Pwm with mixed 12v nominal panels on trailer

jimbuck
jimbuck Registered Users Posts: 46 ✭✭
Hello, I’ve had a look through previous posts but can’t find anything concrete.

I am installing a small pwm based system on a trailer roof.

Ideally I would have put 2x 230w panels in series or parallel down to a MPPT controller to the 12v bank.

Unfortunately I don’t have the room for the larger GT panels so have opted for Pwm and 17/18vmp panels. Can I fill all the gaps between roof vents etc with different size similar vmp panels all wired in parallel down to a pwm controller? As the panels are a current source I presume that each will give what they can when the battery needs charging. Is there any downside to this setup?
I’ve currently got 2x 100w panels and think I can fit a few 50w and 80w panels up there.

Thanks for any help you cast give.

Comments

  • littleharbor2
    littleharbor2 Solar Expert Posts: 2,038 ✭✭✭✭✭
    As long as the Vmp. is within 10% of each other you should be fine. 
     Avoid placing panels up close to vents, antennas and AC units because any shadowing on them will kill their output till clear of shadows.

    2.1 Kw Suntech 175 mono, Classic 200, Trace SW 4024 ( 15 years old  but brand new out of sealed factory box Jan. 2015), Bogart Tri-metric,  460 Ah. 24 volt LiFePo4 battery bank. Plenty of Baja Sea of Cortez sunshine.

  • mcgivor
    mcgivor Solar Expert Posts: 3,854 ✭✭✭✭✭✭
    In addition to littleharbor2's comment, if using a PWM controller the panels would be in parallel, if one were to be shaded it would not affect the others, unlike a series configuration with MPPT, where one shaded panel in a string cripples the ballance of panels in said string. 
    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.
  • jimbuck
    jimbuck Registered Users Posts: 46 ✭✭
    Thanks littleharbour and mcgivor, most of the panels are around 17.5/18.5 vmp so I’m guessing are ok. I can raise them above any vents etc. Do you know the reason that vmp needs to be within 10%, I’m just curious, I like to understand things. Thanks for your help, it’s much appreciated.
    Jim
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    It is getting difficult to find IV and Power curves for solar panels... Here is an older thread where I copied some old diagrams into the thread for archive:

    http://forum.solar-electric.com/discussion/5458/two-strings-in-parallel-with-unequal-string-voltages

    Part of the confusion is the fact that details matter... For PWM charge controllers, they simply connect (through a transistor/switch) from the battery to the solar panel. And no "conversions" occur. Simply, 14.5 volts at the battery, close to 15.0 volts at the input to the PWM charge controller, and ~16.0 volts (wiring drop) at the solar panels. You look that the IV curve for current (under full sun) vs panel voltage, and that is what you get. Think if this as a one speed bicycle.

    However, if you have a MPPT charge controller, this can take high(er) voltage and low(er) current from the solar array, and efficiently (upwards of 95% efficiency) down convert to the low voltage/high current needed to charge the battery bank. More or less, this is like the 18 speed bicycle that matches the human leg force/speed (RPM) to the conditions that the rear wheel is "seeing" (i.e., low gear going up hill, higher gear when on flat ground heading down wind).

    Here you need to "match" the Vmp of each panel/string of the array so that all the panels have the same operating point. I.e., if you put a low speed diesel motor in parallel with a high speed gasoline engine (both are connected together and run at the same RPM), it is not possible to have a transmission gear ratio that will get maximum power from the diesel and the gasoline engine at the same time (i.e., 1,500 RPM diesel vs 4,000 RPM gasoline). And you will probably have another peak in between where the diesel and gasoline engine torque adds together).

    Very roughly, if your paralleled solar panels are rated to Vmp to roughly within 5%-10% of each other, they are close enough that you will not lose very much power (Vmp*Imp=Pmp) from the "ideal" (the Power Curve is sort of a soft rounded peak, and being off by 5-10% is not very much change). If you have a 18 Vmp panel, then +/- ~1.8 volts of Vmp 18 (i.e., 16.2 volts to 19.8 volts matching of panels) is going to be OK (there are other issues, such as Vmp-array minimum of ~17.5 volts recommended for properly charging a 12 volt battery bank over temperature range of panels and battery bank).

