Sharing PV input with two different MPPT inverters

imransb1imransb1 Registered Users Posts: 5 ✭✭
HI There,

I have two questions where your kind help and input is required.

I have installed two 5kva MPPT inverters, each with separate battery bank and separate front end load of appliances. The only thing shared between both is PV input which is a 96 volt 6000 Watts array. Please guide if there is any problem with this scenario where PV input is shared.... so far they are working fine but I am a little bit confuse as I am not seeing a full strength input watts that I assume could be cloudy weather now a days. I have tried disconnecting one inverter from PV input but the average input is more or less same. One thing more concerning is my input current has a lot of variation like within seconds it moves from 300 watt to 1200 watts and then come down again etc.

Please see video of input on ... ............and
This video was made after disconnecting PV input from one unit.

Thanks in advance

Comments

  • BB.BB. Super Moderators Posts: 26,872 admin
    In general, it is usually not a good idea to share a single solar array with two or more MPPT charge controllers/MPPT input devices.

    Basically, each controller thinks it is the only one "in charge of the array" when it "sweeps" current and voltage to identify Vmp and Imp (voltage maximum power and current maximum power).

    Generally, once Vmp-array is identified by the charge controller, it modifies its input current from the array to keep the the array voltage around Vmp-array-measusured.

    If you have two or more MPPT devices, they can "step on each other" and either miss-measure Vmp-array and/or get confused and swing the input current to each controller trying to keep Vmp-array at the correct value.

    In your case, you have disconnected one MPPT charge controller and ran your tests. Which, should have been "reasonably" stable current from the array while holding somewhere around 96 Vmp (or maybe 10-20% less on a hot day).

    Now the debugging. First, realize that Imp array is directly proportional to the amount of sunlight the array is recieving. Get a bit of clouds or overcast, and you can easily get 50% or less current from the array (Vmp-array should be fairly constant once there is "any sun" on the array).

    Next, a charge controller has (typically) three different modes. One is "bulk" where the battery is below the charging set point and accepting as much current as the charge controller can output. Once the battery bank reaches ~14.4-14.8 volts (or ~57.6-59.2 volts for a 48 volt battery bank), the controller goes into "absorb" mode--Which will try to hold the "absorb voltage set point" for ~2-6 hours (to take the battery from ~80% state of charge to >90% state of charge). Once the battery is "fully charged" as determined by the charge controller, it will enter float mode of ~13.6 volts (~54.4 volts) and "hold that voltage" until the sun goes down (or the load on the battery exceeds the capabilities of the solar array).

    Of course, there are other issues. When the charge controller is in "voltage regulation", the output current will be the sum of Battery Charging + DC&AC loads. If you have variable AC loads, the charge controller will output variable current--And less than the array is capable of.

    Also, besides clouds, ANY SHADING on the solar array (trees, bushes, overhead wires/ropes, buildings) can dramatically reduce the array output. Even a heavy power line shadow across the array can output Imp-array by 50%.

    I did not catch the name/model numbers of what equipment you are using, and what voltage battery bank you are running (and type/AH of battery bank)... So, it is difficult to give any specific information on debugging for your product(s).

    Besides running on one MPPT charge controller, I would suggest measuring the current from each (3 panel?) string of your array with your DC current clamp meter--Make sure that each string is outputting is share of the current. Also monitor the PV-array voltage and see if it is (hopefully) constant around 77 to 87+ volts. If the array is > 96-100 VDC, then the array it not being fully loaded by the charge controller (may be perfectly OK if the charge controller thinks not much DC current is needed, or may be other issues if the battery bank voltage/State of Charge is low and you need lots of current).

    Tell us what voltage the battery bank is running while testing (i.e., 48.0/54.4/59.2 volts/etc.).

    Measure the voltage drop from the charge controller to the battery bank (DC cabling). The total voltage drop should be no more than ~0.05 to 0.10 volts (for 12 volt) or ~0.2 to 0.4 volts for a 48 volt bank under load (i.e., how long and what diameter/AWG are the cables between the charge controller and battery bank).

    Check the current and voltage drop of the battery cables (make sure all connections are clean and tight, heavy enough/short cable from battery bank to AC inverters, that all strings of battery bank are sharing current for charging/discharging/etc.).

    At this point, I do not have any other good ideas of why a single MPPT controller is causing you such variations in array wattage... Variable shading on Array, variable loads on battery bank with near 100% state of charge battery bank) are the only two "everything is OK variations I can think of.

