Capacity Testing Methods?

[shaneinseattle]

I've done a fair amount of research and cannot come up with anything firm on how to do this capacity test for my battery bank.

My 12V 540 Ah bank consists of six 6V 180 Ah Gel-Tech batteries and they are over 6 years old but lightly used and on charging almost all the time.

I haven't yet purchased load testers or anything of that nature; want to make sure I get the right equipment. Not sure if I test each battery separately, or can test the bank all at once. Can I test to 50% discharge or must I test to 100% DoD? The questions go on and on...

Any comments from technically astute people that know how to design a test for this size and makeup of battery bank would be MUCH appreciated!

Thanks,

Shane

[Maine Sail]

Originally Posted by shaneinseattle

I've done a fair amount of research and cannot come up with anything firm on how to do this capacity test for my battery bank.

My 12V 540 Ah bank consists of six 6V 180 Ah Gel-Tech batteries and they are over 6 years old but lightly used and on charging almost all the time.

I haven't yet purchased load testers or anything of that nature; want to make sure I get the right equipment. Not sure if I test each battery separately, or can test the bank all at once. Can I test to 50% discharge or must I test to 100% DoD? The questions go on and on...

Any comments from technically astute people that know how to design a test for this size and makeup of battery bank would be MUCH appreciated!

Thanks,

Shane

50% DOD will give you misleading results. For example just yesterday I conducted a test on a pair of 6V golf cart batteries rated at 225Ah. At 12.094V, and under an 11.25A load, (the 20 hour discharge rate) they had delivered 106.5Ah. If I did not know better I could have easily assumed they were on track to come darn close to the factory rating. At the end of the test, 10.5V, they had only delivered 177.2Ah.

The only really accurate test for Ah capacity is to have the batteries at 75-80F, discharge them at the factory rated 20 hour discharge rate (Ah capacity divided by 20) and run the test to a cut off of 10.5V. You can count ampere hours or count hours & minutes and use math to arrive % of new capacity. It is also important to keep the discharge load as steady as possible. I use lab grade 400W programmable DC loads for this and have not seen anything in the "inexpensive" range.

Contrary to popular misconception a yearly or even bi-yearly test to 10.5V followed by an immediate and full recharge is probably less harmful than a few days of PSOC cycling between 50% SOC and 80% SOC. Some batteries even recover some lost capacity by taking them to 10.5V a couple of times and "retraining" the plates as best we can. GEL batteries respond to this better than any other...

I would test the batteries individually this way you will know if you have any out-liars. A DC electronic load is expensive for a DIY but you can make a load bank with switches & resistors to adjust the load and keep it steady as the voltage falls.

[shaneinseattle]

Thanks for the information -- this is good stuff! I'm feeling a bit more confident of planning a capacity test based on your input. For a DC load, I'm considering this load tester from Harbor Freight:

500 Amp Carbon Pile Load Tester

I understand it isn't as precise as lab-quality kit, but with a digital amp meter I think I should be able to keep it close to the 27 amp discharge rates I require for my 540 Ah bank test, knowing that I'll have to tweak it periodically to keep it on target. I think certainly I'll be able to keep it closer than if I were just adding house loads through my inverter and DC devices; I'm not sure I could get the load to 27 amps or keep it there steadily & reliably.

Again, thanks for your input. I will take the V down to 10.5 based on your recent test experience.

[Maine Sail]

Originally Posted by shaneinseattle

Thanks for the information -- this is good stuff! I'm feeling a bit more confident of planning a capacity test based on your input. For a DC load, I'm considering this load tester from Harbor Freight:

500 Amp Carbon Pile Load Tester

I understand it isn't as precise as lab-quality kit, but with a digital amp meter I think I should be able to keep it close to the 27 amp discharge rates I require for my 540 Ah bank test, knowing that I'll have to tweak it periodically to keep it on target. I think certainly I'll be able to keep it closer than if I were just adding house loads through my inverter and DC devices; I'm not sure I could get the load to 27 amps or keep it there steadily & reliably.

Again, thanks for your input. I will take the V down to 10.5 based on your recent test experience.

That won't work as it will overheat and burn itself up. Those are for very short duration testing....

[a64pilot]

The trick is to keep the amps the same, as voltage drops.

