battery capacity testing

[sailorboy1]

My house batteries are 4 years old and I want to capacity test them this Spring. But I don't really want to discharge them all the way down to 10.5V and don't really have a practical way to be at the boat for 20 hours to do it.

If I instead just discharged them halfway to around 12.2V, let them settle a while after, and then measure the voltage and SG and ratio it back out to full, how far off would the measurement be you think.

I don't really care if I completely know the real capacity as long as I can be within 5-10% to use in programming my battery monitor so that I have more faith in the SOC measurement display.

[smac999]

you don't need to be on the boat for 2o hours. you just need to be back ~15 or so after you start. or sooner if you think they are bad and won't last 15.


start it late on a Friday night, come back and hang out sat morning.


12.2v is nowhere near halfway under a c20 load. that's still pretty full.

[sailorboy1]

Where's all the normal expertize, this is your chance to shine.

[btrayfors]

It won't work, for several reasons.

There are only two ways I know of to test health of a LA battery (i.e., remaining capacity):

1. a 20-hour test such as you mentioned, in which the battery is fully charged (100%) and a C/20 resistive load is applied; and

2. a resistance/conductance test using a sophisticated (and expensive) tester such as the Midtronics series testers.

I favor the latter since I made the investment ($600) in such a tester a few years ago, and am familiar with it's strengths and weaknesses.

And, I've done lots of tests of the first type in the course of several research projects, and have had occasion to compare results with those obtained with the second type tester.

Bill

[Maine Sail]

Originally Posted by btrayfors

It won't work, for several reasons.

There are only two ways I know of to test health of a LA battery (i.e., remaining capacity):

1. a 20-hour test such as you mentioned, in which the battery is fully charged (100%) and a C/20 resistive load is applied; and

2. a resistance/conductance test using a sophisticated (and expensive) tester such as the Midtronics series testers.

I favor the latter since I made the investment ($600) in such a tester a few years ago, and am familiar with it's strengths and weaknesses.

And, I've done lots of tests of the first type in the course of several research projects, and have had occasion to compare results with those obtained with the second type tester.

Bill

Bill,

Your Midtronics lines up to actual 20 hour test results...??? I have had to re-learn how to interpret mine after I began offering C/20 testing to customers.... I also have many, many hours on the phone with Midtronics engineers and they just can't do what we would like, which is to conduct a fast test that tells you Ah capacity. Apparently we are not the only ones begging for this the off grid guys are too... They can create good guidance, if you start with a baseline, but the results don't compare accurately, on any of my conductance testers, to a 20 hour test. My EXP-1000 model and the Argus are the best approximations...

[btrayfors]

Maine,

Yes, you have to interpret the readings from the conductance testers. They don't give you a direct reading of AH.

I tend to develop baselines for each type of battery by testing those which are nearly new or just broken in with several charge-discharge-charge cycles, and interpret results against these.

Certainly agree that it would be nice to have a tester which could rapidly and easily give you a close approximation of remaining capacity.

Another thing with the Midtronics testers: you have to be careful to get the clamps very tight in order to get repeatable results.

Had a recent experience which may be of interest. I have four APC UPS650 units in my house for our computers, cable modems/routers, etc. Just replaced all four batteries after the one on my own computer started giving evidence of failure.

The four batteries were all 12V 12AH AGMs of the same approximate age: 3 to 3.5 years old, and all were in 24/7 operation. The one in my computer was the worst, measuring only 10.5 volts after removal from the UPS650.

The other 3 all measured 13 volts or so after resting for 24 hours after removal, but had very different capacities according to the Midtronics tester, viz:

1. (mine) "bad cell"
2. 131 CA made in Taiwan
3. 101 CA made in China
4. 89 CA made in China

The last three were all PowerSonic batteries of the same model. The worst was a PowerWerk battery (China) from Batteries Plus.

Bottom line: four 12V 12AH AGM batteries of the approximate same age, in identical
applications with identical use and charging setup (in the APC UPS650), with very different outcomes after less than 4 years operation.

Go figure :-)

Bill

[Maine Sail]

As a follow up to my last post I had a battery that recently underwent full capacity testing so I ran all three impedance testers on it and compared them with one of the capacity testers I use..

Battery = 100Ah AGM - 575 CCA

Array 40A/400W DC Electronic Constant Load =71.28 Ah's at 77F & 5A Constant Current. Hours = 14:15

Midtronics EXP-1000 = 533 CCA or 92.7% of rated cranking capacity

Argus 500 = 537 CCA or 93.4% of rated cranking capacity

Centech (Chinese) = 640 CCA or 111.3% of rated cranking capacity


This battery looks pretty healthy based on three impedance testers but the C/20 tests tells a very different story... If you can learn to interpret the Midtronics testers they are great field tools. Unfortunately I can't take them a face value for programming things such as a battery monitor.