3 panels in parallel

mike49
mike49 Registered Users Posts: 28 ✭✭
edited December 2015 in Solar Beginners Corner #1
Hi all, I've been searching and reading as much as I can about connecting solar panels in parallel, but still unsure if there are different requirements between connecting 2 panels in parallel versus 3 panels in parallel...

I have a small simple hobby/learning setup - one 100watt 12v panel connected to a pwm charge controller via 10gauge wire running 20 feet with no fuse between panel and controller. charge controller connected to old 12v car battery via 10gauge wire with 15a fuse. I use this to power some led lights via a 200watt inverter for my shed, working great since May.

Since November here in Toronto, I've been seeing several days in a row cloudy/gloomy with almost no sun and maybe one or two days a week getting a few sun hours in total. Much less than supposedly 2 sun hours a day average solar insolation for this time of year for my location that's for sure. So battery is not getting fully charged most of the time now...

I happen to have a 2nd, 40watt solar panel that I use for camping. From what I understand from everything I've read on paralleling different size panels, I said "what the heck" and connected the 2 solar panels in parallel with a pair of parallel Y connectors. The 100watt panel is Vmp 18.2v  Imp 5.5a, the 40watt panel is Vmp17.1v Imp2.3a. I'm guessing this is ok since the voltage difference is not that much and I'm not using an mppt controller so slight drop in voltage from 18.2 to 17.1 isn't too big of a deal...
This seems to be working fine for time being and I am seeing the extra combined current on my charge controller's LCD display.

I'm planning on getting another 100watt panel exactly same as the one I have.
So I'll have 3 panels eventually, two 100watt and one 40watt.

Want to put all 3 in parallel, but not sure of some of the stuff I am reading about 2 versus 3 panels in parallel....
It seems that 3 panels in parallel requires an order of magnitude more complexity than simple 2 panels in parallel?

With 3 panels in parallel, do I need add
a) a fuse for each panel before the connection to the 3-way parallel connector?
b) a blocking diode for each panel before the connection to the 3-way parallel connector?
c) a combiner box with fuses and blocking diodes in lieu of a) and b) ?

I'm hoping the answer to all the above is "no" and I can just simply connect the 3 panels in parallel to the 3-way parallel connector without fuses and diodes, just like what I'm doing now with my 2 panels? What can go wrong with 3 panels in parallel and needing fuses and/or diodes that isn't the case with 2 panels in parallel? Something about one panel shorting out and all the current flowing through a panel? This is the part I don't get...

thanks folks for your help ahead of time...

 
 
 

 


small hobby/learning system in Toronto, Canada:
3x100w parallel via 30A PWM to old 12v car battery to 200w inverter for shed LED lights
(1x100w may-sept)

Comments

  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    edited December 2015 #2
    mike49 said:
    With 3 panels in parallel, do I need add
    a) a fuse for each panel before the connection to the 3-way parallel connector?
    b) a blocking diode for each panel before the connection to the 3-way parallel connector?
    c) a combiner box with fuses and blocking diodes in lieu of a) and b) ?
    Welcome to the forum,

    You don't need blocking diodes at all (with any number of panels) because your controller prevents the battery from backfeeding the panels.

    Yes, you do need a fuse for each panel whenever there are more than two panels in parallel.  It's best to do this with a combiner box, but it can be done with in-line fuses "before the connection to the 3-way parallel connector".  If you use a combiner box, circuit breakers are recommended over fuses.

    Do you know what size fuse/breakers to use?

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Welcome to the forum Mike.

    OK, normally we deal with "matched sets" of panels. It makes the "rules" a bit less complex to understand. When you mix and match, then we have to go into the details.

    For "matching" panels... We like to see the Vmp of the panels within ~10% or better of each other. For PWM charge controllers, as long as the Vmp panel is between ~17.5 and 20 volts--You are OK (MPPT controllers atempt to track the Vmp-array and with mixed Vmp panels in parallel, there is no one "ideal" Vmp voltage--There are several local peaks which can confuse the controller. And if Vmp is >> 18.6 volts or so on a PWM controller, the panels are just less efficinent (battery is charging at ~15 volts and panel is capable of outputting say 36 volts--So, 1/2 of the wattage is lost (power = Voltage * Current -- V=18 volts vs V=36 volts with fixed current to battery).

    For series connection of panels, again, we look for Imp to be withing 10% or better between to series panels.

