best way to wire up 36 280W panels

yrtrnc
yrtrnc Solar Expert Posts: 54 ✭✭
So I have 36 Trina solar TSM-PC14 280W panels. And Im using BlueSolar charge controller MPPT 150/70. My battery bank is 48V. How can you advise me to setup these panels. 4 panels connected in series, and 3 series connected in paralel with 3 charge controllers? Is that ok? Installation location is Nicosia , Cyprus. Lowest recorded annual temp is -1 degrees highest 42 degrees. Thanks in advance!

TSM-PC14 280W

Electric properties
Nominal power 280.00 Wp
Performance tolerance 3.00 %
MPP-voltage 36.00 V
MPP-current 7.78 A
Open-circuit voltage 44.40 V
Short-circuit current 8.33 A
Permissible system voltage 1000 V
PV module efficiency (STC) 14.43 %
Grounding recommendation No grounding
Mechanical properties
Number of cells in the PV module 72
Width 992 mm
Length 1956 mm
Weight 28.00 kg
Cell technology poly
Temperature coefficients
MPP-voltage --- ---
Open-circuit voltage -0.3300 %/°C -146.5 mV/°C
Short-circuit current 0.0460 %/°C 3.83 mA/°C



Charge controller is a Victron 150 70 mppt:

BlueSolar charge controller MPPT 150/70
Nominal battery voltage 12 / 24 / 36 / 48V Auto Select
Rated charge current 70A @ 40°C (104°F)
Maximum solar array input power 12V: 1000W / 24V: 2000W / 36V: 3000W / 48V: 4000W
Maximum PV open circuit voltage 150V

Comments

  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: best way to wire up 36 280W panels

    You do not want to put four of those panels in series on that controller: 4 * 44.4 Voc = 177.6, which exceed the V max of 150 even without and temperature adjustment. Even three in series would be 133.2, and with a low of -1C could still exceed the max.

    It would be best to use only two in series. This would give you a string Vmp of 72, which is sufficient for a 48 Volt system, and a Voc of 88.8 which is not going to exceed the V max of the controller.

    Each controller can handle 70 Amps maximum, so in theory that's about 3360 Watts per controller. That would be 12 panels per controller; six parallel strings of two in series. This works for thirty-six 280 Watt panels on three controllers.

    Most of the time under peak sun conditions the panels will be hot and their output Voltage lower, so you will likely see less than maximum current available from each controller. Most likely you will see about 54 Amps peak per controller.
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Re: best way to wire up 36 280W panels

    you may be cutting it close in 2 ways.

    1stly the voc is too high with 4 pvs in series. now with 3 in series could go within 5v of the cc max during a -1 degree c condition. i'd double check with the cc manufacturer that the controller can operate to 150v and what their opinion would be for having those pvs approach about 145v at -1 degrees. even colder weather than -1 can manifest itself into cc failure even if the manufacturer says ok the -1 point.

    way #2 is you are operating all 3ccs at their max wattage. this will lead to failure for sure in the form of excessive heating. it may be better to run 4 ccs rather 3. this is for a 3x3 pv arrangement per cc and the other with 3 ccs would look like 4 strings of 3 in series per cc rather than the 3 strings of 4 in series you proposed as that would exceed the pv max operating ability.
  • yrtrnc
    yrtrnc Solar Expert Posts: 54 ✭✭
    Re: best way to wire up 36 280W panels

    I guess it would help if I said the temp. is in C. And that its an off grid system.

    This is before I saw the two below posts ;)
  • yrtrnc
    yrtrnc Solar Expert Posts: 54 ✭✭
    Re: best way to wire up 36 280W panels

    Thanks for the speedy reply guys, means alot!

    It does get rather hot, so you think 3x3 is best? so 9+9+9+9? I also like Cariboocoot idea of 2x6 12+12+12. Because I will be using 3 controllers instead of 4. But if it means I will put extra wear and tear on the equipment because of heat I'd rather use 4 cc's. If its not so bad . It says 48V system (solar array up to 4000W) in the cc manual. What do you guys think? Im installing tomorrow :)
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: best way to wire up 36 280W panels

    First, you need to figure out the number of panels per string.

    If you go with 2, then your Vmp derated for 70C (it gets hot in Cyprus, right?) becomes 60V. The controller requires battery temperature +7V to start. If you want to charge at 58V (given 48V system), you need 65V. 60V is less than this, so strings of 2 will not work for you very well.