    And for panels in series, again, that Imp match to within 5-10% is recommended too.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • jimbuck
    jimbuck Registered Users Posts: 46 ✭✭
    Thanks Bill for the comprehensive reply. I think I understand the I V power curve. I think it’s more relevant to a MPPT system?

    For example, in a pwm system the panels will be at VBatt and let’s say that the battery is at 50% soc so accepting lots of current. Does this mean that the panels will be giving all that they can?
    I know that due to the panels being 17.xxv vmp I will loose the power in between panel voltage and vbatt that I would gain with MPPT. But the nameplate Imp of the panel under perfect conditions (I know that doesn’t happen, STC vs NOCT etc) is what I should see?

    Sorry for my lack of clarity in writing, it’s not a strong point. Maybe I’m trying to understand beyond what I need to and in time it will all click.

    I do understand the need for correct Vmp panels for a pwm system which gives sufficient voltage left after cable runs etc to still equalise a battery bank. And that a excessively high vmp eg 22v on a 12v nominal system would be more inefficient as the extra voltage would not be converted to power like in an MPPT system.

    I’ve allways tried to go with the latest technology and started with MPPT but with this system I’m limited with space and funds hence the pwm and mixed sized panels.

    I’ve git a lot more reading to do.

    Thanks, jim
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    For systems smaller than ~400 Watts, PWM is usually just fine and may even be a bit less expensive (pwm controllers are cheaper than MPPT, but 140 Watt and smaller "12 volt" panels are usually more expensive than >200 Watt Vmp~30 volt panels $$$/Watt).

    Also, if you have a longer distance from the solar array (over a few 10's of feet) to the charge controller+battery bank, then an MPPT charge controller can be a better design (higher Vmp-array and smaller diameter copper wire from array to charge controller).

    Many times, your best bet is to do a quick paper design of two (or more) system options and see what works out best for you.

    More or less, when the panel is at Vmp or less (like with a PWM controller), yes, you do lose some power... The equation of Power=Voltage*Current tells you that. Panel operating at Vmp~18 volts vs Vbatt~14.8 volts charging..

    However, remember that solar panel Vmp falls as they get hot, and on a hot/windless day, the panels can lose almost 20% of their output voltage (18 volts * 80% ~ 14.4 volts)--So there is no real voltage headroom/difference between the operating point of a MPPT controller and the PWM controller with nominal operation. If the panels are in below freezing weather (for example), then a MPPT controller can give you some 10-20% more power than an MPPT--But it is usually not worth designing the system for this bit of extra power, vs just getting a 20% larger solar array.

    In summary, larger systems (typically over ~800 Watts), systems with longer distances from array to charge controller, and if you want remote monitoring (MPPT controllers are more expensive, more computer processing power and have extra remote monitoring capabilities than typical PWM controllers), those are the usual reasons for choosing an MPPT controller over an PWM controller.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Photowhit
    Photowhit Solar Expert Posts: 6,002 ✭✭✭✭✭
    I really like the 400-500 watt number for 12 volt systems.

    ...but with a lot of larger, true 24 watt panels, I think there is a good argument to be made for cost effective higher wattage systems in the 24 volt system voltage.

    It's amazing that you can have a 1200 watt 24 volt system with a $140 charge controller like the C60 by Schneider.... and be under $1000, a quality MPPT charge controller would cost $500 by it's self.
    Home system 4000 watt (Evergreen) array standing, with 2 Midnite Classic Lites,  Midnite E-panel, Magnum MS4024, Prosine 1800(now backup) and Exeltech 1100(former backup...lol), 660 ah 24v Forklift battery(now 10 years old). Off grid for 20 years (if I include 8 months on a bicycle).
    - Assorted other systems, pieces and to many panels in the closet to not do more projects.
  • jimbuck
    jimbuck Registered Users Posts: 46 ✭✭
    I’ve got a 21’ round distance 10.5’ each way from the combiner box to the controller so plan on using 2AWG for a 1% drop at 13v 30A. I feel my system is too small for 24v although I can see the benefits, I’m dictated by panel size and roof space only need about 220ah at 12v of battery bank. I think the smallest bank I could make with 6v batteries at 24v would still be 220ah which is twice the capacity that I need. I do like the idea of a 24v system using proper vmp 35v panels and a pwm controller
  • jimbuck
    jimbuck Registered Users Posts: 46 ✭✭
    I lost half the last post.

