It has been a little, while but I’ve collected enough data to assemble another battery update post. The first topic I’d like to cover is battery capacity and changes in temperature. I’ve seen theoretical charts for capacity vs. temperature, but due to variables in battery composition, size, manufacture process, etc… I needed to experimentally find the relationship with our batteries and temperature. I have collected only three samples of capacity and temperature but I consider these to be very good samples given how much trouble I endured to start from a fully charged, rested battery, with a reasonably stable 20-30amp draw (TV and a few lights with the occasional power spike from the well-pump). I used to battery monitor to record how much power was used (Ah) and then took corresponding specific gravity readings to get an estimate. I used specific gravity before and after discharging to determine the change in state of charge.
What’s very interesting about this chart (other than dropping to 50% rated capacity at about 30°F) is how quickly capacity drops off when temperature drops below 60°F. I now consider “Summer” as the season when the batteries are >60°F and “Winter” to be anything below that. It would have been nice to keep the batteries in the basement, but in our situation an insulated garage works. I simply adjust the battery monitor to 850Ah in winter and then back to 1540Ah in summer.
I’ve begun to track the to amount of water I add to the batteries. I had a brief scare a while back when the plates were just barely exposed on a few batteries. The PV charger was running a bit high for voltage – I thought the low capacity of the batteries in the cold was due to not getting enough charge to the batteries so I jacked up the charger set-points. While this did not risk harm to the batteries directly it cooked off the water pretty quick when the weather got warm. Combine a particularly busy time in my life (graduation, first child, moving, new job…) and we were close to doing some real damage to the batteries. Luckily I had a weekend in June to give the batteries the once over. I also recently started tracking equalization dates. The new plan for maintenance is:
- Water as needed,
- Visually inspect every 4 months
- Equalize every 4 months
- Take S.G. of every cell yearly (once before, once after equalization)
The most recent battery health assessment was interesting. It’s been a long time since the last equalization. I may have record of it somewhere but it’s not in either of my two core battery spreadsheets. Here is the S.G. before equalization. Note that the batteries were at 87% charge.
How to read my tables:
- Each battery is 6V and has 3 cells
- One temperature reading is taken for each battery
- Top table is actual readings
- Bottom chart is the variance from average S.G. and standard temperature
- My algorithms flag variances: white is normal, brown is OK, red is greater than ±0.007 from average – which is generally considered an indication to equalize
Here are the effects of equalization (before watering batteries)
I’m happy with this. One cell is just outside of factory specification. I’m not going to worry much since the batteries were cold and the quality of an equalization could be considered questionable because of that. I’m fairly sure a summer equalization would do the trick and leave my tables with all white cells (maybe a few brown ones).
It looks like at 3 years our battery bank is still fairly healthy and that having a maintenance plan is going to pay off long-term. Also, I’m almost ready to declare the off-grid system a success, and stop referring to it as an experiment. I’ll decide that at the five year mark (early 2015). It also looks like the batteries require about 1L water each month and that anything over 5L puts the batteries at risk for exposed plates. This will probably drop to 1/2L water each month in winter due to temperature drop.