    If you go with 3, then your Voc at -1C becomes 145V. This is a little bit too close to 150V controller limit, and the docs for the controller doesn't mention over-voltage protection, so that's on the edge, but in theory it's acceptable. Aside of using different controllers, that's your only choice.

    This leaves you with 12 strings, each consisting of 3 panels.

    Each controller can handle 4000W array for 48V system and they claim they have good temperature protection. 4 strings is 3360W, which is 84% of rated power. Each controller can handle 4 strings.

    So, IMHO, you'll need 3 controllers, each handling 4 strings, each string consisting of 3 panels.
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Re: best way to wire up 36 280W panels

    i think it sounds like you already had your mind made up. i advised to talk to the cc manufacturer on how to best arrange this that won't be detrimental to the cc. 4 in series will kill the controllers as will operating the controllers close to their max power all of the time. 3 in series has potential if they say it will operate well to your specified temperature. 48v nominal battery voltage x 70a cc capacity = 3360w and 36 pvs x 280w = 10080w and 10080w/3 ccs = 3360w per cc and 10080w/4ccs = 2520w per cc. you can still put the pvs in place, but hold off wiring until you are sure about the arrangement working properly.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: best way to wire up 36 280W panels

    70C? No, I don't think so. :D

    25C is normal for panel rating.
    The panel Vmp is 36, so two in series would be 72 Volts. Even at 42C ambient that leaves enough to run about 60 Volts into a 48 Volt system. According to the Trina data I just looked at the panels will lose 18 Volts at their maximum operating temp of 85C (185F). In that case you might have a problem, but this is a highly unlikely scenario.
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: best way to wire up 36 280W panels
    niel wrote: »
    4 in series will kill the controllers as will operating the controllers close to their max power all of the time.

    I believe you meant "4 strings", not "4 in series". 4 in series will definitely kill controllers because of overvoltage.

    The nominal output from the panels is 3360W. If you derate for temperature (because Cyprus is very hot), it'll give you 3000W. This is what you get at the daily peak. And most of the time it'll be less. 3000W is only 75% of the controller rating. So they will be usually operating at 50 to 75% of rated power. Why would that kill controllers?
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Re: best way to wire up 36 280W panels
    NorthGuy wrote: »
    I believe you meant "4 strings", not "4 in series". 4 in series will definitely kill controllers because of overvoltage.

    The nominal output from the panels is 3360W. If you derate for temperature (because Cyprus is very hot), it'll give you 3000W. This is what you get at the daily peak. And most of the time it'll be less. 3000W is only 75% of the controller rating. So they will be usually operating at 50 to 75% of rated power. Why would that kill controllers?

    sorry, but i meant what i said the way i said it as 2 things could kill the ccs with excess voltage as 1 and running the wattage towards the upper end of the cc's ability as the other. the wattage will not be as apparent but will shorten the lifespan to maybe just at or beyond its warranty period. the manufacturers guarantee it before that warranty time period is up so they feel it will be good that long at that wattage. the excess voltage will kill it almost immediately.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: best way to wire up 36 280W panels

    Should mention here that it is difficult to translate ambient temperature to panel temperature directly, and panel temperature is what affects Voltage.

    Orientation can play a role too: if the panels fast easterly then the charging could be complete before maximum panel temperature is reached (usually after midday when sun has done its worst) and the lowered Voltage isn't as much of a problem because you're now into the slightly lower V needs of Float rater than Absorb.

    Wiring will play a role as well: minimizing V drop in the wiring will increase the actual Voltage available to the controller. If there is a very long run from array to controller you might risk the other direction and allow the wire to lose some Voltage while running three in series, but keep in mind that such V drop is at its most under maximum power conditions so any allocation made to keep Voc at safe levels will be much worse under full power charging.

    This is a case where it would be nice to be able to have a thermostatically controlled "extra" panel of the same current but just a couple of Volts that would be in series when temps are high but bypassed when temps are low. In theory this is possible, but I'm sure it's not practical to bring about.

    Curiously the Trina 270 Watt panels with Vmp 32.1 and Voc 38.3 might actually work better if used three in series (Vmp 96.3 is more than enough to charge 48 Volts even in very hot temps, Voc 114.9 won't go over 150 at -1C). You would lose 360 Watts over-all from the array.
  • solarix
    solarix Solar Expert Posts: 713 ✭✭
    Re: best way to wire up 36 280W panels

    If its not too late, consider using a charge controller with higher Vin like the Midnite Classic 200 which can do 4 maybe 5 modules in series. This unit is designed with a "Hyper Voc" feature which allows for exceeding the Vin in certain conditions. Check out their string sizer on line.
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: best way to wire up 36 280W panels
    70C? No, I don't think so. :D.

