Sunny Island Logic Madness
SpikeandKari Registered Users Posts: 7 ✭✭
edited April 2022 in Construction #1
Hello! Installation details listed at the end and am prepared for all the criticism as to having chosen an AC coupled system for Off Gid but there you go. I have had this system for 5 years now and am now replacing the first set of batteries which were Rolls AGM 270 AH at 48 Volts. Not lasted as long as I would have wished so watched new battery charging very carefully. What I seem to have found is that the logic that SMA use to start a Full Charge or Equalisation seems nuts. So, anybody out there got experience which indicates my observations are wrong?
So Sunny Island is meant to do a Float, Boost, Full and Equalisation.
Voltages can be programmed for all 4.
Times can be programmed for all 3 that are not Float with return to Float at the end.
When state of charge drops to 70% Float ends and Boost starts.
Unless the Timer for Full or Equalisation is zero in which case Full or Equalisation starts.
So far, so good BUT
If timer for Full or Equalisation finds a Float is in progress when it ticks down to zero it waits just one 'tick' to see if Float ends. The tick is 2.4 (so one decimal day) long and then it resets.
If timer for Full or Equalisation finds Boost is in progress when it ticks down to zero it just resets.
So this seems to set an unrealistic set of conditions to start a Full or Equalisation charge so it just never happens.
I don't want my batteries to never have a full charge. I also don't want to overcharge them. So I want a regime of charging to say 90% daily and 100% once a week. How do I get SMA to do that?
Thanks in Advance, Mark
3KW of panels
Sunny Boy producing 220 V AC
6KW Sunny Island
500 AH , 24 cells at 2V
0 · Share on Twitter
I know pretty much next to nothing about the Sunnyboy charging setup.
So, lets go back to your Rolls AGM batteries. Are your new batteries also Rolls AGM? If so, here are the Rolls charging specifications:
The Rolls' charging voltage of 29.4 volts (or 14.7 volts for 12 volt bank) seems to be pretty aggressive. Most AGM batteries seem to use 14.4 volts/28.8 volts for charging voltage. But this is Rolls' specs and they are battery engineers--So I would start with that.
For charging time, Rolls has their Absorb Time equation (basically charging from 80% full when Absorb voltage is "reached", to 100% full). At the point that absorb is reached, the battery bank starts to limit charging current (from rated/available charging current at 80% SoC) to less than 1% rate of charge (as reading 100% SoC)... More or less, the ideal way to know when a sealed battery is fully charged is when the charging current falls to 1% or less (AGM can be down to 0.1% or less rate of charge--Very low self discharge current).
And with "variable" solar charging current---Hitting abosrb voltage (28.4 volts at 25C) and holding until charging current falls below 1% of bank AH capacity (or 500 AH @ 20 hour rate * 0.01 rate of charge = 5 Amps for "tail" of charging curve) is the ideal charging cutoff point.
I don't know if the SunnyBoy/SunnyIsland has a shunt (current measuring resistor) for the battery bank--But you can get a battery monitor to measure/display present battery current for you to do some quick monitoring.
These are US links--I am sure you can use Amazon.UK or other sites more local for you to search.
The Amazon stuff--There is so much to choose from and cheap enough to try and see if works for you or not...
Be careful, there are some very nice DC Amp*Hour and Watt*Hour metering systems out there... Some may not "recognize" battery Charging and Discharging (add to capacity when charging, subtract from capacity when discharging)... They are pefectly fine meters, but not true Battery Monitor Systems.
What you don't want is the battery bank to overheat while charging and/or vent (gassing H2 and O2) while charging. Venting is a permanent loss of water--And since AGM batteries are sealed, there is no way to add (distilled) water back to the bank.
Rolls does not list an EQ voltage (which can be 15.0-16.0+ VDC for flooded cell lead acid batteries).
[10% min, 20% recommended, 30% max]
t = 0.38 x (C/I)
t = Absorption Charge Time (Hours)
C = 20 hr Rated Capacity (AH) [ex: 2 strings x S6-460AGM-RE models (415 AH) = 830 AH rated capacity]
I = Charging Current (Amps) [charger output min 10% up to max 30% of 20 hr rate]
2 strings of S6-460AGM batteries
20 hr rate = 415 AH x (2 strings) = 830 AH
I =20% of 830 AH = 166 Amps OR If the charger output limit is 120 Amps max, then 120 is used
T = 0.38 x 830/166 = 1.9 hrs OR T = 0.38 x 830/120 = 2.63 hrsTo figure out "absorb time", need the size of your solar array...
However, as we all know, with solar panels, you do not get "Charging Current" at XXX Amps for 24 hours per day... Instead, we get 0-77% of Panel rating for charging current.
More or less-You need to hold Absorb Voltage for probably a full day of charging after a night (and/or previous day) of charging for pretty much the whole sunny day (or what passes for a sunny day in UK?).
And only drop to float voltage if the bank is not cycled (i.e., you are there on weekends, and no loads during week, etc.).
There is an AGM/Sealed battery "EQ" that many users have suggested. That is to hold the Absorb Voltage for 8-24 hours to "soft EQ" the sealed battery bank. Do this once every ~6 months.
When to EQ... For AGM 2 volt cells... Ideally they should all have the same resting voltage. If you see the resting voltage from high to low cells reading >~0.030 volt difference)--Then would be a good time to "soft EQ" the bank. EQ should be long enough to bring the difference, ideally, to 0.0 volts between cells... But the reality is simply soft EQ until all cells stop rising in voltage--And those voltages (and differences) are your new 100% SoC values.
By the way, do you know why (or how) your first AGM bank failed? From daily cycled "good AGM" in 25C or less ambient temperature "climates", getting 7 years from the bank would seem to be "not bad". 5 Years, not great, but if heavily cycled, or not correctly charged--Certainly possible.
-Bill "not a battery engineer" B.