    @Bill, I took a panel out of storage a few days back and checked voc and then took it into the sun and watched it slowly drop and after about 4 minuted it levelled off, that was an eye opener for me to see how temp affected the panel voltage, even in the uk.

    Jim
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Yep. There is a relatively substantial connection between cell temperature and Voc&Vmp.

    The marketing numbers are at standard test conditions (cells at from temperatures and a few seconds under a sun lamp to get & confirm specifications.

    In real life systems, we have to adjust designs so that the systems work well over time and temperatures.

    Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Raj174
    Raj174 Solar Expert Posts: 795 ✭✭✭✭
    jimbuck said:
    I’ve got a 21’ round distance 10.5’ each way from the combiner box to the controller so plan on using 2AWG for a 1% drop at 13v 30A. I feel my system is too small for 24v although I can see the benefits, I’m dictated by panel size and roof space only need about 220ah at 12v of battery bank. I think the smallest bank I could make with 6v batteries at 24v would still be 220ah which is twice the capacity that I need. I do like the idea of a 24v system using proper vmp 35v panels and a pwm controller
    Are you sure that the PV in to the CC is 13 volts? I hope not. You'll need more than that to properly charge the battery.
    17 to 18 volt panels, so where's the additional loss?
    4480W PV, MNE175DR-TR, MN Classic 150, Outback Radian GS4048A, Mate3, 51.2V 360AH nominal LiFePO4, Kohler Pro 5.2E genset.
  • jimbuck
    jimbuck Registered Users Posts: 46 ✭✭
    edited April 2018 #13
    The panels are vmp 17/18v but because pwm draws the voltage of the panels down to Vbatt I calculated my voltage drop for 13v as if the battery was being charged in bulk mode. If I was using an MPPT controller I would have calculated the drop at the vmp of the panels because the controller would then down convert that to the battery charging voltage. I think I’m correct in this calculation?
  • jimbuck
    jimbuck Registered Users Posts: 46 ✭✭
    @Bill I’m interested to see what temperature my panels will be at here after a few hours sun. When looking at nameplates on panels I rarely see NOCT figures which I think are given for around 45c? This is definitely a learning curve for me. A 100w panel in full sun in summer is not going to give 100w of power due to heat. I think I’ve seen you use a 77% derating figure for panels and cc efficiency and a 52% overall system efficiency. Something I need to take into account.

    Jim
  • Photowhit
    Photowhit Solar Expert Posts: 6,002 ✭✭✭✭✭
    jimbuck said:
    The panels are vmp 17/18v but because pwm draws the voltage of the panels down to Vbatt I calculated my voltage drop for 13v as if the battery was being charged in bulk mode. If I was using an MPPT controller I would have calculated the drop at the vmp of the panels because the controller would then down convert that to the battery charging voltage. I think I’m correct in this calculation?

    No what you are seeing is the charge controller doing it's job of controlling the voltage. Interestingly, the battery will accept less and less current as it approaches fully charged state.
    Home system 4000 watt (Evergreen) array standing, with 2 Midnite Classic Lites,  Midnite E-panel, Magnum MS4024, Prosine 1800(now backup) and Exeltech 1100(former backup...lol), 660 ah 24v Forklift battery(now 10 years old). Off grid for 20 years (if I include 8 months on a bicycle).
    - Assorted other systems, pieces and to many panels in the closet to not do more projects.
  • jimbuck
    jimbuck Registered Users Posts: 46 ✭✭
    Am I right in thinking that the voltage traveling down the cables from the pv array to the controller in a pwm system will be Vbatt as it is a straight through connection in bulk mode?

