Helping someone with a SMA/Midnite System (really a FLA Battery issue)

Surfpath
Surfpath Solar Expert Posts: 463 ✭✭✭
Hi Forum buddies,
I'm helping someone with a Sunny (SMA) system that has battery issues. Can someone help to double check my logic?  

Here is the basic situation:

Existing set up: Sunny Island 5048, Sunny Boy 2400 (I think), 3900 watts of panel. The panels are in 2 arrays: 1) 2400 watts controlled by the Sunny Boy and 2) 1,500 watts of PV being controlled by a Midnite 250 connected to the SMA equipment as a 'slave' using the Midnite Comm adapter.  These 2 arrays charge a 6 year old [sorry I said 4 year earlier tonight] 640 amphr 48V Trojan RE-A FLA bank (configured in 2 strings of 8 batteries each).

Here's the 1st problem:
About 18 months ago their 1.5KW windmill died (they are in a very windy area). So there goes nightime & cloudy charging!  So they are on solar alone now. But 3900 watts (*0.77/48) = ~62 amps, which should give them potentially a ~10% charging rate? Should be fine right?  

Here's really the problem:
  • One of the cells in one of the batteries (#7) has a dead cell and the others have pretty disparate recent SG's (I will post these below).
  • The owner has never EQ'd this bank - The SMA digital readout is old and really challenging to program (plus it is practically in code). The owner has always had a hard time programming the unit.
  • IMHO, SMA equipment has pretty unique charging profiles, often going to float very early, and only absorbing to 100% SOC periodically). There is some debate in this forum about if this is beneficial or not, but so I don't want to argue this point, but state that it's difficult for the owner to go to 100% very often. 
  • The old 2016 Trojan battery charging parameters are still programmed into the unit (about 2.5 years? ago Trojan increased their recommended charging voltages for absorb/EQ pretty significantly.  
My opinion:
Obviously the bank needs an EQ ASAP.  Also the dead battery needs to be replaced with 2 GC Batts once, and if, it is determined that the rest of the bank can be salvaged.
continued....
Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.

Comments

  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    Recent SG's for the bank are below
    • Note these are the SG values for for each of the 16 batteries (3 cells each).
    • See #7 with the dead cell (the middle cell). 
    • From the top....#2, and #15 also look very disparate. 
    • 8 of the batteries have SG disparities of 30 points or more. 
      
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    edited April 2021 #3
    SMA makes performing a manual EQ look easy. But I checked out the LED readout and it is hard to decipher the weird codes and symbols.  The display is also faded. I can see how the owners have not been able to perform an EQ easily. 

    If we can't figure out how to EQ using the SMA gear, one of the options we considered is decoupling the Midnite controller and using it to manually EQ. Even that looks like a pain, given the state of the display.   

    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,728 ✭✭✭✭✭✭
    As you know, the clock is ticking! You have to do something ASAP, like today. Can' t you just turn off the inverter and do an EQ with the CC? Lose string #2 for now!
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    edited April 2021 #5
    Thanks Dave, by the way EQing a bank with a dead cell?  Will this be an issue?
    Also I dont know the answer to your question. SMA does things differently and I am an Outback user.  
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,728 ✭✭✭✭✭✭
    My last sentence!    I only use Outback and Schneider.  
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    A quick update: they were able to access and change? the charging parameters menu, but have still not been able to start an equalization as of yet.

    i’ve dropped the idea of decoupling the midnight controlled array to use that to EQ. This would be a pain to do (again, the SMA one line readout is literally unreadable nonsense) and it also could not be a long-term solution, since The sunny and the midnight charging parameters would never be able to synchronize independently. They must communicate through the COMM device.
    So they are stuck for now.
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • boB
    boB Solar Expert Posts: 1,030 ✭✭✭✭
    Just disconnect the Classics from the communications box temporarily and try starting the EQ from the Classics.

    Sunny Island doesn't always do what you want it to.


  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    edited May 2021 #9
    boB,
    I agree. 
    I am at the location now, and despite the display issues, I think we have changed the eq parameters to 2.70 per cell - or 64.8v for the bank, as Trojan suggests for this battery. 

    We are still trying to perform the Eq using the SMA system. Decoupling the midnite controller looks more complex.
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    After changing the eq parameter (see the red arrow), and initiating a manual EQ, the charging voltage set point rose from 61.7 to 63.0. So not quite the 64.8 that Trojan stipulates.
    However, the battery voltage (see yellow arrow) never rose to 63v. The highest I saw it was 58volts.

    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    edited May 2021 #11
    This was at 345 pm with enough sunlight (and 3900 watt of panels) to raise the charging voltage to 64v. My guess is that there was at least 1200watts of PV for this time of day.
    So, is this the SMA gear being stubborn again? Even though it was a manual EQ?
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    As an aside I’ll give you a little taste of the display issue...and no it’s not programmed in Japanese.


    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Vic
    Vic Solar Expert Posts: 3,208 ✭✭✭✭
    Surfpath said:
    This was at 345 pm with enough sunlight (and 3900 watt of panels) to raise the charging voltage to 64v. My guess is that there was at least 1200watts of PV for this time of day.
    So, is this the SMA gear being stubborn again? Even though it was a manual EQ?
    Hello Surf..,

    Good that you are helping this person with the battery charging/EQ..

    IMO,  1200 W of PV power is seldom sufficient to reach EQ voltage on a 640-ish Ah battery bank,   especially with fairly aged batteries.   A battery bank that really needs an EQ often requires more that 5% of Capacity to begin to reach EQ voltage.   AND,  that is after the battery bank has been about as fully-charged as it will get in Abosrb.