    Try the other MPPT charge controller (swap to the second MPPT controller) and see if you have the same issues with array wattage.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • imransb1imransb1 Registered Users Posts: 5 ✭✭
    Hi Bill, Really appreciate the detailed reply.

    Here are few things that I know more about the system.

    While the system is connected to one MPPT inverter I checked voltages the from the disconnected units cable ends which were 90 to 92 volts stable without variation during the test.
    The load on inverter is constant which is 2300 Watts (one 1.5 ton A/C and fridge)
    Battery voltage during test are ranging from 47 to 49
    Inverter specifications are attached below in picture
    We are using 10 mm good quality DC cable that is 70 feet in length from PV to inverter .... while 4 mm DC cables are connecting each 3 panel bank to main 10 mm.
    Battery to inverter cable is also 10 mm and distance is 4 to 5 feet.
    We are using 150 Amp Flooded brand new batteries with one bank and 120 amp batteries with second. The test uploaded is with the unit which has 150 Amp flooded battery bank. Both have 4 x 12 volt batteries - 48 volt system.
    I have already checked each 3 member bank separately which are contributing 1 to 2 amps per bank.
    Also checked short circuit amps of few panels separately  which are up to 8 amps in the same sunlight
    Other than the PV input variation there is no abnormality observed in both inverters. Both inverters work fine when charging and discharging with main input.

    Once again bundle of thanks for sharing the detailed input... I have discussed same with another senior engineer who has recommended me the same and asked to check the same with second inverter.


    One more thing...if we used a shared battery bank with two inverters and keep the PV input shared does that makes the system more sensible or it will be same as with separate battery banks with both units.

  • mcgivormcgivor Registered Users Posts: 1,304 ✭✭✭✭
     @Imransb1 said. One more thing...if we used a shared battery bank with two inverters and keep the PV input shared does that makes the system more sensible or it will be same as with separate battery banks with both units.

    Although it is possible to parallel the 2 battery banks, it probably is not the best thing to do, keeping series strings is preferable to a series parallel arrangement, even if they were the same capacity. Sharing a single array with 2 charge controllers is not possible, each controller must have its own separate array.
      1500W, 6× Schutten 250W Poly panels , Schneider 150 60 CC, Schneider SW 2524 inverter, 8×T105 GC 24V nominal 

  • westbranchwestbranch Solar Expert Posts: 4,873 ✭✭✭✭
    Imransb1, Did I read your statement that you tested the open circuit of the panels and got ~ 8A and that the STRING of 3 panels  yielded 1 to 2 A ?  This should not be that low unless the batteries are full or very near full... You could redo that test when the batteries are just starting to be charged. .. 
     
    KID #51B  4s 140W to 24V 900Ah C&D AGM
    CL#29032 FW 2126/ 2073/ 2133 175A E-Panel WBjr, 3 x 4s 140W to 24V 900Ah C&D AGM 
    Cotek ST1500W 24V Inverter,OmniCharge 3024,
    2 x Cisco WRT54GL i/c DD-WRT Rtr & Bridge, Hughes1100 Sat Modem
    Eu3/2/1000i Gens, 1680W & E-Panel/WBjr to come, CL #647 asleep
    West Chilcotin, BC, Canada
  • BB.BB. Super Moderators Posts: 26,872 admin
    OK... Let me talk a bit about the installation. We use some rules of thumbs here to design a reliable (over time/temperature/battery state of charge/etc.) to size everything. Usually we start with your loads to design the battery bank. And then the battery bank (+loads) to design the solar panel/array (usually 10% minimum rate of charge for full time off grid system). And based on where you live (hours of sun to replace energy used every day).

    Nominally, I would suggest combining the battery banks into a a single system. You have have two or more solar arrays+MPPT charge controllers, charging one battery bank. But, for this discussion, since you have two independent systems, I will do a design for the larger one just to show you the way I run a quick design.

    The design will assume 1x 5,000 Watt AC inverter and a 2,300 Watt load--Averaging 50% duty cycle for 10 hours per day/evening. Knowing your loads is critical to designing a cost effective/reliable system. And remember that your loads are pulling "battery charging" current if operated all day long (vs, for example, charge your battery bank until 2-3pm/14:00-15:00 hours, then the battery is full, and you run your system).

    I have some concerns about the design of your system vs your loads (I try for a "balanced" system design). Note that I am not there, and I may be misunderstanding your system+loads and their usage.