    When paralleling panels, there is a possibility of a short circuit in a panel (rock, hail damage, animals chew on wiring, etc.)... And so each panel has a series fuse rating--The maximum size fuse you can put in the wiring to the panel to reduce the risk of fire (older, lower voltage panel many times do not have a series fuse specification).

    So--If you have a panel that has Isc~10 amps and a series fuse rating of 15 amps, if you have a single panel, no fuse is required (charge controller is assumed not to be a source of current back flowing to array).

    If you have two panels in parallel, and one panel gets shorted, the second panel cannot back feed more than ~10 amps of current, so a 15 amp fuse is not required.

    However, if you have three panels, and one gets shorted, now Isc of two panels is 20 amps and will exceed the 15 amp fuse rating--So, for three matched panels, we would suggest that almost all 3 or more parallel strings of panels woudl need a series protection fuse per panel (or series string).

    Now, your confusion... You have panels with different Isc (and series fuse) ratings. One panel is 2x the rating (watts, current) of the smaller panel. So, it is possible for the 100 Watt panel to back feed a 40 watt panel and (possible) start a fire if the conditions are right (or "wrong"). In your case, you would need a series fuse rating for the 40 watt panel and not need one for the 100 watt panel. However, you need to know the series fuse rating for each panel (on the data sheet)--However, I would guess there is no series fuse rating for the 40 watt panel or even possibly the 100 watt panel.

    You can estimate the appropriate fuse by taking Imp of the panel and multiplying by a fudge factor and rounding up to the next standard size fuse or circuit breaker:

    2.3 amps Imp * 1.25 NEC wiring derating * 1.25 NEC solar derating = 3.6 amps ~ 4 to 5 amp fuse/breaker

    The Imp of your 100 Watt panel is 5.5 amps, and the estimated Isc is roughly (1.25 x 5.5 amps Imp=) 6.9 amps--So its Isc rating exceeds the 4-5 amps (estimated) maximum fused rating of your 40 watt panel.

    It is a close call--If you are really worried about possible fire, then install the fuse/breaker (at least on the 40 watt panel). If the panel is mounted someplace where a possible fire is not an issue (over gravel, no dry grass, etc.)--Most likely, nothing bad will ever happen. Fuses/breakers are not "exact" and feeding 5.5 amps through a 5 amp fuse may take minutes to hours to trip (if ever). Obviously, adding a second 100 Watt panel would mandate the use of a fuse on the 40 watt panel (probably no fuses would be needed on the 2 paralleled 100 Watt panels as 100 watt+40 Watt panels do not exceed the series fuse spec. for the second 100 Watt panel).

    You could use blocking diodes instead of fuses/breakers--But I would tend to avoid that. Diodes have voltage drop (0.2 to 1.0+ volt drop) -- So they get hot (need heat sinking and electrical insulation) and can fail (diodes may fail either open or shorted) and (for NEC/UL purposes), you need two series diodes when using them in "protective" situations. Blocking diodes were used in decades past, but modern charge controllers and use of fuses/breakers has pretty much obsoleted the use of blocking diodes and paralleling strings.

    Lastly, as you make your system larger (higher wattage solar array)--You are going to get into the problem of sending 12 volts any distance. For example, say you have 2x 100 watt plus 40 watt panel for Imp-array~13.3 amps. Using a generic voltage drop calculator for 20 feet of #10 AWG cable and 13.3 amps, the voltage drop will be:

    http://www.calculator.net/voltage-drop-calculator.html?material=copper&wiresize=3.277&voltage=15&phase=dc&noofconductor=1&distance=20&distanceunit=feet&amperes=13.3&x=73&y=10

    Roughly ~0.53 volt drop... That is getting up there (~3.5% drop)... We suggest a maximum of 3% voltage drop and aiming for ~1% voltage drop for solar wiring (less than 1% drop is very expensive and does not gain much).

    Looking ahead, this is when you think about putting 2x 100 Watt panels in series and using a MPPT type solar charge controller... These can take high voltage/low current from the array and efficiently down convert to low voltage/high current needed to charge the battery bank (using much less copper--Longer runs and/or thinner cable). So, say you have 2x 100 watt panels  in series and 2 parallel strings (4 panels total), for Imp-array ~ 11 amps and Vmp-array ~ 36 volts with 20 feet and 10 awg cable:

    Voltage drop: 0.44
    Voltage drop percentage: 1.22%
    Voltage at the end: 35.56

    Same 10 AWG cable, almost 2x the power, and close to 1% voltage drop--Of course, you would have to get a "not cheap" MPPT charge controller.

