SMA 6048 rec-bms SOC wrong problem

[westvandude]

I have two SMA 6048 sunny islands off a 24.5kwh nissan leaf pack gen2 modules. Anyhow, everything was working fine until we had a power outage 14hrs long and the system ran for about 7hours and quit depleting the battery. I didn't have the ABB disconnect relay/breaker wired up for disconnect by the bms but the bms did shut down the sunny islands with it's RS485 connection. When I measured pack voltage it was about 47.4v which isn't crazy low.

Anyhow, I stuck a small 5A 60v lab power supply to the pack for an hour and brought the voltage up slightly so that the bms was not flashing red led and beeping alarm.

I restarted the sunny islands as I had to power my laser printer briefly to get an invoice printed. Had a generator running separately for fridge/freezers (not connected to my system). At this point the sunny islands showed 1% SOC.

Yes... I know, I need to wire up a transfer switch and connect DigIn on the islands, etc and possibly get a larger generator or diesel etc. Also need to connect one of my 41series 6kw gridtie inverters to the secure loads panel as well as changing settings on the islands and the gridtie to offgrid. I've got the grid code to do that. Boy these things have way too many settings lol.

Once power came on the SOC went up to 12% rather slowly. Eventually landing on 14%. It took me a while to get my old laptop going with the rec-bms software application and then guess at the COM settings and all the junk (drives me nuts that it's not intuitive!!).

So I first thought why the heck is it stuck at 14% SOC after like 6hrs of on-grid recharging? Once I got into the bms and checked voltages, etc everything was charged up to my orig settings of 3.9v per cell and I upped the END of Charging voltage from 3.9 to 4.0 and eventually back to 3.95 to watch the charge current. It indeed did ramp up to 50A and then down to 20A. So my settings in the sunny islands are essentially 60A or whatever is the higher setting for charging current ongrid. I had thought that the sunny islands were limiting charge current. That wasn't the case.

So at the end of the day, the sunny islands and the bms are working correctly, except the SOC is not showing 100% and I'm not sure how to fix that. I've emailed REC. Lets see what they say.

Figured I'd post this up just because I haven't seen this issue posted anywhere before. Up until this point the pack was always floating at 100% SOC except for one power outage that was brief maybe an hour, which didn't screw up the SOC reading.

On a side-note, one of my kitchen lighting breakers is an arc-fault unit and that has gotten real flaky since running straight off my generator (which wasn't holding rpm super accurately - stupid carburetor!) I'm thinking the generator may have pooched it. Personally I think the arc-fault breakers and gfci's as well as child-proof outlets that push your plugs out over time are straight up TRASH.

 

[toms]

It will reset itself to 100% SOC, otherwise you can manually set SOC in the REC to 100%, follow instructions in the manual, from memory it is command SOCS 1.00

 

[sunshine_eggo]

A few things... (I know nothing about the REC)

Not sure it matters, but IIRC, the Leaf modules are NMC, though they may be NMC/LMO.

Have you tested your battery for capacity? If you've programmed your battery capacity based on rated, it may be off.

In your image...

With high cell at 3.912 and low cell at 3.826, that's a pretty large deviation.

If you're only charging to 3.95V/cell, you're in the 75-80% range, not 100%. To get to true 100%, you need to go to 4.20V. If you've factored that in, i.e., true 80% is 100% and you're only using 80%, that's your choice.

Incorrect SoC issues usually result from errors in programming.

 

[westvandude]

A few things... (I know nothing about the REC)

Not sure it matters, but IIRC, the Leaf modules are NMC, though they may be NMC/LMO.

Have you tested your battery for capacity? If you've programmed your battery capacity based on rated, it may be off.

In your image...

With high cell at 3.912 and low cell at 3.826, that's a pretty large deviation.

If you're only charging to 3.95V/cell, you're in the 75-80% range, not 100%. To get to true 100%, you need to go to 4.20V. If you've factored that in, i.e., true 80% is 100% and you're only using 80%, that's your choice.

Incorrect SoC issues usually result from errors in programming.