    The voltage travelling down the cables from the pv to the controller in an MPPT system will be the vmp of the array at that time because the array and battery are not directly connected. The controller uses a st
  • jimbuck
    jimbuck Registered Users Posts: 46 ✭✭
    step down transformer.

    This has been my understanding of how to calculate voltage drop in each system.
    Jim
  • littleharbor2
    littleharbor2 Solar Expert Posts: 2,038 ✭✭✭✭✭
    jimbuck said:
    @Bill I’m interested to see what temperature my panels will be at here after a few hours sun. When looking at nameplates on panels I rarely see NOCT figures which I think are given for around 45c? This is definitely a learning curve for me. A 100w panel in full sun in summer is not going to give 100w of power due to heat. I think I’ve seen you use a 77% derating figure for panels and cc efficiency and a 52% overall system efficiency. Something I need to take into account.

    Jim
    There needs to be a level playing field for rating a module's output and 25c is where that standard was set. NOCT numbers are more, real world for most of us. It does work both ways though. If you place that same panel in a very cold, clear and sunny conditions it can put out higher than it's rated wattage, mostly due to higher voltage output. This is when an MPPT controller can take that higher voltage and give you a substantial boost in energy harvest.

    2.1 Kw Suntech 175 mono, Classic 200, Trace SW 4024 ( 15 years old  but brand new out of sealed factory box Jan. 2015), Bogart Tri-metric,  460 Ah. 24 volt LiFePo4 battery bank. Plenty of Baja Sea of Cortez sunshine.

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Jim,

    You are correct that for a PWM controller the solar panel/array voltage is basically Vbatt (with some wiring and controller voltage drops). Imp=Ibatt (current from panel/array=currenv into battery bank) when the charge controller is in "bulk" charging mode.

    And for an MPPT controller, the equation is (basically) Pbatt=Imp*Vmp*0.95 MPPT controller efficiency.

    The issue is that for a 'hot' solar array, Vmp-stc~18 volts approaches Vmp-hot~Vbatt -- And there is no addition voltage for the charge controller to down convert.

    So, it works out that a "properly designed" PWM system is about the same "efficiency" relative to a MPPT charge controller in nominal operation.

    The ~77% "derating" for PWM vs MPPT controllers works out to be about the same number, but for different reasons. And it does not really help to be "more accurate" as it just gets confusing when penciling out a system design. For example, in sub zero weather, a MPPT system can give you more energy as Vmp-cold is > Vmp-hot -- But you can have so few hours of sun in winter (~1-2 hours of sun for winter vs 5-6+ hours of noon time equivalent sun in summer), that the 20% more energy in winter is simply 20% more than a relatively small number to begin with.

    Note that solar system calculations are at best "close" to real numbers... If you are within ~ +/- 10% of predicted values, that is "dead on" for a system design. Between measurement errors, panel temperatures, efficiency differences between morning and noon time operations, actual sun vs average "predicted sun", etc. -- It does not really help to get much closer than our estimates.

    I try to be conservative, but the numbers we use here are pretty close to what most people get on their real world systems.

    There are still many details--For example, a 12 volt MPPT controller should run a Vmp-array-std of ~24 volts or higher... If you run the Vmp-array-std at ~18 volts, the MPPT controllers are not really a good match (they need some voltage drop to operate correctly. PWM controllers have very little voltage drop in normal operation). Each brand/model of charge controller is different... You need to follow the manual to ensure that your system is properly and safely installed.