    That's not an ambient temperature, that's a panel temperature, which is usually 20-30C higher than ambient. At 25C ambient, the average operationg temperature is 49C (commonly reffered to as Normal Operating Conditions - NOCT). At higher ambient and without wind it can easily go to 70C.

    At any rate, it is not a good idea to keep the voltage so close to the charger's cut-off point. I would rather allow some margin.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: best way to wire up 36 280W panels
    NorthGuy wrote: »
    At any rate, it is not a good idea to keep the voltage so close to the charger's cut-off point. I would rather allow some margin.

    Me too. Hence the different panel idea or solarix's suggestion of a MidNite that has a higher V max in.

    And if you think this install is fun try calculating for Lillooet, BC where the yearly temperature swing is over 100C!
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: best way to wire up 36 280W panels
    And if you think this install is fun try calculating for Lillooet, BC where the yearly temperature swing is over 100C!

    We have only 90C swing. I first bought controllers, then started shopping for panels (bad idea), and I realized that I cannot use many of the panels because when I combine them into the strings it's either too little or too much.

    They probably have lower swing in Cyprus, so it should be easier. OP just has a bad match. Classics 200, 250, or even XW80-600 would be easier to fit to these panels, but they may not be available in Cyprus.
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Re: best way to wire up 36 280W panels

    northguy,
    don't forget that it does get to -1 c there at times and most likely it would occur in the morning right about sunrise allowing for quite a high voc. if anything else is involved that may amplify the light even more or they get colder than the -1 c, then it could easily pop the cc. you can't lump it like you did for average times as you must go by a worst case scenario.
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: best way to wire up 36 280W panels
    niel wrote: »
    northguy,
    don't forget that it does get to -1 c there at times and most likely it would occur in the morning right about sunrise allowing for quite a high voc. if anything else is involved that may amplify the light even more or they get colder than the -1 c, then it could easily pop the cc. you can't lump it like you did for average times as you must go by a worst case scenario.

    I agree. 2 panels in series is almost too little, 3 in series is almost too much. The best way is to use a different controller, or perhaps different panels.
  • gww1
    gww1 Solar Expert Posts: 963 ✭✭
    Re: best way to wire up 36 280W panels

    Would a small resistor in the line from the pv keep the voltage below voc? I read a post on the outback forum that mentioned this.
    gww
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: best way to wire up 36 280W panels
    gww1 wrote: »
    Would a small resistor in the line from the pv keep the voltage below voc? I read a post on the outback forum that mentioned this.
    gww

    No, it won't.
    The reason being Voc is virtually unaffected by line resistance due to the zero current. I did mention line loss due to resistance before, but also with the caveat that enough to affect the Voc would severely deplete the Vmp.
    Since you need more Voltage drop to keep Voc from going over the max and less to keep the Vmp high enough to charge you get a paradox for the resistance function.
    You can check this with a V-drop calculator:
    Set a low Voltage like 12, low current like 1, short distance like 20 feet, and typical 10 AWG wire size. V-drops to 11.95. Now change to smaller wire size 14 AWG to give greater line resistance. V-drops to 11.89. Hardly any difference. (That's about 1 Ohm per 1,000 ft vs. about 2.5 Ohms per 1,000 feet, so either is hardly much resistance at all.)
    Now run the calculations again with 10 Amps current. You get 11.58 and 10.96 respectively. The higher current which will be present at Vmp will cause a greater drop in Voltage than the change in line resistance.
    At the Voltage range of this proposed three or four panel string (around 120) the effect is even less: 20 feet of 10 AWG @ 1 Amp drops to 119.95 and with 10 Amps 119.59. Decrease wire size to 14 AWG (again, more than double the line resistance) and you get 119.89 @ 1 Amp and 118.96 @ 10 Amps.
    To take it to an extreme using the high Voltage, to lose 1 Volt from 120 Volts @ 1 Amp over 20 feet you need 24 AWG wire (about 25 Ohms per 1,000 feet, so 0.5 Ohms). Increase the current to 10 Amps on the same wire (as you would find while charging) and the Voltage drops to 109. 10X the current, 10X the Voltage drop. As I said this functions exactly backward of how you want it to work for this situation.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: best way to wire up 36 280W panels

    I think the suggestion would be to put a load resistor across the output of the solar panel... But it would appear that you would need significant loading to really reduce the output Voc significantly and reliably--Somewhere north of 10% of the rated output current of the array. Which would significantly reduce the panel's useful output.

