Battery Capacity Testing in the Real World

[conachair]

Getting an accurate handle on how much capacity your battery bank actually still has left in it seems tricky for off the grid cruisers, a number of issues come to mind:

it goes against the grain to discharge down to 10.5v to attempt to get an accurate c20 figure to compare with the manufacturers data
.
It's hard enough to keep the batteries charged without having to charge from empty.

Proper test equipment is expensive .

As getting back to 100% charge takes a long time often this just doesn't happen.

Equalising seems to be accepted as necessary occasional to maintain battery health when they don't get back to 100% after every discharge.
...................

With this in mind any comments on doing a "half" discharge capacity test.

For instance, on a fully charged trojan T105 225Ah 6v have a motorbike headlamp or something to draw around 10A and an accurate method of measuring how long it takes for the voltage to reduce to, say, 12.3v. Then compensate for temperature and hopefully you will have a figure to compare next time, even if it doesn't relate exactly to the Ah data of the battery, at least it's something.


If such a test is carried out with accurate voltage cutoff point and exactly the same load each time would the data be helpful?

Any thoughts/suggestions ? .




As for measurement, I have an arduino onboard somewhere which shouldn't take much to get fairly accurate logging of voltage, temp and current data.

TIA

[Maine Sail]

Originally Posted by conachair

Getting an accurate handle on how much capacity your battery bank actually still has left in it seems tricky for off the grid cruisers, a number of issues come to mind:

it goes against the grain to discharge down to 10.5v to attempt to get an accurate c20 figure to compare with the manufacturers data
.
It's hard enough to keep the batteries charged without having to charge from empty.

Proper test equipment is expensive .

As getting back to 100% charge takes a long time often this just doesn't happen.

Equalising seems to be accepted as necessary occasional to maintain battery health when they don't get back to 100% after every discharge.
...................

With this in mind any comments on doing a "half" discharge capacity test.

For instance, on a fully charged trojan T105 225Ah 6v have a motorbike headlamp or something to draw around 10A and an accurate method of measuring how long it takes for the voltage to reduce to, say, 12.3v. Then compensate for temperature and hopefully you will have a figure to compare next time, even if it doesn't relate exactly to the Ah data of the battery, at least it's something.


If such a test is carried out with accurate voltage cutoff point and exactly the same load each time would the data be helpful?

Any thoughts/suggestions ? .




As for measurement, I have an arduino onboard somewhere which shouldn't take much to get fairly accurate logging of voltage, temp and current data.

TIA

I've tried it and it is not very representative because the discharge curve is not as linear as one would expect. If you want a true capacity test then a C/20 load would be applied until you hit 10.5V then an immediate recharge.

Usually only one series pair or one battery of a parallel bank is required to get a good estimate of bank condition.. As the voltage drops your load changes so if you want really close accuracy keep up on your C/20 to keep it a C/20 load... Best to do this dockside so recharging to full is easy.

[conachair]

Originally Posted by Maine Sail

I've tried it and it is not very representative because the discharge curve is not as linear as one would expect. If you want a true capacity test then a C/20 load would be applied until you hit 10.5V then an immediate recharge.

Usually only one series pair or one battery of a parallel bank is required to get a good estimate of bank condition.. As the voltage drops your load changes so if you want really close accuracy keep up on your C/20 to keep it a C/20 load... Best to do this dockside so recharging to full is easy.

That was sort of the point, does it matter if you can't interpolate to the data sheet so long as you get an idea of capacity relative to the last time you did the same test? For many long term cruisers dockside just isn't an option.
I did have a graph earlier of cell voltage/capacity for different loads but can't find it again

[conachair]

Still can't find that graph but came across this...
http://offgridcabin.wordpress.com/20...sg-soc-charts/



EDIT.
Found it



http://autonopedia.org/renewable-ene.../soc-vs-volts/

Sent from my SGP312 using Tapatalk

[Maine Sail]

Originally Posted by conachair

Still can't find that graph but came across this...
batteries :: new SG / SOC charts | offgridcabin

Sent from my SGP312 using Tapatalk

SOC is easy to find. The actual Ah capacity the battery can deliver is what is more difficult....

Bottom line is if your batteries used to be at 12.24V in the mornings, and your overnight loads have not changed, and they are now at 11.95V in the mornings then you most likely have lost capacity...

[conachair]

Originally Posted by Maine Sail

SOC is easy to find. The actual Ah capacity the battery can deliver is what is more difficult....

Bottom line is if your batteries used to be at 12.24V in the mornings, and your overnight loads have not changed, and they are now at 11.95V in the mornings then you most likely have lost capacity...

Again, that's sort the point, doing a regular test to get something much more accurate than noticing the voltage is a bit down in the mornings so you have some sort of handle on how healthy the batteries are, and also to see if an equalising session has helped.

I suspect most cruisers have no way of getting any accurate regular data on what is a vital part of the boat.

Temperature correction might be a source of error though, wonder if it would be better to adjust the cutoff voltage due to temp instead of trying to adjust the results? It does seem fairly linear.