I load test aircraft batteries with a landing light, it is not really all that accurate, but will tell you if the battery has enough reserve capacity to run the avionics to get you on the ground in the event of a charging system malfunction, which is the point.
I don't know of a way without spending a lot of $ to keep the amp draw the same for a whole bank, but one battery at a time could be as simple as a big rheostat and a light bulb? Yes you would be adjusting the rheostat every hour or so, but how accurate do you really have to be? Would a simple resistive load and an average amp draw be good enough? Back to the landing light type of low buck answer

[a64pilot]

OK, just brainstorming here, but say for instance in the above example he needs 350W to get his 27 amp draw. would an inverter compensate and keep the amp draw the same once it was established?
What I'm asking is since the inverter will regulate itself to a constant 120V at 60 Hz throughout it's allowable DC voltage range, once you establish the required amp draw with AC resistive devices, like light bulbs, won't that draw remain constant as the batteries discharge?
Of course you can't forget losses in the inverter, but won't those also remain constant?

On edit, output would remain constant, draw from batteries amp wise would have to increase as voltage drops, TANSTAAFL ,never mind.


If batteries were rated in Watt hours instead of Amp hours, we wouldn't have this problem?

[smac999]

get a few 50w or 60w 12v light bulbs. they are a few bucks each. get a 20w or 30w too. then you can mix and match to get correct amps.


from a single battery you'd only need 1 or 2 bulbs. you'd only need 140w for 2 GC batteries or 60w for a group 31.

[smac999]

Maine Sail do you have a device that shuts off load at 10.5v or do you have to watch it?


if so does it also shut down a timer so you know how long it went for? or do you only go by AH on a meter?


I need a better way of testing them then manually watching.

[wsmurdoch]

Originally Posted by smac999

Maine Sail do you have a device that shuts off load at 10.5v or do you have to watch it?


if so does it also shut down a timer so you know how long it went for? or do you only go by AH on a meter?


I need a better way of testing them then manually watching.

See the circuit and discussion on page 145 of the old edition of "The 12 Volt Doctor's Practical Handbook".

http://kb-kbh.dk/shipslib/el_ombord/12volthandbook.pdf

[Maine Sail]

Originally Posted by smac999

Maine Sail do you have a device that shuts off load at 10.5v or do you have to watch it?


if so does it also shut down a timer so you know how long it went for? or do you only go by AH on a meter?


I need a better way of testing them then manually watching.

My DC electronic loads are fully programmable and can count Ah's as well as duration of discharge in hours, minutes & seconds. I can set cut off voltage out to the thousandths of a volt and the discharge current out to the thousandths of an amp. It holds the discharge current very steady almost always within 2-4 thousandths of an amp. Once the batteries are fully charged, equalized and floated for a few days I let them rest for 24 hours then start the discharge capacity testing.

When the electronic load hits 10.499V the load is automatically stopped and the screen retains the Ah's delivered as well as the duration the load ran for until I physically clear the screen... If I'm not there my bench-top power supply is set to turn on at 20:10 minutes, with used batteries, and 20:30 with new or barely broken in batteries. Usually I program it so that I am around and have developed a sense for how many hours the batteries might run, so I can turn on the charger manually, but having it automated is nice too. In the off season I keep the batteries at 77F using a temp controlled water bath. In the summer my shop is usually pretty steady at 74-80F. This time of year I am using the water bath. Oh and the charger I use as well as the electronic load both have dedicated voltage sensing circuits so that voltage is not measure over the charge or discharge wires.

Over the years I used a variation of home made devices using Ah counters, relays resistors and even a PWM dimmer. I then had Mark Grasser build me a custom DC electronic load but it was just not user friendly so I ponied up for a pair of Array 3721A 40A loads and two BK Precision Model 1900 1-16V, 0-60A variable power supplies with dedicated voltage sensing. These things get a work out and barely a day goes by when they are not being used. Today is a large bank of Lifeline 6 volt batteries that are 4 years old and have a LOT of ocean miles on them, three trips from Maine back and forth to the BVI's and probably 20 months of 24/7 on the hook cruising.

This 110Ah rated battery is pretty healthy, and had already delivered 7.68Ah over 1:24 minutes and was still holding 12.766V under the 20 hour discharge rate. If your AGM batteries are not holding voltage like this......


This is what my customers get:



If more marine electricians actually capacity tested batteries it would be a HUGE eye opener and would hold manufacturers feet to the fire. Capacitance testing simply does not work for predicting battery Ah capacity, I own three of these expensive tools and they fall flat..

I often bite my tongue when the discussions of "My batteries are 7 years old and are in perfect health." come up..... Gives me a good chuckle anyway....

[missourisailor]

MaineSail
You use to tout the Balmar Smart Gauge as the end all of monitors.

My understanding of said unit, all you do is hook it up, answer a question or two, and presto, it starts giving you capacity info..

More accurate and easier to use than a Coulomb counter.

No need to do any 20 hour tests.

If this is true, why would any one want to do a full-blown capacity test?

[Maine Sail]

Originally Posted by missourisailor

MaineSail
You use to tout the Balmar Smart Gauge as the end all of monitors.