    If your system is actually "on grid" and used to ride through common afternoon brown outs (~4-6 hours an evening), the sizing below is not going to be right and is going to way oversize the system. But I think this is still a good time to go through a full off grid system design so you can see my process.

    Air Conditioning is a very large load for an off grid solar system--We usually suggest a well insulated home and mini-split A/C (+heat pump if needed) as they are very efficient, and the "inverter type" mini-splits can be operated at low cooling levels and current (i.e., on low, a 2.5 kW system may run at 600 watts or less, with no starting surge).

    For the battery bank, a good rule of thumb is to size the bank for 2 days of "no sun" and 50% maximum discharge. Using my above assumptions:
    • 2,300 Watt load * 0.50 duty cycle * 10 hours per day * 1/0.85 AC inverter eff * 2 days of storage * 0.50 maximum battery discharge (for longer battery life) * 1/48 volt battery bank = 1,127 Amp*Hour @ 48 volt battery bank
    Also, for a quick sannity check--There is a calculation for average power drawn from the battery bank, and maximum surge power. For an off grid system that uses the above design rules, this step is usually skipped. For an afternoon brownout on-grid system, this step is usually needed.

    Flooded cell batteries are generally operated at ~C/20 (20 hours to draw a battery dead) to C/8 discharge rate (C/5 rate can work for a few minutes/hour for short/heavy loads). So, if you have a 2,300 Watt load, and the smallest suggested battery bank would be:
    • 2,300 Watts * 1/0.85 AC inverter eff * 8 hour discharge rate * 1/48 volt battery bank = 451 AH @ 48 volt minimum battery bank suggested
    Your 150 AH battery bank is being discharge at way too heavy of current draw if your A/C system is drawing 2,300 Watts for hours a day.

    Next, there is sizing for AC surge... A typical AC inverter can surge (a few seconds of surge) about 2x rated output power. The maximum surge current from a flooded cell lead acid battery bank should be around C/2.5 discharge rate:
    • 10,000 Watt AC surge * 1/0.85 AC inverter eff * 1/48 volts * 2.5 rate of discharge = 613 AH @ 48 volts for 10,000 watt surge
    If you do not plan on supporting a 10,000 Watt surge, then plug in the actual surge you plan on supporting (note that items like AC compressors and well pumps may have very short surge times and you need a peak hold current clamp DMMM (digital multi meter) to catch the actual surge current.

    So, for sake of discussion, say you have a 1,127 AH @ 48 volt battery bank. You want to charge that at 5% to 13% rate of charge typically. 5% for seasonal/weekend usage... 10%+ for full time off grid. Here is the 10% calculation (minimum suggested) for full time off grid system:
    • 1,127 AH * 59 volts charging * 1/0.77 panel+charger deratings * 0.10 rate of charge = 8,535 Watt array nominal off grid system
    And there is sizing the array based on where you are located (hours of sun based on location, average weather, etc.). For Lahore Pakistan, fixed array, tilted for best year-round collection:
    http://www.solarelectricityhandbook.com/solar-irradiance.html

    Lahore
    Average Solar Insolation figures

    Measured in kWh/m2/day onto a solar panel set at a 58° angle from vertical:
    (For best year-round performance)
    Jan Feb Mar Apr May Jun
    4.72
     
    5.66
     
    6.16
     
    6.50
     
    6.68
     
    6.34
     
    Jul Aug Sep Oct Nov Dec
    5.52
     
    5.46
     
    6.00
     
    6.43
     
    5.73
     
    4.91
     
    Your daily usage would be:
    • 2,300 Watts * 0.50 duty cycle * 10 hours per day = 8,309 Watt*Hours per day
    • 8,309 WH per day * 1/0.52 off grid AC system eff * 1/6.0 hours per day (summer sun) = 2,663 Watt array minimum
    And here is where we have an issue... A large battery bank (to support heavy loads of A/C) and you have lots of sun in your location. Assuming a minimum of 5% rate of charge (~4,250 Watt array) vs the actual minimum array of ~2,663 watts based on hours of sun per day (lots of sun). Not knowing your loads--I am a bit at a loss--Is the battery bank "oversized" or if the battery bank really needs to be that large, need a large solar array to meet the minimum rate of charge to keep the battery bank "happy".

    Anyways--That is the back of the envelope calculations for the battery bank and solar array based on my guesstimate of loads and your location.

