Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

How to quickly estimate lead acid battery capacity ?

Status
Not open for further replies.

azaxi

New Member
Need to quickly estimate capacity of SLA batteries without doing full cycle and without spending hundreds on equipment.

Looking at the discharge curve, fully charged is about 2.25V/cell and fully discharged 1.75V and the curve is fairly linear. In theory, for 6-cell 12V battery, the voltage range between 0% and 100% DOD is approx 3V - 10.5~13.5V. If battery is within this range to begin with, and relatively large load is connected, measure current, voltage, then after 1min voltage drop, approximate capacity in Ah should be I * 3V / (deltaV*60)

Right ?
 
Last edited:
I think that there are too many variables to be able to rely on that, unless you calibrate against some known batteries. After just one minute of discharge, the voltage change will be small, but the battery may be hotter, which will alter the voltage. You will need to measure the voltage very accurately.

If your discharge rate is 1C, then 1 minute is only 1/60th of the charge, so that will be 50 mV (if the 3 V is correct), and you will need to measure the 50 mV change quite accurately, to better than 10 mV. Voltmeters that can measure to 1 mV on a 12 V system are quite expensive.

You will also find that you will need to leave the battery to stabilise for a few minutes after the current draw before the voltage can be measured.

If you discharge the battery faster, the stabilisation will take longer and you have more energy to dissipate.

If the battery has just been charged, the voltage before the test will be affected by "surface charge", mentioned here:-https://batteryuniversity.com/article/bu-804c-acid-stratification-and-surface-charge
 
The voltage will change significantly with load, so both readings need to be taken after the voltage has stabilised in whatever state (charge, open, discharge) you intend to measure.
 
I would guess that weighing the battery would give an indication of capacity. Most of the weight is lead and the more lead the higher the capacity. If this varies between car batteries and deep cycle batteries is another matter.

Mike.
 
Thanks to all.

I only need to estimate if 10yr old SBS190 batteries have 190Ah or 30Ah. Weighing wont work because they all weigh the same, approx 130lbs, and I have no way to know how much of plates are unusable, dried up or permanently sulfated. And since they are sealed with pressure valves, I have no easy way to refill or desulfate and replace new electrolyte so I would just use whatever capacity is remaining but would like to know approximately what it is before I commit to particular units. The batteries have been sitting in a warehouse for a long time, so surface charge will not be a problem.


Currently have one SBS190, which turned out to be disappointment: was expecting 90% original capacity since these are from power backups and were mostly sitting at room temp fully charged without cycling, but I got only 30Ah from it. I planned to build a bank of 8-10 units to capacity 20kWh, dragged this one sample to my off-grid cottage, attached to solar system, scheduled my consumption around expected 2kWh capacity only to find my lights went off after 2hrs with massive beep from the inverter. I did have energy monitor on the battery, so I know exactly how much was pulled from it.

The person who sold it to me has plenty, he says many his customers are using these for solar application and is willing to exchange for another unit. So I did not give up on the idea of 20kWh power bank for under CAD $1,500, maybe just got a bad unit. He just tests crank current and when I was getting this one he pulled 1.1kA from it.

Or if someone can suggest a way to restore the capacity on these. I do not know what is inside, if acid or gel, was thinking of drilling the valves, adding distilled water by syringe to the brim and reseal with tape or small rubber plug. That should replenish water in electrolyte whatever type it is. But that would prevent me to exchange. Will be my last resort if it turns out all his batteries are dried up
If your discharge rate is 1C, then 1 minute is only 1/60th of the charge, so that will be 50 mV (if the 3 V is correct), and you will need to measure the 50 mV change quite accurately, to better than 10 mV. Voltmeters that can measure to 1 mV on a 12 V system are quite expensive.

Was thinking of using INA226, module costs under $5 on Ali and I already have a few on hand. It measures two voltages with 15-bit + sign precision. System voltage up to 36V with LSB 1.25mV, and shunt voltage +-82mV with LSB 2.5uV, which would be 3.3mA on 75mV/100A shunt.
 
Last edited:
If they are not dried out and have not been abused, an electronic "desulfator" may restore some capacity.

The concept is that it electrolyses lead sulphate (which forms on the plates over time) back to lead, with the sulphur being returned to the acid.



I've had mixed success trying one on old UPS batteries that have been on permanent float charge; some recover well, others are unaffected.

There are dozens of types on ebay & they are also quite simple to build from either a kit or just buy the components.


Also on amazon etc.


ps. Pulling high current from a badly sulphated battery may damage it, as only the "clean" area of the plates are highly reactive so the current per plate area is far higher than it should be and that can overheat or warp them.
 
If they are not dried out and have not been abused, an electronic "desulfator" may restore some capacity.

The concept is that it electrolyses lead sulphate (which forms on the plates over time) back to lead, with the sulphur being returned to the acid.



I've had mixed success trying one on old UPS batteries that have been on permanent float charge; some recover well, others are unaffected.

There are dozens of types on ebay & they are also quite simple to build from either a kit or just buy the components.


Also on amazon etc.


ps. Pulling high current from a badly sulphated battery may damage it, as only the "clean" area of the plates are highly reactive so the current per plate area is far higher than it should be and that can overheat or warp them.

I have this charger: https://www.amazon.ca/gp/product/B094VQ88X2

It claims to have desulfation function under "Repair" mode. Perhaps its the same functionality as the devices you shown ?
 
In theory it should do the same, but if it does not stay on that mode for hours or days, a dedicated unit would be better.

(I have a Ctek charger with near identical functions - that seems to just skip that stage quickly regardless of the battery).
 
Follow-up.

No need to build a tester, they already exist. I got one called V311B for C$45 off Amazon. It is meant for automotive purposes and cranking amps, but it also has Chinese standard battery test (GB) that estimates Ah capacity rather than cranking amps.

Tested several batteries. Brand new 24DC marine battery I bought few days ago and is rated at 95Ah is reading 72Ah. I see how this may spook some, but knowing lead-acid batteries have a burn-in, I say close enough.

My old car battery 24F that came with car when I bought it second hand and turned out it was too small for it and failed to start car at one time. Still measuring more CCA and Ah capacity than it should have, see pics.

The SBS190 I mentioned above that failed on me is now showing 95Ah. Before I got the tester I run it through repair cycle on Nexpeak charger so not sure if this is an improvement or a fluke. I will test the real discharge capacity soon.
 

Attachments

  • IMG_20220616_221506_939.jpg
    IMG_20220616_221506_939.jpg
    177.9 KB · Views: 203
  • IMG_20220616_221330_157.jpg
    IMG_20220616_221330_157.jpg
    175.3 KB · Views: 196
Last edited:
Follow up.

V311B battery tester measured my SBS190 capacity at 95Ah. Run a real discharge test with 12VDC/110VAC inverter with 2-2.5A and got 30Ah from it.

Acquired car battery that tested with 9Ah left out of original 95Ah. Run Nexpeak repair overnight and it measures 25Ah now, internal resistance also dropped from 20 to 10mOhm.

It may be interesting to note that when running repair on SBS190, it runs for an hour and finishes. Desulfator must have way to detect sulfation. SBS190 is gel battery and is likely not sulfated but dry, unfortunately I have no way to rehydrate cells. Out of curiosity I took waveforms from repair, its 2 second periodic process, first 1s sweeps from 100kHz to 20Khz range and other 1s then stays in 20kHz range.

Nexpeak (and others) may be following expired PulseTech patent. Pulsetech makes claims in their promo video that their PowerPulse device extends battery life up to 3x by continuously breaking up forming crystals that are easier to break when small rather than when there is huge buildup in dead batteries.

There are some simplistic NE555 implementations as rjenkinsgb pointed out, but this study shows overdesulfation may lead to capacity decrease! Intelligent desulfators are attach and forget.

UPDATE: got cleaner waveforms from another battery. First pic is DC coupled and shows how charger charges for 1sec and then lets go (or discharges?) for another 1 sec. Rest of the waveforms are AC coupled otherwise I would be unable to zoom voltage. Pulse frequency sweep happens during charging phase, then its static frequency during discharging phase.
 

Attachments

  • PulseTechVideo2qTzKwujoW8-1.jpg
    PulseTechVideo2qTzKwujoW8-1.jpg
    93 KB · Views: 186
  • PulseTechVideo2qTzKwujoW8-2.jpg
    PulseTechVideo2qTzKwujoW8-2.jpg
    84.5 KB · Views: 167
  • 20220620-B101-repair1.jpg
    20220620-B101-repair1.jpg
    180.8 KB · Views: 185
  • 20220620-B101-repair2.jpg
    20220620-B101-repair2.jpg
    172.7 KB · Views: 168
  • 20220620-B101-repair3.jpg
    20220620-B101-repair3.jpg
    210.8 KB · Views: 180
  • 20220620-B101-repair4.gif
    20220620-B101-repair4.gif
    1 MB · Views: 178
Last edited:
Interesting!

The pulsetech patent appears to be nothing more than a kind of PWM regulation to maintain float charge voltage, there is no high voltage generator for desulfation as far as I can see. I think they are claiming the rapid rise time has such an effect, but I've never seen that documented before.

(There were some functionally rather similar pulsing charge circuits published in electronics mags the 1970s, I built one for charging NiCd cells).

The voltage waveforms you recorded also appear to be very low voltage, if the scope display is accurate - 50mV per division?

The study you link to does say that the authors do not know the cause of the degradation, and it could just be from their discharge capacity testing..

I've never used mine for more than a couple of weeks on one battery, and only on those that have already failed in service. I'm rather wary of using something capable of giving high voltage pulses to any electronics the batteries are intended to operate with!

I even have ferrite sleeves with a few turns on each, on the trickle charger output leads, to provide some isolation from the high frequency pulses.
 
Out of curiosity, I just tried to have a look at the circuit of the one I have; it turns out the thing is rather well encapsulated and cannot be disassembled without destroying it; I could pry out the pressed in end panel, but that is as far as I can go... [The end panel pressed back in without damage after the photo].

What can be seen is quite impressive though; its definitely not an empty box or lightweight tiny component version!
There is a large inductor looking something like an E-core transformer behind the LEDs, a 3A or 5A diode behind the connecting leads and either a decent size electrolytic or cylindrical inductor behind that - just in the first third or so of the length of the device.

It's one of these, though I got mine from ebay probably ten-fifteen years ago:


Desulfator_Internal.jpg
 
The voltage waveforms you recorded also appear to be very low voltage, if the scope display is accurate - 50mV per division?

Scope sensitivity is correct, probe was 1:1 with AC coupling. I captured cleaner waveforms, check post #10 again.

In last 3 days was able to partially recover old 95Ah battery from around 12Ah to current 32Ah using this pulse repair charger. Disadvantage - it refuses to pulse without being plugged to mains, and shuts off when battery is full. I check it once or twice a day and if its stopped, just restart repair process.

About my original question - estimating capacity. For now I am happy with V311B. Even though shown capacity differs a lot from real capacity, it provides means to compare various batteries.

The SBS190 has real capacity about 30Ah @ C/10. V311B standard GB test shows capacity around 110Ah. When I was doing 1.5hr discharge test, measured voltage drop 12.90V -> 12.83V at average current 1.78A and battery monitor shows 2.67Ah. Using my suggested linearized method in post#1 I would get 51Ah, which is incorrect. Turns out, curves are not as linear as I thought (see page 2 here). If I were to build the estimator, I would program these curves as tables and looked up battery operation point.

Anyway, my quest to estimate capacity is done for now. Thanks everyone, especially rjenkinsgb for being so supportive. In the future I would like to discuss desulfation, and reviving LA batteries in general, but I need to gather more data and it would be another thread.
LA-SOC.png
 
Last edited:
Need to quickly estimate capacity of SLA batteries without doing full cycle and without spending hundreds on equipment.

Looking at the discharge curve, fully charged is about 2.25V/cell and fully discharged 1.75V and the curve is fairly linear. In theory, for 6-cell 12V battery, the voltage range between 0% and 100% DOD is approx 3V - 10.5~13.5V. If battery is within this range to begin with, and relatively large load is connected, measure current, voltage, then after 1min voltage drop, approximate capacity in Ah should be I * 3V / (deltaV*60)

Right ?
can't be without kits
while discharging, here yours 1 min, must be kept read how many Amp current is, it coulde be still, ie. no V decrease yet, or decrease ie. on V decrease. So ensure write it down the time range of each A value

IMHO your right, but 'large' if it means low RL, should be the other way, have hi RL, to get low Amp discharge, just do it with proper measuring tool so might get rough quite approximation
 
I designed a method for capacity estimation.

It uses State Of Charge table that correlates discharge voltage to SOC

The method requires load, voltmeter and battery energy monitor. It uses the SOC table and linear approximation between two closest points to evaluate exact SOC at given two voltages.

I have discharged one battery and took several points 11.87V, 11.75V, 11.64V and 11.46. The yellow fields are entered values, cyan are calculated. "Discharged Energy" is measured by energy monitor and then divided by delta SOC.

Energy monitor can be substituted by stop watch and average current/voltages: E = (I1+I2)/2 * (V1+V2)/2 * time

As you can see the estimated values between measurement points slightly differ, but they are not massively different.

Tried to attach the spreadsheet in attachment, but forum wont accept it.

BatteryEstimator.png
 

Attachments

  • BatteryCapacityEstimator.zip
    11.9 KB · Views: 176
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top