Absorbtion Time for AGMs

[conachair]

Originally Posted by sailinglegend

I have done 20 hour capacity tests on my five 210 Ah AGMs and they are all at about 55% of original capacity.

Interesting, always a problem onboard - to test how to you maintain constant temperature (25degC) and constant current C/20 pls ?

[sailinglegend]

Originally Posted by conachair

Interesting, always a problem onboard - to test how to you maintain constant temperature (25degC) and constant current C/20 pls ?

You can't maintain a constant temperature - and its probably not necessary if the batteries are in a constant temperature environment.
Maintaining a constant current for 20 hours is a huge problem as you have to do each battery separately - so 5 batteries take me 5 nights!!!
If you start an hour before you go to bed and check before you go to sleep and turn on extra loads if you have to. If you get up in the night you can keep adding small loads, but if in the morning your 200Ah battery with a 10 amp original load has gone from 10 amps to 8 amps then over that period was on average 9 amps. In the day time you keep the load back at 10 amps until the voltage falls to 10.5 volts. You might then say the average amp draw was 9.5 amps not 10.

But if you test falls to 10.5 volts well before the 20 hour test is finished then your batteries have losts capacity. So if it finishes at 18 hours your battery capacity would appear to be 18X9.5 =171Ah not 200Ah. But if this is the case the original test load should have been 171/20=8.55 amps not 10 amps. But does this real matter - I don't think so! The next time you do the test estimate a lower capacity and set the load accordingly.

[conachair]

Originally Posted by sailinglegend

You can't maintain a constant temperature - and its probably not necessary if the batteries are in a constant temperature environment.
Maintaining a constant current for 20 hours is a huge problem as you have to do each battery separately - so 5 batteries take me 5 nights!!!
If you start an hour before you go to bed and check before you go to sleep and turn on extra loads if you have to. If you get up in the night you can keep adding small loads, but if in the morning your 200Ah battery with a 10 amp original load has gone from 10 amps to 8 amps then over that period was on average 9 amps. In the day time you keep the load back at 10 amps until the voltage falls to 10.5 volts. You might then say the average amp draw was 9.5 amps not 10.

But if you test falls to 10.5 volts well before the 20 hour test is finished then your batteries have losts capacity. So if it finishes at 18 hours your battery capacity would appear to be 18X9.5 =171Ah not 200Ah. But if this is the case the original test load should have been 171/20=8.55 amps not 10 amps. But does this real matter - I don't think so! The next time you do the test estimate a lower capacity and set the load accordingly.

That sounds like it may indicate any change from the last test if temperature was identical and loads were changed at identical times, but may be well off from manufacturers figures.

No data logging? How do you know what changes from one test to the next?

Probably not so important in the real world but doubt if you're getting a real resolution of capacity down to a few percent of manufacturers figures.

Maine sails site has some posts about how involved getting accurate capacity figures actually is, and the test gear required.

In our world it might well not matter much if you can replicate the conditions of each test and accurately log the data to compare. Raspberry pi/arduino can to that.

[john61ct]

Originally Posted by conachair

Raspberry pi/arduino can to that.

Greatly appreciate if you could link to relevant HowTo's, GitHub FOSS projects etc.

[Maine Sail]

Originally Posted by conachair

Maine sails site has some posts about how involved getting accurate capacity figures actually is, and the test gear required.

You're really just trying to get a feel for where you stand in relation to the factory rating of the battery. Light bulbs, resistors and even a PWM dimmer dialing the load in can be "accurate enough" for a guide to SOH.

You don't really need to data log anything just count time and amperage. Any iPad or phone can be paired with a time-lapse or intervalometer app to snap a shot every minute or 2 minutes etc.. You can then track the timer and voltage, after the fact, for what ever duration you set it for. Pretty much anywhere between about 72F and 82F is going to get you a decent ballpark. I have a time-lapse/intervalometer app I use called "Oh Snap"..

Before capacity testing any bank I do advise at least one or two back to back deep discharges, to 20% SOC +/-, with full recharges in-between each deep discharge.

The industry guideline for "End of Life" is when a battery can no longer deliver more than 80% of it's original capacity rating. Can they go beyond this? Sure they can, especially with light house loads, but the risk of failure is increasing the lower you go. For RV's or coastal cruising it's not as big a deal if the bank fails than it is in the middle of the North Atlantic..

With Lifeline, Odyssey, Northstar etc. I will personally recommend replacement for coastal users at about 65% and for offshore work 70-75%. I used to recommend 65% across the board but a few suden failures have caused me to up that to 70-75% of new capacity..

I have had a number of AGM batteries, including Lifeline, drop off the proverbial cliff once they get into the 70% of rated capacity range or lower.

Somewhere I think I have an interesting string of text messages from one customer who lost his house bank well off shore (18 hours out of Beaufort heading for Tortola). They hand steered to Marsh Harbor and the Lifeline batteries were 4X the cost of what I had quoted him three weeks prior. This quote was after finishing a capacity test of his bank, and knowing he was heading straight for Tortola. The bank tested in the high 60's to 70% range (can't look it up right now).. I had strongly urged him to replace the bank but the next time I heard from him was via text after the bank failed. He got 7 years of use out of them but insisted on pushing them just a bit more. Sometimes capacity testing can just be expensive and meaningless if the owner does not choose to pay attention to the results.....

[john61ct]

Originally Posted by Maine Sail

intervalometer app to snap a shot every minute or 2 minutes etc...

Love learning new words!

Lapse It Pro, on Android

And thanks for the real-life guidance on lifespan.

Since I'd replace a bank whole, what if I did a 20- hour at new, then benchmarked against that?

Save doing each cell separately?

[a64pilot]

During annual inspection of an airplane, in order to remain airworthy the battery has to be capacity tested.
I have done it with a landing light, now we are not doing a real 20 hr test, but a landing light will tell you if you lose your alternator in the clouds if you have enough battery to operate navigation and communications equipment to get on the ground or not.

It is probably not the best way, but I'm hanging my hat on the Smart Gauge, if it is accurate as it is supposed to be, I will notice that my nightly average of 15% capacity used will slowly become 20% then 25% as my bank loses capacity

[conachair]

Originally Posted by john61ct

Greatly appreciate if you could link to relevant HowTo's, GitHub FOSS projects etc.

Sorry, don't have much to hand right now but this is from an arduino logging a fixed load on a couple of very tired US batt 6v a while ago. Cost was a few beers and lots of hours on google getting the code working.

Google battery tester arduino or something like that should come up with some instructables and other links hopefully.

[conachair]

Originally Posted by Maine Sail

You're really just trying to get a feel for where you stand in relation to the factory rating of the battery. Light bulbs, resistors and even a PWM dimmer dialing the load in can be "accurate enough" for a guide to SOH. .

Completely agree (Of course )

The post was more in response to the idea that the 55% of new capacity was actually assumed to be accurate - so many variables.

[sailinglegend]

Originally Posted by conachair

No data logging? How do you know what changes from one test to the next?.....

Good old fashioned graph paper - one for each battery and then use a different colour pen next time. 12 months from my last test two batteries were about the same and the other three were down. If there's a big change in the voltage you don't even need to complete the test - you know you're in trouble.

A simpler way is to run in new batteries first and do a 10 hour test plotting on graph paper which will take you down to about 50% - this figure may not be 50% because the fall in voltage is not linear and depends on the battery age. Each year repeat the test and plot the difference.