    That is enough for part 1.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • BB.BB. Super Moderators Posts: 26,872 admin
    And now for part 2... Some specific questions about your cabling... Just to be clear, you use mm wire diameter and not mm^2 (mm squared) wire area for your your sizing (i.e., 10 mm diameter, not 10 mm^2). Here is a chart to US AWG (American Wire Gauge--Which I am more familiar with):

    http://www.rapidtables.com/calc/wire/awg-to-mm.htm#chart

    For a 3,000 Watt solar array with 70 feet (one way run) of 10 mm or 000 AWG --- really, that is quite heavy and expensive (if copper cable)????
    • 3,000 watt / 96 votls Vmp-array = 31.25 amps
    http://www.calculator.net/voltage-drop-calculator.html?material=copper&wiresize=0.2028&voltage=96&phase=dc&noofconductor=1&distance=70&distanceunit=feet&amperes=31.25&x=49&y=15

    Voltage drop: 0.27
    Voltage drop percentage: 0.28%
    Voltage at the end: 95.73

    Typically, we design for 1% to 3% votlage drop. Going below 1% voltage drop is usually a lot of money for the wiring without much return in investment.

    The second concern is the AC inverter... The name plate on the inverter does not look very accurate. Typically, the inverter DC input current would be (assuming 4,000 Watt maximum continuous rating???). Note that your inverter is rated for 5,000 VA maximum, and 4,000 Watts maximum. VA is not Watts (we can have a longer discussion if needed about Power Factor / Watts / VA):
    • Watts = Volts * Current (VA) * Power Factor = Volts * Amps * Cosine (angle between voltage and current)
    • 4,000 Watts = 5,000 VA * 0.80 PF
    And for the DC power input to AC inverter:
    • 4,000 Watts * 1/0.85 AC inverter eff * 1/41.0 battery cutoff voltage = 115 Amps maximum DC battery bank continuous current
    And in the US/North America, I would highly suggest a derating of 0.85 for wiring and circuit breakers/fuses (required by National Electric Code):
    • 115 Amps continuous current * 1/0.85 derating = 135 Amp rated wiring and fuse/breaker
    From this table (the "real" NEC table/calculations are more complex):

    https://lugsdirect.com/WireCurrentAmpacitiesNEC-Table-301-16.htm

    The minimum cable size should be ~1 to 00 AWG wire... Or ~7.4 to 9.3 mm wire diameter (copper, size depends on insulation temperature rating).

    You have 10 mm (diameter) @ 5 feet (one way) of cable, the voltage drop would be:
    http://www.calculator.net/voltage-drop-calculator.html?material=copper&wiresize=0.2028&voltage=41&phase=dc&noofconductor=1&distance=5&distanceunit=feet&amperes=115&x=68&y=11

    Voltage drop: 0.071
    Voltage drop percentage: 0.17%
    Voltage at the end: 40.929

    The length/diameter of cable is not an issue--Very little voltage drop (suggest a maximum of 2 volt wiring drop at max DC current on a 48 volt battery bank). And 000 cable (10 mm is 1/2 way between 00 and 000 cable) is good for ~150-200 Amps (in conduit).

    I am still unsure if you are talking about 10 mm diameter vs 10 mm^2 surface area of cut cable--So you will have to confirm that back (and decide what maximum current per conductor should be). Note that marine rated wiring (usually not run in conduit) allows for much higher current capacity--NEC tends to be very conservative (I try for conservative, fewer problems and safety issues).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mcgivormcgivor Registered Users Posts: 1,304 ✭✭✭✭
    Basic question, @BB  can two controllers, MPPT or PWM   share the same array? From my understanding this is not possible, but I've been proven wrong, on more than one occasion.i
      1500W, 6× Schutten 250W Poly panels , Schneider 150 60 CC, Schneider SW 2524 inverter, 8×T105 GC 24V nominal 

  • BB.BB. Super Moderators Posts: 26,872 admin
    MPPT charge controllers--No, each should have its own array so it can mange Pmp=Vmp*Imp.

    I can never say never--Because some engineer may have designed a MPPT controller with "shareable" Vpanel inputs somewhere (i.e., the controllers have a communications circuit between them to manage the shared array parameters).

    For PWM controllers, if the two battery banks share the same ground (negative battery bus), then two PWM controllers can share teh same array--However, since PWM controllers cannot throttle the input current from the array, then each PWM controller has to be large enough to manage the entire array Imp output current (i.e., a 20 amp maximum array, then you need two 20 amp PWM controllers minimum).

    Obviously, when both PWM controllers are "on", the current is shared (based on battery voltage/wire resistance/etc.) between the two battery banks. When only one PWM controller is "on", then that battery bank+controller will receive full array charging current.