Thanks for the info. Yes I purposely didn’t set voltage to 4.2 because 80% or so is better long term lifetime. Actually for backup not daily cycling it’s super relaxing easy on the cells vs an electric car (depleted daily perhaps). I guesstimated capacity somewhat lower than theoretical. But no, I didn’t test actual cell capacity. It was a low kms car and charged fairly high voltage when I got the pack from the auto wrecker many years ago.

 

[sunshine_eggo]

Thanks for the info. Yes I purposely didn’t set voltage to 4.2 because 80% or so is better long term lifetime. Actually for backup not daily cycling it’s super relaxing easy on the cells vs an electric car (depleted daily perhaps). I guesstimated capacity somewhat lower than theoretical. But no, I didn’t test actual cell capacity. It was a low kms car and charged fairly high voltage when I got the pack from the auto wrecker many years ago.

I'm running a 450Ah (tested)/14S DIY NMC/LMO bank built from 294 Ford Energi PHEV cells. 3.92V is supposed to be something of a magical happy place for these cells and is 75-80% based on my actual tests. I've assigned an arbitrary 75%. I have to manually reset it as SoC accuracy drifts (maybe 1-2 times per month).

I've run them up to 4.18 and run them down to 3.00 and confirmed their capacity. Due to inclement weather, I will occasionally run it up to 4.05 or 4.15 for additional reserves. As such, I like to let it represent the actual true SoC. IMHO, artificial SoC limits tend to introduce uncertainty when things go out of limits, so it's more reliable to make a mental adjustment, i.e., 75% is 100% of my typical daily usage, but I always have the option of going higher if needed.

 

[westvandude]

Yes that’s a good idea before a bad weather incident to up voltage just in case.

I’m looking for a reasonably priced auto transfer switch that the sunny island can control for generator input.

Here is quick reply back from REC:
Set the Balance voltage end to the same value as the End of charge 3.95 and Balance voltage start @ about 90%.
Is the BMS measuring proper current?
Attached please find the latest update.
BMS is waiting for the lowest cell to come near the End of charge voltage to reset SOC to 100%. To reset SOC Manually go to BMS-> Communicate, Set send to address of the BMS (default is 1) then Type SOCS 0.99 and Query for 99% SOC.

 

[westvandude]

Well that was sure weird.... I was able to go in change the SOC and cell max voltages as instructed. Then on the sunny islands there was no "G" for grid connection and for about 1/2hr it was running my loads off the battery, up to like 45A draw. I was starting to freak out a bit searching forums for why the grid didn't connect and checked some settings, even set grid setting to START. No change. So after looking online for possible solution vs draining my battery to nothing, I go back and take a look and the damn grid is lit up "G" and it's brought the battery up to 100% SOC.

Whew.

 

[1201]

Well that was sure weird.... I was able to go in change the SOC and cell max voltages as instructed. Then on the sunny islands there was no "G" for grid connection and for about 1/2hr it was running my loads off the battery, up to like 45A draw. I was starting to freak out a bit searching forums for why the grid didn't connect and checked some settings, even set grid setting to START. No change. So after looking online for possible solution vs draining my battery to nothing, I go back and take a look and the damn grid is lit up "G" and it's brought the battery up to 100% SOC.

Whew.

Check your event history. There might be a clue there

 

[westvandude]

Not that related to this bms query, but anyone using one of these on an SMA sunny island? https://www.amazon.com/Transfer-Bri...3a-af4e-92a49041acb5&pd_rd_i=B0B63WXZYZ&psc=1
or

RV Transfer Switch, Briidea Automatic Transfer Switch 50 Amp Applied to RV, Speedboats, 240 VAC

NEW GENERATION: Our generator automatic transfer switch sets the generator preferentially to power supply. When the generator stops, it can automatically switch to the mains power supply. When the generator restarts, power will go to the panel from the Gen side and theshore side will be locked...

www.briidea.com

Odd how this sells on amazon.ca for $335!!!

 

[Hedges]

Also need to connect one of my 41series 6kw gridtie inverters to the secure loads panel as well as changing settings on the islands and the gridtie to offgrid. I've got the grid code to do that. Boy these things have way too many settings lol.