    And if you really want to reduce output votlage in cold weather, you are probably looking at 80% of Imp to hit the "knee of the curve". for example, look at this I*V curve:

    http://lib.store.yahoo.net/lib/wind-sun/Trina-TSM-240PA05-specs.pdf

    The voltage slope is pretty vertical (right side of the curve) until the Pmax (Imp/Vmp) point of the curve gets into the current limited mode of the solar array.

    In the olden days (pre-Midnite and Xantrex "high input voltage" controllers), I have suggested looking at adding a switch to the solar array to cut out one panel out of the string when very cold weather is forecast (or during the winter).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: best way to wire up 36 280W panels

    You'd want to take that load resistor out when charging.

    This sort of thing is what we Canadians sit around thinking about on cold Summer nights: how to handle that hyper Voc caused by superconducting at -70C temps on Baffin Island. Lots of schemes about thermostatically controlled panel switching. Every one having some flaw in it somewhere. Usually a question of reliability and the dire consequences of failure.

    Our other hobby is watching the Aurora Borealis and thinking of ways to harness that energy. No practical answers there yet either. :p
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: best way to wire up 36 280W panels
    Our other hobby is watching the Aurora Borealis and thinking of ways to harness that energy. No practical answers there yet either. :p

    We see northern lights much less in the last years that it used to be. Perhaps, someone already harnessed the energy.
  • gww1
    gww1 Solar Expert Posts: 963 ✭✭
    Re: best way to wire up 36 280W panels

    Does a midnite classic voltage clipper only work with wind?
    gww
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: best way to wire up 36 280W panels
    gww1 wrote: »
    Does a midnite classic voltage clipper only work with wind?
    gww

    Interesting question.
    It is probably overkill for a PV array (and $1,000 for the 1500 Watt DC model).

    In theory, if each panel had a Zenor diode across its output rated for Voc and the panel's current this would clip any V over Voc. But in practice ... Where do you get a 45 Volt 10 Amp Zener diode?

    Other wild ideas for Voc clipping:

    Thermostatic controlled 'last panel in string' shorting.
    Low Voltage, same Amperage 'tag end' panels to compensate for V-drop due to heat without adding too much Voc.
    Hi-V capacitor clipping

    Sane ideas:

    MidNite Classic 200 or 250
    Using panels of lower Voltage so the problem doesn't arise to begin with
    Xantrex XW 600V 80 MPPT for those extreme cases

    :D
  • gww1
    gww1 Solar Expert Posts: 963 ✭✭
    Re: best way to wire up 36 280W panels

    Just trying to learn as I go. Here is the thread that I previously refered to.

    http://www.outbackpower.com/forum/viewtopic.php?f=9&t=346&p=4341&hilit=gvfx3648#p4341

    I am waiting for an answer to how the below expeirment turned out.

    http://www.fieldlines.com/index.php/topic,147100.0.html

    Thanks
    gww
  • yrtrnc
    yrtrnc Solar Expert Posts: 54 ✭✭
    Re: best way to wire up 36 280W panels

    Thank you everyone for all your comments and ideas. Much appreciated.

    We didnt have a chance to change controllers since we already bought them (next time I will be more carefull when considering!). So we opted for the 3x4 (12+12+12) . Not sure how you say that, 4 strings of 3 panels? Was that correct? Yeah, thats what we did. I know we are cutting it close , so if it ever does go down to -1C Im installing 40 amp dc breakers so they can be manually switched off at such a rare event. This happening is highly unlikely. Victron says that their controllers will handle the amps. So all is looking good for the moment.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: best way to wire up 36 280W panels
    yrtrnc wrote: »
    Thank you everyone for all your comments and ideas. Much appreciated.

    We didnt have a chance to change controllers since we already bought them (next time I will be more carefull when considering!). So we opted for the 3x4 (12+12+12) . Not sure how you say that, 4 strings of 3 panels? Was that correct? Yeah, thats what we did. I know we are cutting it close , so if it ever does go down to -1C Im installing 40 amp dc breakers so they can be manually switched off at such a rare event. This happening is highly unlikely. Victron says that their controllers will handle the amps. So all is looking good for the moment.

    Yes, four parallel strings of three in series.

    The breakers are a good idea, but not for why you think. With more than two parallel connections you need circuit protection in case something shorts in one of the strings because the others can then feed it more current than it is designed to handle.

    The breakers will do nothing for the Voc problem. There is no high current involved with that because the panels are not under load. If they were they wouldn't be at Voc. You will still have the risk of zapping the charge controllers' inputs with more Voltage than they are designed to accept.