Considering downlead losses

65DegN
65DegN Solar Expert Posts: 109 ✭✭
Been thinking about this lately as I design a 3,750 watt array into a system that could require 500' (RT) of array downlead. I've come to the conclusion that there is going to be a curve associated with the inefficiencies of this system balanced between the number of panels and the percentage of losses in the wire. Generally speaking I think it is better to not dump a massive amount of money into heavy wire (in this case it would take 500' of 1/0 to provide for ~2.4% loss or 82 watts) Rather than invest ~$1,000 into downlead I would rather put more of that capital into more panels and maybe add a tad more, to make up for losses. With this arrangement the array will also have better cloudy day performance (particularly if the controller is maxed out on a sunny day) at lower current levels, and consequently less loss than on sunny days, and greater output due to increased surface area.

If #2 copper were used instead, the losses ( at the array's full rated output) would be about 3.5% for a total of 131 watts or an additional loss of 49 watts over using 1/0 wire and the savings in the cost of wire would be about 30% or $250.

With panels at ~$1 a watt and less it makes sense to go with another 250 watt panel, where technically and logistically feasible, as the system would gain roughly an additional 200 watts in full sun with 3.5% downlead losses. This could become more substantial as the system size increases. One variable that I can not absolutely quantify right now is exactly how this increase in downlead resistance would effect controller performance but it seems to me it should remain linear.
Granted there are many considerations including racking costs etc. But just something to think about.
As panel prices continue to remain low and even drop some, more panels are the answer to a lot of system design and system cost problems.

Comments

  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: Considering downlead losses

    Smart analysis. I am very concerned that my electric consumption be as efficient as possible, but as production (solar panels) becomes cheaper and cheaper I have become less concerned with the efficiency of production.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Considering downlead losses

    3,750 Watt array at what Voltage? Up the Voltage, lower the current, reduce the losses.
    Up to the point where the array Voltage (especially Voc with cold temps) becomes too high for the controller and/or the down-conversion from a high Voltage reduces controller efficiency.

    Without the Voltages involved the other numbers don't mean much.
  • 65DegN
    65DegN Solar Expert Posts: 109 ✭✭
    Re: Considering downlead losses

    Caribou, array voltage is at ~112 VDC OC. Three poly REC 250 watt panels in series, five strings in parallel. So a few more panels would max out a Midnite 150 controller. Same situation I came up with as far as controller choice. I found I'm better off with more panels than using a 250V controller since it won't handle as much power out and is less effecient.

    Seems the new Tristar 600V controller doesn't have adjustable current limiting and doesn't have any function to use excess solar power so I don't like that either. Worse, Morningstar hasn't repiled to any of my queries about the unit.

    In short I believe the Midnite 150 is the most efficient way to go and instead of going to a 200 or 250 (except for small systems) I'm better off adding panels where applicable.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Considering downlead losses

    Yep; that works.
    And is better than paying $1,200 for the XW 600 Volt controller. :p You can buy a lot of panel for $600.

    Of course the losses drop as the current demand goes down too, so they won't always bee there at the max all day every day.
  • mmag
    mmag Solar Expert Posts: 57 ✭✭✭✭
    Re: Considering downlead losses

    yeah but how much money in wire can you save by runnng a xw 600 volt controller?
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Considering downlead losses
    mmag wrote: »
    yeah but how much money in wire can you save by runnng a xw 600 volt controller?

    Probably none.
    The point here is you will have losses and how you make up for it: larger wire, higher Voltage charge controller, or add some panel so that there's more power to begin with and the loss at the other end no longer matters. With copper prices high, the XW 600 2X a standard controller, and panels being in the $1 per Watt range the additional panels may be the cheapest option.
  • mmag
    mmag Solar Expert Posts: 57 ✭✭✭✭
    Re: Considering downlead losses
    Probably none.
    The point here is you will have losses and how you make up for it: larger wire, higher Voltage charge controller, or add some panel so that there's more power to begin with and the loss at the other end no longer matters. With copper prices high, the XW 600 2X a standard controller, and panels being in the $1 per Watt range the additional panels may be the cheapest option.
    Isn't it smaller wire and higher voltage charge controller?
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Considering downlead losses

    Yes, as you wrote it... Marc (Cariboocoot) was saying "or"... Heavier wire or higher voltage controller.