Setting Sunny Boy to "Rule 21" ought to work both directly on grid and on Sunny Island.
Although there may be some issue with that for Sunny Island; SMA America in a video recommended Offgrid (Island) mode, doesn't disclose why.

By default, SI makes excursions down to 59 Hz to get clocks back on time. SB setting for UL-1741 disconnects below 59.3 Hz. Under Rule 21, SB is allowed to remain connected down to 58.5 Hz so I tweaked values for UL-1741 in my older SB that don't have Rule 21.

Any reason not to have SB always connected to SI AC1 output?

Pass-through of SI relay is limited to 6.7 kW per SI. For more SB than that (up to 2x wattage of SI), I figure manual interlocked breakers would work, enforcing the required 5 second off time. I use interlock to feed main breaker panel when grid is down, but I have more SI than SB so all SB are on AC1 already.

 

[westvandude]

Thanks for the info. One of my 6.0kw SB 41’s is on critical loads panel but I have it temp wired to welder outlet on main panel direct gridtie.

I’ve got my grid code for changes…. Was told turn SB to offgrid. Rule21 … I’ll do some reading.

Will be good to get SB hooked up so we survive grid failure longer without generator but generstor is necessary.

I’m toying with buying a second leaf battery to double storage capacity but i got my
Existing pack about 2018 for about $3700cdn delivered from about 4hrs drive away. Same now is $5-6k. But this is still a lot cheaper than rack mount 48v bms batteries. The downside is the cost of the rec-bms. As a % of battery cost it’s a lot.

 

[Hedges]

Looks nice, except I keep seeing batteries under inverters, and have seen a couple aftermath of battery fire pictures.

4x SB, of course only about 2x +/- SB fit 56A pass-through relay. So I imagine 2x SB on main panel, 2x SB on SI, and interlocked breakers which transfer main panel from grid to SI. Have to manually switch off breaker feeding SI AC2 so it doesn't try to feed itself.

Now that my SI are fed from a load-side tap I would need a different mechanism to disconnect excess SB. (but because I have 4x SI and not that much SB, it isn't a problem.

I haven't used SB -41 on SI yet, others here have. Offgrid should be better, allows wider voltage and frequency. Rule-21 is documented setting for grid backup. Maybe it gets knocked offline for 5 minutes if SI overshoots 61 Hz, whereas Offgrid remains connected up to 64 Hz. Don't know if there are other issues. Certainly Rule-21 performs anti-islanding, probably OK if SI is stiff enough.

 

[westvandude]

So.... the SOC and battery voltage is varying all over the place. Today it's doing this weird thing of using 0.4-1A from battery negative flow (discharging) and SOC dropped to like 87%. SOC is supposed to be close to 100% in standby all the time in case the grid drops.

So I was trying to get it to charge back up to 100%. Was fooling with Vcell_min and upped that... well bms thought SOC was 1.5% and then shut down both sunny island immediately!! My dumb a** didn't realize after rebooting bms to communicate SOC to what I guessed was about 85% and didn't put the "S" after "SOC" in the text command. When I realized my error I managed to correct SOC and after sunny islands were disconnected from the battery for probably 15mins I was able to restart them. However, it's been like 20-25mins and the grid again is not reconnecting to start charging and I'm drawing 55A from the battery. This is driving me nuts!

Where is a setting for the delay time for the sunny island to reconnect the grid and start charging?? Imagine if your battery SOC was so low that it barely fired up the inverters.... the sunny island manual states to not connect any AC loads to it until the grid comes on and it starts charging the battery. Well I waited some 20mins at least... still no grid showing in display. I can't leave my other UPS to die and have other more critical equipment on the system. This is bs.
-----------------
Update: so eventually I got into grid setup settings and gave a grid "START" command vs "AUTO", etc. It still took a couple of minutes but then it started charging about 68A. Early next morning SOC was 99%. Pretty nerve-racking tho.

In a situation where you have grid power but the inverter/battery/bms is not working, you almost want to have a bypass circuit to power your critical loads panel or lockout switch.... so either have power to panel from main panel or from inverters. If you can't get your inverter restarted or lets say a battery cell dies or you need to do some repairs, having a bypass becomes vital for when SHTF.