    Then you start looking at >>140 Watt solar panels with Vmp~30 volts--Many times these are less expensive panels (used for grid tied systems--Although shipping a few large format GT panels can be very expensive)--Keep an eye on the end results. Do not expand without looking ahead to the eventual goals--It is very difficult to cost effectively expand a small solar power system into a large solar power system a piece at a time (like trying to upgrade a VW bug into a 1 ton pickup---Can it be done, use, but is a wise thing to do, not usually).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mike49
    mike49 Registered Users Posts: 28 ✭✭
    vtmaps said:
    mike49 said:
    With 3 panels in parallel, do I need add
    a) a fuse for each panel before the connection to the 3-way parallel connector?
    b) a blocking diode for each panel before the connection to the 3-way parallel connector?
    c) a combiner box with fuses and blocking diodes in lieu of a) and b) ?
    Welcome to the forum,

    You don't need blocking diodes at all (with any number of panels) because your controller prevents the battery from backfeeding the panels.

    Yes, you do need a fuse for each panel whenever there are more than two panels in parallel.  It's best to do this with a combiner box, but it can be done with in-line fuses "before the connection to the 3-way parallel connector".  If you use a combiner box, circuit breakers are recommended over fuses.

    Do you know what size fuse/breakers to use?

    --vtMaps
    Hi, thanks for the reply back, no diodes needed is good news, since I'm not familiar with wiring those, but I do have spare mc4 connectors, 10gauge wire and fuse holders/fuses, so adding a fuse for each panel's positive wire should be easy enough.

    So, for the two 100watt Imp5.5a panels, would a 10a fuse each  be ok? (I don't have any 7a fuses), and for the 40watt Imp2.3a panel, would a 5a fuse would be ok? (i don't have any 3a fuses)

    And with these fuses, is it supposed to be preventing a worst case scenario that the 40watt panel is shorted and 5.5a+5.5a=11a flows through that shorted panel? would the shorted panel catch fire or something if there was no fuse? just trying to understand the reasoning for the fuses, thanks again!
     
    small hobby/learning system in Toronto, Canada:
    3x100w parallel via 30A PWM to old 12v car battery to 200w inverter for shed LED lights
    (1x100w may-sept)
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Yep... You got it... The fuse rating to to protect both the wiring used on the panel, and the panel interior connections from fire if there is an unplanned event. If your panels do not list a series fuse rating, then approximately:

    5.5 amps * 1.25 NEC wiring derating * 1.25 NEC Solar derating = 8.6 amps or ~9-10 amp fuse

    Just to give you an idea of what "gray market" panels (Not UL/NRTL Listed) can do when things go wrong:

    http://forum.solar-electric.com/discussion/3375/panel-fire-question

    The above was not specifically because of a series fuse issue--But to show that you put enough solar panels on your roof, and bad things can happen. 40 Watts, a short is not likely to do much... Get 500+ watts, that is a lot of energy that dumps into a short/arc fault.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    BB. said:
    5.5 amps * 1.25 NEC wiring derating * 1.25 NEC Solar derating = 8.6 amps or ~9-10 amp fuse


    That double 1.25% rating is interesting and not always correct... it depends on the fuse or circuit breaker. 

    Some fuses (breakers) will blow (trip) when they carry their rated current for a few moments.  Other fuses (breakers) are designed to carry their rated current indefinitely.  

    If your fuse (breaker) is designed to carry its rated current, then you only need to use one of the 1.25% deratings.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • mike49
    mike49 Registered Users Posts: 28 ✭✭
    Thanks BB for the great infos, I think I get it with your detailed explanations (finally yay)!

    Let's see if I really get it:

    So even with my 2 panels in parallel right now, if I'm understanding things correctly, since I have no idea what the series fuse rating is for the 40watt panel and go by your suggested estimation for it:
    2.3 amps Imp * 1.25 NEC wiring derating * 1.25 NEC solar derating = 3.6 amps ~ 4 to 5 amp fuse/breaker

    It would be possible in my 2 parallel panel setup currently for my 100watt Imp5.5a panel to backfeed 5.5a to the 40watt panel if it gets shorted, which would exceed the estimated 3.6a the 40watt panel can handle, so I should put a fuse right now on that 40watt panel, right?

    small hobby/learning system in Toronto, Canada:
    3x100w parallel via 30A PWM to old 12v car battery to 200w inverter for shed LED lights
    (1x100w may-sept)
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Yes.

    Note that 1x 100 Watt panel is 5.5 amps Imp. 2x 100 Watt panels in parallel is 11 amps Imp.