Sent from my SGP312 using Tapatalk

[conachair]

Downside is you might not get anything completely reliable even with the very best kit....



Capacity fluctuations on two identical charge/discharge tests of 91 starter batteries.*The capacities differ +/–15% between Test 1 and Test 2.

http://batteryuniversity.com/learn/a...asure_capacity


Sent from my SGP312 using Tapatalk

[Maine Sail]

I have a few batteries I have been doing long term testing on and 20 hour capacity tests.. These batteries have undergone 20 hour capacity tests each year for the last 5.

The results have been quite repeatable and consistent with minor decreases in capacity as the batteries have aged. I am simply measuring delivered Ah capacity at a C/20 load at a 77F battery temp and my Ah counter is set to cut off voltage at 10.5V. Very simple test... I adjust the load as the voltage falls away to maintain as close to C/20 as I can. It is kind of a PITA and not a test that is conducive to a long term away from shore power cruiser... The batteries (US Battery DCXC Group 31's) are in their 8th season.....

I have not seen large amounts of degradation from conducting yearly capacity test to 10.5V but likely because the batts are then immediately recharged at C/20 to 14.8V and once full are then equalized.

[conachair]

Interesting, thanks.
Looking again at the arduino, should be easy and cheap to log voltage, temperature and amps during the test. Probably with a few power resisters it would be fairly straightforward to maintain a constant current as the voltage drops as well.

https://www.youtube.com/watch?v=lisprJs5sNU

Sent from my SGP312 using Tapatalk

[conachair]

And micro sd card for logging data only a few pounds , aren't more people using these onboard?
http://www.ebay.co.uk/itm/Ethernet-S...p2054897.l4275


Sent from my SGP312 using Tapatalk

[SV THIRD DAY]

Originally Posted by conachair

aren't more people using these onboard?

Sure....and most cruisers that don't live at the dock with shore power also always have enough power...ha ha ha ha....

[Lake-Effect]

Originally Posted by conachair

... any comments on doing a "half" discharge capacity test.

For instance, on a fully charged trojan T105 225Ah 6v have a motorbike headlamp or something to draw around 10A and an accurate method of measuring how long it takes for the voltage to reduce to, say, 12.3v. Then compensate for temperature and hopefully you will have a figure to compare next time, even if it doesn't relate exactly to the Ah data of the battery, at least it's something.

The new-fangled pro digital battery analyzers do something like that - you enter the nominal stats of the battery into the analyzer, the analyzer applies known loads over a set (short) time interval, continuously records the current and voltage over the intervals, then compares the readings to stored nominal data (eg the discharge curves already seen in this thread) and from that, you get a printout of the battery's condition.

I think what MaineSail's driving at is that if you have good batteries, charged correctly and often, avoid deep-discharging them, and keep an eye on the voltage, you will become accustomed to their behaviour and you will notice when their performance starts dropping off, at which time you can either make your own more extensive tests, or have someone with a battery analyzer make some measurements.

It would be a fun project to make an Arduino-based (or MicroChip PICs, my favourite) data-logger... but it could be 4 or 5 years before you have enough data to detect that your battery capacity is starting to drop... at which time we'll all be on 3rd-generation lithium batteries running electric engines, and recharging with cheap Honda fuel-cell generators

(Seeed Studio - one of my favourite sources for micro-computer sensors, parts etc)

[conachair]

Originally Posted by Lake-Effect

I think what MaineSail's driving at is that if you have good batteries, charged correctly and often, avoid deep-discharging them, and keep an eye on the voltage, you will become accustomed to their behaviour and you will notice when their performance starts dropping off, at which time you can either make your own more extensive tests, or have someone with a battery analyzer make some measurements.

Though with the cost of sensors now you should be able to make a unit to test to whatever depth of discharge you want whenever you want while measuring current, voltage and temperature of every cell. And have lots of fun doing it human perception and memory is very flawed, data is much more fun

Sent from my SGP312 using Tapatalk

[OceanPlanet]

Originally Posted by Maine Sail

I have a few batteries I have been doing long term testing on and 20 hour capacity tests.. These batteries have undergone 20 hour capacity tests each year for the last 5.

The results have been quite repeatable and consistent with minor decreases in capacity as the batteries have aged. I am simply measuring delivered Ah capacity at a C/20 load at a 77F battery temp and my Ah counter is set to cut off voltage at 10.5V. Very simple test... I adjust the load as the voltage falls away to maintain as close to C/20 as I can. It is kind of a PITA and not a test that is conducive to a long term away from shore power cruiser... The batteries (US Battery DCXC Group 31's) are in their 8th season.....

I have not seen large amounts of degradation from conducting yearly capacity test to 10.5V but likely because the batts are then immediately recharged at C/20 to 14.8V and once full are then equalized.

Do you mean that monitoring, maintaining, and properly charging Pb batts actually makes them last longer?! Amazing...

[lonesoldier0408]

I'm still waiting for the Capacitor and the Ultra Capacitor technology to catch up and replace the batteries. Twenty years and less weight and expense to produce. Don't expect those financial advantages to be passed on after the word, "Boat," falls from your lips. Just saying.. .