My understanding of said unit, all you do is hook it up, answer a question or two, and presto, it starts giving you capacity info..

More accurate and easier to use than a Coulomb counter.

No need to do any 20 hour tests.

If this is true, why would any one want to do a full-blown capacity test?

Because you could be at 50% SOC of a bank that is now 54Ah's, or what the bank originally was, and that may have been be 230Ah. It tells you SOC, very accurately, but does not tell you the actual capacity in the bank. Most of the capacity test I do are for expensive GEL or AGM batteries to determine state of health for cruising and so the owner can properly program an Ah counter.

Here is a prime example of the Smart Gauge doing its job at tracking SOC but the owner murdering his batteries.

How to Murder Batteries in Half a Year


.

[Sternwake]

This thread answers a few questions about suspicions I had floating in the back of my Mind.

I now cycle a single group27 90AH Northstar AGM heavily 4 to 5 nights a week, and my Amp hour counter reveals that when I hit the 50% zone usually at a discharge rate exceeding that at which the battery achieved that 90 Ah rating, my voltage, even while still under a light load, is still above the 12.11v that Northstar states as the 50% level.

So I was believing this battery was underrated in capacity, and while I have about 210 Cycles to 50% or deeper on this battery over a 2 year timespan, I've not been able to notice any significantly lower voltage when there are 45+ AH removed from the battery now, than I could on Day one, as long as I actually high amp recharge the thing to 0.5% of capacity at 14.46v. Low and slow solar might get to that 0.5% level, but it does not get there fast enough and performance loss is noticeable when there is no high amp recharge.

I will not be discharging to 10.5v, and do not trust my tools to actually read precisely enough to make any determination of actual remaining capacity, but I will high amp recharge this battery as often as possible, and recharge fully, and keep an eye out for performance degradation, which I determine by voltage held for AH removed and the load it is under at the time.

So far it appears to be still above factory capacity doing the 50% discharge thing and comparing voltage, but I know this is an impossibility.

But this battery is performing significantly better than any previous flooded battery I have owned, but my knowledge and tools to keep previous batteries fully charged was not as extensive as they are now.

So Thanks Mainesail, I always learn something from your posts. I'd love to know the actual remaining capacity of this battery, but I simply cannot perform a 10.5v 100% discharge test accurately enough to give any meaningful answer.

I'll just keep using it hard and charging it fully as often as possible, and see how many total accumulated deep cycles I can get from it

[missourisailor]

Originally Posted by Maine Sail

Because you could be at 50% SOC of a bank that is now 54Ah's, or what the bank originally was, and that may have been be 230Ah. It tells you SOC, very accurately, but does not tell you the actual capacity in the bank. Most of the capacity test I do are for expensive GEL or AGM batteries to determine state of health for cruising and so the owner can properly program an Ah counter.

MaineSail,
As always, excellent job.

I do have a quick question.
When you installed the Smart Gauge, did it or has it ever read SOC of 75% or above?

I must confess, due to your excellent research, writing and recommendations, about two years ago I installed a Smart Gauge in our smaller IP. But as she sits in a slip with StoS power plugged in while we are not on her, her house batts (300aH FLA) were fully charged when the Smart Gauge was installed. As per the Balmar instructions, set the SOC to 75% and went through several complete cycles over a few days.
Rarely have we seen 100%SOC but most of the time, when we get to the boat, it is reading 90%~95%.

Granted, we are not "live aboard's", only week end sailors, and the batt's have plenty of time to fully charge while we are gone. But they are an honest 3 years old and even with a two night outing, very seldom does the Smart Gauge reads below 50%. And this is with a normal house load w/ refrigeration.

For the prior 5 years, I was constantly messing with a very old Coulomb Counter that NEVER told me anything of any use. And this was after doing several controlled discharges to find out true capacity. On both old and new batt's.

For meter heads like you and I (and others of course), doing a controlled discharge is no big deal.
We both have the tooling necessary to do it properly.
But for the average Joe, it is a different story. Not only do they not have the tooling to properly set up a Coulomb Counter, they also don't have the knowledge, will nor desire to keep it calibrated.

I guess what I'm trying to say is...regardless how much true capacity a new or used set of batt's, actually have, I feel it is almost a waste of time, for the average Joe, to determine the true capacity for use with a Coulomb Counter.
Buy a decent batt or set of batt's, keep them properly charged while the boat is not in use and use the Smart Gauge to keep track of capacity.

[smac999]

Originally Posted by Maine Sail

so I ponied up for a pair of Array 3721A 40A loads...

that thing looks sweet. now to convince my boss to pony up


do you normally test single 6v batteries or in pairs at 12v?