    And Isc is a bit larger than Imp. Short curcuit vs max power current.

    Also, most standard fuses and breakers should run at a maximum continuous current of 80% of the marked rating. There are a few solar companies breakers that will run at 100% of marked rating. As always check the specifications.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mike49
    mike49 Registered Users Posts: 28 ✭✭
    BB. said:
    Looking ahead, this is when you think about putting 2x 100 Watt panels in series and using a MPPT type solar charge controller... These can take high voltage/low current from the array and efficiently down convert to low voltage/high current needed to charge the battery bank (using much less copper--Longer runs and/or thinner cable). So, say you have 2x 100 watt panels  in series and 2 parallel strings (4 panels total), for Imp-array ~ 11 amps and Vmp-array ~ 36 volts with 20 feet and 10 awg cable:

    Voltage drop: 0.44
    Voltage drop percentage: 1.22%
    Voltage at the end: 35.56

    Same 10 AWG cable, almost 2x the power, and close to 1% voltage drop--Of course, you would have to get a "not cheap" MPPT charge controller.

    Then you start looking at >>140 Watt solar panels with Vmp~30 volts--Many times these are less expensive panels (used for grid tied systems--Although shipping a few large format GT panels can be very expensive)--Keep an eye on the end results. Do not expand without looking ahead to the eventual goals--It is very difficult to cost effectively expand a small solar power system into a large solar power system a piece at a time (like trying to upgrade a VW bug into a 1 ton pickup---Can it be done, use, but is a wise thing to do, not usually).

    -Bill
    Hi Bill, thanks for your comments regarding getting an mppt controller and putting panels in series. This is actually the next step I have planned for my hobby/learner system, but am holding off on it until my purchasing power gets back to something reasonable since the canadian dollar is in the dumps right now and probably going to take awhile to get back up there so I can economically by an mppt controller....

    But since I've been reading all this stuff about putting panels in parallel and series, I just wanted to make sure I understand - with 2 identical 100w panels and a 40 watt panel, I would only be able to put the 2 100 watt in series and NOT put the 40watt panel in series, because the current from the other panels would not be able to flow through the 40watt panel, right? thanks again :)
    small hobby/learning system in Toronto, Canada:
    3x100w parallel via 30A PWM to old 12v car battery to 200w inverter for shed LED lights
    (1x100w may-sept)
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    That is correct. Unless you get a second 40 Watt panel to put in series worth the first.

    Although, as the system gets bigger, you are generally better off not using small wattage panels. The extra wiring connections, fuses, etc cost per Watt are high for small panels.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mike49
    mike49 Registered Users Posts: 28 ✭✭
    I found a 3a fuse and have it inline now on my 40watt panel which is parallel with my 100watt panel, now I feel really safe, I'm also ready for the 3rd panel when that comes in and will likely fuse all 3 since it's easy enough to do and i like to be safe as possible.
    Thanks for all the great info, it's fun learning about this stuff.

    --

    So, when I eventually get do get an mppt controller and put identical or Imp within 10% panels in series, I'm confused by this website that is suggesting the use of bypass diodes in the event of an entire panel being shaded:

    http://www.mpptsolar.com/en/solar-panels-in-series.html

    Shading of a panel is possibility here when snow blows around and accumulates on one panel more than the others etc.
     
    What I don't get is how the wiring for these bypass diodes "in parallel" is done between the panels when in series?
    a) wouldn't wiring things like how they have it in the diagram create a short circuit in the series through the parallel branch with the diodes?
    b) how would I physically wire these diodes like they have in the diagram, is there some sort of special connector/splitter where I can make that "parallel" branch with the diodes?

    thanks again :)


    small hobby/learning system in Toronto, Canada:
    3x100w parallel via 30A PWM to old 12v car battery to 200w inverter for shed LED lights
    (1x100w may-sept)
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Most (all?) good quality solar panels have bypass diodes already wired in the panel.... More or less, solar cells go "high resistance" when they are shaded and block current flow. And, if the cell gets more than ~12 volts or so of "reverse bias", then the cells can be damaged (ruined).

    The bypass diodes are usually installed in the Junction box of the solar panels (the black cylinders with the silver polarity marking band).



    Imagine you had a 6 cell (1.5 volts per cell) flashlight... And you had one battery go "dead". You can put a reverse biased diode on each cell--That way, if there was one dead/open cell, you would still get enough voltage (and current) to the load to keep it running. That is what the diodes are doing for solar panels. A shaded solar cell looks like an "open" circuit (which can be damaged by too high of reverse voltage).

