Originally Posted by JohnnyFry
Wanting a load close to C20. I used the fridge on DC for 14.9 amps...about C15.
The good news was that the battery held up fairly well to the 50% discharge level.
So my question to the group: while it is clear that I am OK drawing down to 50% what conclusion can be drawn about the overall condition of the batteries considering the differences in apparent discharge percentage between the actual Amp Hour readings compared to the SOC VS Voltage comparisons to the Trojan chart?
We agree with markopolo that the ‘load condition’ (or recency of load connection) is hugely important to any measure of state of charge (done other than by ampere/coulomb counting). Let’s take this a little further.
You indicated that your battery “held up fairly well to 50% discharge level.” Did it? How do we know? How do we know that you still had 50% left? That 50% determination was a ‘calculation’ as much as a ‘measurement’ and it was a calculation that your SoC meter made based on your previous programming - - you told your SoC meter that you had 232 ah batteries. What would that meter have said if you switched batteries . . . connected, instead, a 116 ah battery? It would have said the same thing based on its programming. Although your 132 ah battery would have, in fact, been completely discharged, your SoC meter would have reported a successful discharge to 50%.
The same would be true for your 4 year old batteries if they had, over that period, deteriorated. All we know for certain is that they’re still good for at least 50%.
We decided after initial installation of our system to run (and record progressive data on) a full discharge cycle with the goal being to confirm the capacity of the battery and, secondly, to create a ‘starting spot’ record against which we will be able to compare future capacity tests.
Now it is true that we’re running lithium and we are informed that we can safely fully discharge these batteries. But for the purposes of a very occasional test, we’d argue that you, too, can fully discharge your lead-acid batteries. You could trust your SoC meter and only discharge to 50% as you did in your recent test. But then, as noted above, you really don’t know if you have 50% left - - particularly when your charts are telling you that you are well below the 50% discharge level. So we’d run a full discharge cycle.
We, too, needed a ‘load’ to run our test. We happened to have a spool of several thousand feet of #18 wire and discovered that something in the order of 130' represented a 10 amp load at 13 volts. We trimmed and cut to get a fairly accurate 10 amp load. (The actual resistance of our length of trimmed wire actually changes enough over temperature that we can see small current changes as the wire gets hotter. We use this as a ‘feature’ to tweak our load . . . unrolling, fully or partially, to cool [increase the current] and visa versa.)
We have a 500 ah pack, so we run 5 or 10 hour discharge cycles for corresponding 10% or 20% discharge increments. After each cycle, we disconnect all charging and discharging (including our BMS) and let the batteries rest 10 hours. We have found that it takes several hours for the batteries to reach a steady-state voltage so the 10 hour interval is longer than necessary but assures a real ‘resting’ measurement.
Our three tests to date have netted the following full capacity measurement: 495, 509, and 518 ahs (interestingly, the 518 is the most recent).
As an interesting aside for those dabbling in lithium, one of our tests revealed a resting battery voltage of 13.36 volts represented both an 80% and 90% SoC while a second test netted 13.36 volts for 80% SoC and 13.37 volts for 90%. Either way, you can see why they say lithiums have a very flat voltage vs SoC curve. This 13.36 volt measurement is what we have adopted as our target ‘go to’ charging voltage for our lithium system - - wanting to keep it charged to something in the range of 80-90%.