    Regarding some basic solar math... Here are a couple of examples. Say you have 2x 6 volt @ 200 AH "golf cart" batteries in series. Or a 12 volt @ 200 AH battery bank. And we recommend 5% to 13% typical rate of charge (5% for weekend/sunny weather use system. 10%+ for full time off grid living):
    • 200 AH * 14.5 volts charging * 1/0.77 panel+charger deratings * 0.05 rate of charge = 188 Watt array minimum
    • 200 AH * 14.5 volts charging * 1/0.77 panel+charger deratings * 0.10 rate of charge = 366 Watt array nominal
    • 200 AH * 14.5 volts charging * 1/0.77 panel+charger deratings * 0.13 rate of charge = 490 Watt array "typical" cost effective maximum

    And then based on where you camp and if the panels are mounted flat to the RV roof (closer you are to the north/south poles, the more tilt the solar panels should have):

    http://www.solarelectricityhandbook.com/solar-irradiance.html

    London
    Average Solar Insolation figures

    Measured in kWh/m2/day onto a horizontal surface:
    Jan Feb Mar Apr May Jun
    0.75
     
    1.37
     
    2.31
     
    3.57
     
    4.59
     
    4.86
     
    Jul Aug Sep Oct Nov Dec
    4.82
     
    4.20
     
    2.81
     
    1.69
     
    0.92
     
    0.60
     
    More or less, 3+ hours of sun per day is usually "sufficient" to give you a decient output ffrom a solar power system. Say you camp April through September, and we use 2.81 hours of sun (flat mount panel on RV) for hours of sun per average day:
    • 366 Watt array (10% rate of charge) * 0.52 end to end AC off grid system eff * 2.81 hours of September Sun per day = 535 WH per day of 120/230 VAC electric power (enough to run some LED lighting, laptop computer, and cell phone charger).

    -Bill

    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mcgivor
    mcgivor Solar Expert Posts: 3,854 ✭✭✭✭✭✭
    jimbuck said:
    Am I right in thinking that the voltage traveling down the cables from the pv array to the controller in a pwm system will be Vbatt as it is a straight through connection in bulk mode?

    The voltage travelling down the cables from the pv to the controller in an MPPT system will be the vmp of the array at that time because the array and battery are not directly connected. The controller uses a st
    jimbuck said:
    step down transformer.


    This has been my understanding of how to calculate voltage drop in each system.
    Jim
    The voltage will be pulled down to slightly above the battery voltage in a PWM system, there has to be a potential difference for current to flow, so it's not a straight through connection, the ability of the panels to always exceed the voltage the battery requires is of paramount importance in order to achieve a positive current  flow, if the conductors are sized correctly to minimize voltage drop it is unlikely a panel rated at ~20 V  for a 12V nominal system  would have a problem supplying current, unless the ambient temperature is approaching 50°C without wind, not something of concern in the UK, but something I deal with occasionally without problems with panels rated at 36.8 Vmp on a 24V nominal system.

    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.
  • jimbuck
    jimbuck Registered Users Posts: 46 ✭✭
    Thanks

    @Photowhit , it’s something that pains me, running a 2kw generator whilst in absorption and putting 5A into my battery’s. I look forward to the solar finishing the last stage of charging in the near future!

    @Bill, I’m a full timer of 11 years so know what to expect from a British winter
    (Grey skies and rain) so I think I may need to overpanel and go to 13% plus charge rate. I have a generator here and a sterling PF corrected charger for the dark months. I’m just doing paper plans for MPPT and PWM. This system is a big compromise. If I could use 60cell panels I would just parallel 3 into an MPPT cc and save on the cable run. I can get used 260w vmp 30v panels for £115 each delivered.
    I think I’m going to have to go PWM and 1AWG cable and a mix of panel sizes on the roof. Thank you for the insolation figures, I’m about 200 miles west from London so not much difference I expect.

    @mcgivor, I’m sure that I too will be ok with the panels not getting too hot and falling below adequate battery charging voltages, it would be very rare here unfortunately! Thanks for the info regarding the voltage differential to allow current flow, I’m learning everyday from the combined knowledge of you guys!

    Jim
  • jimbuck
    jimbuck Registered Users Posts: 46 ✭✭
    edited April 2018 #22
    Hi everyone, I would like to thank all that helped me. I’ve built my system, it’s up and running with what I had at hand but will be rewired from the roof down to the controller when I can afford the propper size cable. I’ve attached a few pictures.

    Thanks again, Jim

    Still need to find a gromet for the Tristar knockouts too and a few other bits.