    Older panels (12 volt or less designed for battery charging) may not have bypass diodes, and when used in simple battery charging systems--A blocked cell simply stops current flow in that panel. And the shaded cell(s) will not be damaged. Plus, "low Vmp-array" voltage systems, if you have a cell shaded, the string voltage fall too much to be useful to charge the batter bank (remember a Vmp~18 vdc panel will typically have two bypass diodes--18v/2=~9 volts--not useful for charging your 12 volt battery bank).

    When you have 2-4 or more panels in series (very common to have 10 panels or more in series--Vmp~30 volts, Vmp-array 300 vdc or higher), the current GT type (UL/NRTL approved) solar panels must have by by pass diodes in each panel--Or the cells will be ruined and possibly cause a fire.

    This can be an issue when using older/small (non-UL/NRTL listed panels) in mixed/expanded solar arrays. Placing (for example) multiple small solar panels in series for MPPT solar charge controllers (Vmp-array upwards of 100 VDC), or with higher voltage battery battery banks (24/48 vdc)--Then you may need to put a bypass diode on each smaller/older 12 vdc solar panels.

    When you expand your system, I highly recommend that you get UL/NRTL Listed solar panels and do not use older/smaller panels in your MPPT string. Adding diodes externally is a pain (they need insulating, heat sinking, and weather proofing).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mike49
    mike49 Registered Users Posts: 28 ✭✭
    Hi Bill, thanks for the clear picture and explanation of the bypass diodes being already in the junction box.
    I finally had a chance to open the junction boxes on my panels:
    a) my 100w panel's junction box is identical to your picture.
    b) my 40watt is different - picture of that attached here

    So I'm wondering about these diodes in the junction boxes:

    a) for the 100w panel with the 2 diodes, since those 2 diodes are bypass diodes, does that mean that it doesn't have a "reverse current blocking diode" that is supposed to prevent nighttime backflow of current to the panel from the battery in case of charge controller malfunction?

    b) for my 40w panel, there is a single diode wired differently than the 100w panel, am I guessing correctly that this diode is not a bypass dioide but is a "reverse current blocking diode"?

    Thanks again for the fun learning and happy new years to all :)

    small hobby/learning system in Toronto, Canada:
    3x100w parallel via 30A PWM to old 12v car battery to 200w inverter for shed LED lights
    (1x100w may-sept)
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    The 40 watt panel appears to only have a blocking diode, no bypass diodes--As you said.

    The 40 watt panel appears to be designed to be directly connected to a battery bank... And should not be placed in series for 24+ battery bank (unless you add your own bypass diodes).

    Also, panels that are designed for direct connection to battery banks without a charge controller, typically have Vmp~15 volts or so (30 cell panels instead of the common 36 cell panels with Vmp~18 volts for use with a solar charge controller).

    Do you have the Vmp/Imp ratings for your 40 watt panels?

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mike49
    mike49 Registered Users Posts: 28 ✭✭
    My 40w panel is Vmp 17.1V / Imp 2.3A
    (In parallel currently with my 100w panel Vmp 18.2V / Imp 5.5A)

    Curious - could the blocking diode on my 40w panel prevent backfeed from the parallelled 100w panel in the event the 40w panel get shorted as we were discussing earlier in the thread? Since I already made a 3a fuse for this 40w panel I'll still keep using it, but just curious to know if this blocking diode would serve the same purpose :) thanks again...

    small hobby/learning system in Toronto, Canada:
    3x100w parallel via 30A PWM to old 12v car battery to 200w inverter for shed LED lights
    (1x100w may-sept)
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Yes, the diode will block reverse current flow.

    However, diodes can fail sorted or open. The powers that be, require that two diodes be used in series to protect against a single point of failure. Or one NRTL rated fuse/breaker.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mike49
    mike49 Registered Users Posts: 28 ✭✭
    Well, it ain't NTRL anything but I sure feel double safe with both a blocking diode and 3a fuse on that little 40w puppy o:)

    And since a little knowledge is a dangerous thing, if one day I was silly enough want to add a bypass diode for whatever reason to the 40w panel (I can't think of one right now since I don't intend to put this little panel ever in series), would it be a simple matter of just soldering a bypass diode to the positive and negative terminals inside the junction box? >:)
    small hobby/learning system in Toronto, Canada:
    3x100w parallel via 30A PWM to old 12v car battery to 200w inverter for shed LED lights
    (1x100w may-sept)
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Yep. You connect the diode so no current flows during normal operation.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset