost a cell. A simple hand-held tester set to 20V DC range will suffice. However, reading between the lines in your opening post, I suspect you've already confirmed that all four house batteries are fine.

P.S. You can certainly believe everything chrisjones tells you, above. He comes from a long line of well-nurtured batteries

 

[Vegable]

Thanks Chris, John, and Mac. I now appreciate what's going on and what I need to do when I'm back up on the 7th June. Chris and Mac, your explanations have been superb and very detailed and I appreciate the time you've spent on me. My boat's called Conspiracy (easily recognisable by its dim lights ) and if our paths cross please make yourselves known and we'll share some malt extracts together.
Thanks to all
Mike
As macd suggested, all the cells have have been checked and are within expected readings, and yes thanks I have access to sg readings v state of charge.

 

[Birdseye]

Golly - enough detailed answer here to make Brian Cox's eyes glaze over.

The thinking in your intial post is right Vegable but I am less sure about both your data and your assumptions.

With a 60 amp alternator, a Sterling digi controller, and a 300aH battery bank it takes my boat 6 hours of engine running to get up to float voltage, and thats with a similar equipment current draw to your situation. The reason is simple. Even when half charged the current from the alternator into the battery doesnt exceed 45 amps and it rapidly falls down from this as the battery voltage build up. As the batteries approach fully charged the current is below 10 amps. Reason is simple - the current flow depends on the difference between the alternator output voltage and the battery voltage so as the batteries get more charged and their voltage rises., this difference shrinks and the current flow shrinks with it. So your 1 hour of engine running will only just tickle the surface of battery charging

Next is voltage measurement. How are you doing this? Is it a digi meter and if so does it have a good internal battery. My digi meter reads all sorts of wrong data when the internal 9 v battery is on its way out. An analogue meter is a waste of time.

As others have said you need to establish your 50% charge point with the hydrometer but because this is cumbersome and dodgy at sea, you need to establish what battery voltage this 50% corresponds to. This depends on what load the batteries are under. So chose a standard configuration of electronics switched on and then measure the volts to get your reference point. Could be that 50% charge corresponds to 12.2 or 12.4 or 11.9 all depending on what is switched on.

If, once you have done all this, you still find that the batteries are going from full to half full in a lot less than the 60 hours or whatever you reckon then there are only two possibilities. One is that your reckoning of the load is wrong. The second is that you have one or more duff batteries.

 

[Vegable]

Hi, I was wrong in mentioning the hour charge up. I'm fully aware of alternator charging times etc and it's led some people in the wrong direction. What's been important is that I'm starting my trip with fully charged batteries and after 3 hours of a small amount of usage the battery voltage fell to what I thought was half capacity ie 12.3 volts. I now know that this assumption is wrong. Voltage readings under load cannot be indicative of state of charge, (unless other factors are known, he hastily adds) and most probably I've been worrying/concerned unneccessarily. My booze cruise back to Pwllheli on the 7th will be over 10 days and you can bet your bottom dollar I'll be switching all sorts of equipment on and sucking with my little hydrometer pippette for most of the journey and recording everything! No booze for me, sulphuric acid and soda will be my tipple.
And as you so rightly say, if things are still not right then it HAS to be the batteries. My measuring equipment isn't in doubt backed up with 3 voltmeters and 2 ammeters and in all cases with the manufacturers power usage specifications. I'm one of those wierdos who devour instruction books!
Once again thanks for all your help and I'll report back in mid June.
Mike

 

[pvb]

What's been important is that I'm starting my trip with fully charged batteries and after 3 hours of a small amount of usage the battery voltage fell to what I thought was half capacity ie 12.3 volts.

Firstly, are you absolutely sure you're starting with fully charged batteries? Ignore SG readings, just measure resting voltage with a decent meter. And what sort of float current are these fully charged batteries taking?

Secondly, you say you have a Rutland wind generator constantly connected, so it would be reasonable to expect that this ought to provide some of the 4 to 6A you're using. If the batteries are way down after only a few hours, there's something draining them. Is the Rutland connected through your Ah meter?