    There are a couple Positive Ground PWM controllers out there--I am not sure how to share a single array with them (don't know how they are designed internally). So I cannot say for those types.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • bill von novakbill von novak Solar Expert Posts: 711 ✭✭✭✭
    BB. said:
    MPPT charge controllers--No, each should have its own array so it can mange Pmp=Vmp*Imp.

    . . .

    There are a couple Positive Ground PWM controllers out there--I am not sure how to share a single array with them (don't know how they are designed internally). So I cannot say for those types.

    -Bill
    The switched-positive vs switched-negative is also big issue.  A system that uses two dissimilar charge controllers on the same array - one that switches on the positive side, one that switches on the negative side - will apply full array voltage to the battery, with no way to stop it other than the breakers  (disabling the charge controllers will not stop the current.)  This, of course, will rapidly destroy the battery.

    If you are using the same charge controllers for both that MIGHT work; the big problem then will be the MPPT algorithms fighting as you mention above. 

  • imransb1imransb1 Registered Users Posts: 5 ✭✭
    HI Every one,

    I have separated both units PV input. Unit A has now 750 watts solar panels connected and Unit B has 5200 Watts Solar panels connected. Both units are 5 KVA MPPT. 

    Please see videos below and share your valuable feed back... i guess Unit B charge controller is malfunctioned and requires warranty claim.

    Unit A:

    Unit B:

    Regards,
    Imran
  • mcgivormcgivor Registered Users Posts: 1,304 ✭✭✭✭
    Try switching inputs and see what you get, if possible, can't exactly see what the figures represent, could you describe?
      1500W, 6× Schutten 250W Poly panels , Schneider 150 60 CC, Schneider SW 2524 inverter, 8×T105 GC 24V nominal 

  • imransb1imransb1 Registered Users Posts: 5 ✭✭
    Unit A is connected to 750 watts of panels and is getting constant input of 450 to 500 Watts , while battery is on 48 volts. PV array voltage are 79, front end load is 800 watts. Every thing seems to be fine and as per my expectation.

    Unit B is connected to 5200 watts of panels and is getting variable input of 260 to 800 Watts , while battery is on 46.6 volts. PV array voltage are 95 to 97, front end load is 2300 watts. Batteries and panels both are connected to 10 MM cable.


  • mcgivormcgivor Registered Users Posts: 1,304 ✭✭✭✭
    If it is possible to swap over the inputs that would confirm wether the unit B is the problem, or the array has an issue, a verification of sorts before sending it out on a warranty claim.
      1500W, 6× Schutten 250W Poly panels , Schneider 150 60 CC, Schneider SW 2524 inverter, 8×T105 GC 24V nominal 

  • imransb1imransb1 Registered Users Posts: 5 ✭✭
    Just swapped the PV input between Unit A and Unit B... and Unit B is again having fluctuation in input current while Unit A is stable.
    Lesson Learned ... Sharing PV input can be fatal for MPPT charge controllers ... sending Unit for Warranty Claim now. 
    Thanks every one for support.
  • BB.BB. Super Moderators Posts: 26,872 admin
    Sharing the Array between two MPPT controllers should not cause damage (that I can think of).

    Except--MPPT charge controllers (as far as I know) are NOT isolated electrically between array and battery bank.

    For example, if (somehow) you have (for example) a load between battery bank A+ and battery bank B-, the return current will flow through the ground bond in the two charge controllers (typically rated to ~60-80 amps or so maximum) -- And if that return current is very large (a large DC battery bank load+charging current), you could damage the ground path in the charge controllers.

    You may very well just have an honest failure that was not the result of anything you did.

    But, if you got water/dust/electrostatic discharge/excessive heat/even a bug living in the inverter in/round the charge controllers/inverters--That can cause problems/failures.

    There sometimes can be wiring issues too. Most MPPT controllers have a "common ground" connection between the array negative and the battery negative. And it does not matter how you wire the array negative ground to the charge controllers (for example, you could run Array negative directly to battery negative. And all will be fine.

    However, there are (a few?) MPPT charge controllers that (for example) measure the array current through the negative connection inside the controller (from Vpanel negative to Vbattery negative). If you do not route the array +/- to the Vpanel input +/- (and instead bypass the controller Vpanel- and Vbatt- ground connections), the MPPT controllers will not work correctly because they cannot measuring the Varray current (to manage Vmp/Imp of the array) because the shunt in the ground connections was bypassed.

    Hope that made sense....

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
Sign In or Register to comment.