    You can save a bunch of money on wiring with higher voltages for long runs.... But if you have nominal runs and are already using a MPPT controller that can take Vmp~100 VDC STC, then you are not too bad off (in general) when compared to somebody using a PWM controller on a 12 volt battery bank.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Considering downlead losses
    mmag wrote: »
    Isn't it smaller wire and higher voltage charge controller?

    It's a $1,200 charge controller originally designed to accommodate GT arrays for people who wanted to use them during power outages as a means of recharging 24 or 48 Volt back-up battery systems.

    Whereas it is true that for any given amount of power (Watts) the current goes down as the Voltage goes up this only relieves one type of loss; the Voltage drop in the line between the array and the controller. It accentuates another type; the conversion efficiency of the controller from high array Voltage to lower system Voltage.

    So where do you want to lose the power, how much do you want to lose, and for what price?

    It is never as simple as "this solution is the best one for everyone".
  • mmag
    mmag Solar Expert Posts: 57 ✭✭✭✭
    Re: Considering downlead losses

    my point was 500 feet is alot of copper, two strands of 2 awg is around $1200 if you go with a 600v charge controller you could spend less than half on smaller wire with less line loss.
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: Considering downlead losses
    mmag wrote: »
    my point was 500 feet is alot of copper, two strands of 2 awg is around $1200 if you go with a 600v charge controller you could spend less than half on smaller wire with less line loss.

    But don't forget... it's 500 ft round trip... really a 250 ft span for the 2 conductors. (and maybe a ground conductor?)

    Spending a few hundred dollars on copper and an extra solar panel is probably a better investment than spending a few hundred on electronics.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • mmag
    mmag Solar Expert Posts: 57 ✭✭✭✭
    Re: Considering downlead losses

    Sorry I thought it was 500 feet away
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Considering downlead losses
    mmag wrote: »
    Sorry I thought it was 500 feet away

    So did I. :confused:
    If it's 250 feet away that's a big difference.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Considering downlead losses

    More or less, the basic wiring differences between array voltages:

    2x higher voltage and 1/2 the current:
    • 1/2 less current * 1/2 higher voltage drop allowed at higher voltage = 1/4 the thickness of wire
    • Every 3 American wire gauge difference is 1/2 the surface area of copper: 1/4 thickness is 6 AWG smaller wire
    For 4x higher voltage and 1/4 current:
    • 1/4 current * 1/4 higher voltage drop allowed = 1/16 the thickness of wire
    • 1/16 thickness of wire is 12 AWG smaller wire
    3,750 Watt system... 250 foot run. Vmp-array = 90 volts (3x 30 Vmp) or 390 volts (13x 30 Vmp panels). 3% voltage drop maximum:
    • 3,750 Watts / 90 volts = 41.7 Amps
    • 3,750 Watts / 390 volts = 9.62 Amps
    • 90 volts * 0.03 drop = 2.7 volt drop
    • 390 volts * 0.03 drop = 11.7 volt drop
    Using a generic voltage drop calculator for 250 Foot round trip:
    • 41.7 amps, 250 foot round trip, 2.7 volt drop max = 3.1 volt drop @ 1 AWG cable
    • 9.62 amps, 250 foot round trip, 11.7 volt drop max = 9.2 volt drop @ 12 AWG cable
    So--Very significant reduction in cost of copper... The cost to trench and bury PVC Conduit (your choice--I like to bury several large diameter conduit at the same time--Allows me to run other cables later [more AC cable, communications/networking/etc.] if ever needed without having to re-trench).