    Please keep trying to EQ this bank.   FWIW,   (IMO),   Good Luck, Vic
    Off Grid - Two systems -- 4 SW+ 5548 Inverters, Surrette 4KS25 1280 AH X2@48V, 11.1 KW STC PV, 4X MidNite Classic 150 w/ WBjrs, Beta KID on S-530s, MX-60s, MN Bkrs/Boxes.  25 KVA Polyphase Kubota diesel,  Honda Eu6500isa,  Eu3000is-es, Eu2000,  Eu1000 gensets.  Thanks Wind-Sun for this great Forum.
  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    Thanks Vic,
    You’re right, my personal bank is only 370 amp hrs, so it tends to be more “responsive” to an eq. 

    So more EQ tomorrow for this bank. hopefully it will be a sunny day. We’ll see if the battery voltage gets to the actual 64 volts that’s needed.
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,728 ✭✭✭✭✭✭
    Might want to check for surge protection. When electronics fail that is often an issue offgrid. Can't you run a genset to speed this up?
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    Hi Dave,
    Their Honda 2000 is recently down. Plus if there were any related programming or setting changes associated with using a genny the display may be an issue. Just to figure out the password yesterday to log on as an operator took us 45 mins.
    The Sunny inverter uses a 6 digit password code that changes hourly, you then have to add up all 6 digits to enter that into the password field. Go figure.
    No simple one code like 141.
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    This recent update from the owner today...

    The set voltage has been on 63.0 since 10.30am. The actual (battery) voltage fluctuated between 55.3 V & 57.7 V for the longest while but it has now (at 12.30pm) gone up to 61.5 V and holding steady so we’re getting a bit more bubbling.
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • mcgivor
    mcgivor Solar Expert Posts: 3,854 ✭✭✭✭✭✭
    edited May 2021 #18
    @Surfpath

    Just curious as to wether the battery is being cycled between the EQ attempts, it's difficult to achieve positive results if doing so, particularly with marginal input available. To achieve a complete equalization may take many hours of continuous charging or done over multiple days without discharge, tough to achieve if energy is required and its the only source.
    1500W, 6× Schutten 250W Poly panels , Schneider MPPT 60 150 CC, Schneider SW 2524 inverter, 400Ah LFP 24V nominal battery with Battery Bodyguard BMS 
    Second system 1890W  3 × 300W No name brand poly, 3×330 Sunsolar Poly panels, Morningstar TS 60 PWM controller, no name 2000W inverter 400Ah LFP 24V nominal battery with Daly BMS, used for water pumping and day time air conditioning.  
    5Kw Yanmar clone single cylinder air cooled diesel generator for rare emergency charging and welding.
  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    Hi,
    The owners are using the bank at night with probably a 20% discharge? 

    On Saturday I believe we were able to input the right charging parameter for a proper EQ. However, despite there being lots of sunshine on Sunday, the Sunny Inverter only raised the the bank voltage to 61.7v, ie not enough to EQ. They have a C10 charge capacity, so EQ is totally possible.

    Basically, it was sort of an above average absorb.
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    So perhaps the right thing now is to unhook the midnite controller from the Sunny Island and have it do the EQ like what bOB said. I’m a little hesitant to do this if it involves any kind of programming or reprogramming of the Sunny Island. 

    Im no electrician and we have little or no technical support in the area.

    The set up for the midnite (classic 250) is that it controls 1500w, has a Pv breaker box, and then a shunt with a whiz bang. It connects to the SMA communications adapter and follows the Sunny charging protocol. I’ll post pictures.
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    Above is the Shunt and whiz bang.
    Below is the comms adaptor.

    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    edited May 2021 #22
    Is it a matter of unplugging the green and blue canbus/modbus jacks, then programming the midnite controller to perform an EQ. 
    After that is done then reprogramming the Midnite controller to go back into slave mode? Then plugging the green and blue wires back in again?
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,728 ✭✭✭✭✭✭
    You should not have to reprogram anything. It is in non volatile memory and should stay. However, I do not use this equipment so I am just saying that you should be able to do a complete DC shutdown of solar and battery. Plug the network cables in and out several times and then restore DC to the battery and then the solar. Observe system. Then program all of the solar for an EQ for at least 3 hours at 65vdc.It may take days of this!  Observe temperature and be safe. You should do what Bob and Vic have told you. Did you check for SPD's in the system? They probably caused this if they were not there! Good Luck!

    Kind of weird (to me) that you have both a canbus and modbus input networks? 
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • Saggys
    Saggys Solar Expert Posts: 189 ✭✭✭


    Kind of weird (to me) that you have both a canbus and modbus input networks? 
    What he said, No output?
  • Surfpath
    Surfpath Solar Expert Posts: 463 ✭✭✭
    Hi Dave,
    Yes there is an SPD connected to the midnight array, but not the Sunny 
    I have no idea why the comms setup looks off to you and Saggy, I’ll pass this on to the owner.
    Thanks!
    Outback Flexpower 1 (FM80, VFX3048E-230v, Mate, FlexNetDC) 2,730watts of "Grid-type" PV, 370 AmpHrs Trojan RE-B's, Honda 2000 watt genny, 100% off grid.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,728 ✭✭✭✭✭✭
    Who installed this? It does not sound like the owner is up to speed. Very rare (in my experience) that a system would use both modbus and canbus for a power system network. To me it looks like someone mixed up the ins/outs on your picture. That it has taken this long to somehow get a genset to fix this is also weird. The movie groundhog day where the same thing keeps happening. Good that you are trying though!
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net