Guru
Joined: 13/10/2014
Location: AustraliaPosts: 2199
| Posted: 01:24am 18 Dec 2023 |
I started testing the worst battery when it suddenly failed. The Terminal voltage was correct, but it would not take a load nor a charge. I decided to take that one back to the supplier and see what they thought, with a view to possible get a pro rata warranty as it had been connected to a faulty Active balancer port supplied with the batteries.
I was annoyed because these banks have never gone below 70% SOC and usually charge to 93%. A full top charge carried out every few months. However due to my health issues, they hadn’t been cycled a lot, that is a problem if LiFePO4 is held at anything much above 50% for an extend time. I wondered if that might have been the cause.
While waiting for the prognosis on the returned battery, I decided to pull the 2nd worst battery down, and dam the pro rata, and see exactly what I had been sold - like how many house bricks were added for weight etc.
There was good news and some disappointing news.
Now this battery, and others as I later found, would do things like drop instantly to 12.8v with a 40A load, then drop to 12.4v a few seconds later, sometimes jumping back to 12.8v ? Now that does not sound like a cell problem. Once at the low voltage they would supply that constant 40A load without dropping.
I mentioned above that I found two Active balancers that had partly failed, one had 1 port open and 1 port on another was continually boosting output.
After Removing the Cells from the battery case, I found them in perfect physical condition, completely flat with no swelling, obviously I can’t tell the quality of the cells, and the bar code does not indicate the maker.
The cells were held in place with a poured rubber molding, not a good idea as the batteries were on their sides and the lower terminals were covered by the liquid molding before it hardened. There were no insulating sheets between the battery casings, just relying on the thin plastic covering, not good in a battery aimed at off road and caravan solar.
When dismantled, I found the rubber molding material was still liquid around some terminals. I think capillary action and heat cold cycles due to the poor workmanship in terminal assembly and the terminal thread tapping, the result was over 250mv voltage drop @ 40A between some terminal studs and the plated solid copper links. Most were at least 25mv @ 40A.
One set of batteries had longer link bars cut down “roughly” with a hacksaw and a hole drilled through and bolted on without de-burring. To compound this, the drop across the common line connected BMS was 17mv and the drop across the battery plus lead from the Cell to the external battery plus Terminal was 15mv @ 40A.
Add all that up and you wonder how it even worked a all. Obviously over time the spring washer tensions reduced and the heat cold cycle through what started out as a smallish voltage drop due to poor connections just grew and then snowballed as the terminals got really hot on occasions of sustained current and the resistance from corrosion, melted rubber compound and heat increased.
FYI the BMS was a Daly 100A.
The picture is after I cleaned and de-burred the copper links and slightly rounder off the hacksaw cuts, again, a 40A load was connected across the battery, this time the voltage stayed at 13.2V a big difference to 12.4v. The heat previously generated by the 250ma loss was enough to almost burn my fingers when I touched the Link Bar. The resulting heat is the likely cause of the liquefied rubber compound.
Just the thing you want for long life with a LiFePO4 cell.