    Which is better for you.. Don't know. Either would work fine

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • 65DegN
    65DegN Solar Expert Posts: 109 ✭✭
    Re: Considering downlead losses

    Using the Xantrex 80-600 might be feasible if the system were designed specifically for it. Considering cold weather effects it drops the upper VOC limit to ~500 V or 13 panels. Cutting 500 watts off a 3,750 watt array.
  • SkiDoo55
    SkiDoo55 Solar Expert Posts: 414 ✭✭✭
    Re: Considering downlead losses

    Max VOC is 600 VDC including temp compensation. Operating range is 195 to 550 VDC. MPPT range is 195 to 510 VDC. Shows start voltage at 230 VDC.
    Schneider site is currently down for maintenance as I was going to check the numbers on the array to see what was panel configurations with temp and your REC panels. Assuming you get -40 C or colder.
    GT3.8 w/4600W Trina 230W, TX5000 w/5000W ET-250W, XW4024 w/1500W ET-250W, 4 L16, 5500W Gen. (never had to use) Yet!!
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: Considering downlead losses
    65DegN wrote: »
    Using the Xantrex 80-600 might be feasible if the system were designed specifically for it.

    One design change might be overcurrent protection... no 150 volt breakers .... 600 volt fuses AND disconnects. I guess that with a single string you won't need a combiner, but it would be nice to have a disconnect at or near the array to de-energize the line to the controller.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: Considering downlead losses
    65DegN wrote: »
    Three poly REC 250 watt panels in series, five strings in parallel. So a few more panels would max out a Midnite 150 controller.

    Another thought.... how about aiming part of the array SE and part of the array SW? This is often called virtual tracking. Because the two parts of the array are not giving maximum current at the same time, you don't have as much line loss on the 250 ft run. You can also 'over panel' your controller a bit more.

    The panels in each string of a virtual tracker should have the same exposure. You may want some asymmetry... two strings facing east and one string facing west... depends on your climate and exposure and loads. e.g. southerners with air conditioning might want to have more strings facing west for the late afternoon load.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • 65DegN
    65DegN Solar Expert Posts: 109 ✭✭
    Re: Considering downlead losses

    vtmaps,Customer won't do that. The ice can build on roofs where he is to a couple feet thick and he is worried about it taking the array off the roof. So he only wants a top of pole mount.

    SkiDoo, 550 V in to the controller would likely exceed input voltage on cold days. I figure about to 638 V.

    Using a high voltage controller won't improve cloudy day performance like additional panels will.
    The Midnite controller will handle 96 A @ 48V. Thats about another 500 watts over the Schneider at about half the price which pays for the additional 500 watts.
    Granted wire costs are reduced with HV which would also pay for the panels but then another HV controller is needed and another run of wire.
    Also a big consideration is the fact that the Midnite has a superior 'use it or lose it' function allowing for excess solar power to be metered off to perform other functions like heating water. Looks like the Schneider unit has simple relay closure based on battery voltage while the Midnite uses PWM output.
  • SkiDoo55
    SkiDoo55 Solar Expert Posts: 414 ✭✭✭
    Re: Considering downlead losses

    "SkiDoo, 550 V in to the controller would likely exceed input voltage on cold days. I figure about to 638 V."

    that is why I would recommend 2 srings of 9 series that would add panels keep cold VOC reasonable and within range. All are suggestions to a problem presented. XW MPPT 80 600 from sponsor with a ComBox would be about $1500 plus shipping.
    GT3.8 w/4600W Trina 230W, TX5000 w/5000W ET-250W, XW4024 w/1500W ET-250W, 4 L16, 5500W Gen. (never had to use) Yet!!
  • 65DegN
    65DegN Solar Expert Posts: 109 ✭✭
    Re: Considering downlead losses

    SkiDoo, thats probably a workable solution. I was concerned about the lower end of the operating range but this seems to reasonably quite well at ~21.6 volts per panel(loaded ) to hit op range and a 30.5 V-MPP for these panels.
    If the people at Morningstar would communicate, their HV offering might have merit with a boost function starting from 5 volts.