DC to AC then back to DC

Organic Farmer
Organic Farmer Solar Expert Posts: 128 ✭✭
in General Solar Power Topics #1
Is there a practical method to convert DC power from a Photo-Voltaic array into AC. To then transmit the power to a location 100 yards away where the battery bank is located?


I am looking at heating requirements for a battery-bank if I locate it near the Photo-Voltaic array. I would like to keep the battery-bank in my home. But the array will be at a bit of a distance.

Thank you.
· Share on Twitter

Comments

  • tons001
    tons001 Solar Expert Posts: 71 ✭✭
    #2
    Re: DC to AC then back to DC

    How many and what kind of panels? How is your array wired? Is it connected to a MPPT charge controller?
    · Share on Twitter
  • BB.
    BB. Super Moderators, Administrators Posts: 33,613 admin
    #3
    Re: DC to AC then back to DC

    And other questions.. Why AC?

    Are you just looking for the most cost efficient/low maintenance method move the electricity 300 yards to the point of use/battery storage/etc.?

    And back to my broken record--You probably are going to have to do 2-5 different paper designs and see which works best for you. I general, I would suggest keeping most of the electronics+batteries in the same place (i.e., don't build an array+charge controller+battery bank+AC inverter at the remote site just to bring AC back to the main power site--Unless you are prepared to service the remote site/need significant power there too--perhaps even a backup genset--I.e., everything keeps growing).

    So, the major options would include:

    SMA Sunny Island system. Put array+SMA GT Inverter (in Island Mode) at the remote array, connect AC to your local grid and back charge through a Sunny Island inverter+charger (ideally) or through other compatible (Schneider/Xantrex, Magnum, possibly Outback, etc.) inverter charger (SMA will vary GT output via line frequency "dithering"; others will change frequency by +/- 1 Hz to knock GT inverter off line).... Or you can add some sort of dump load controller or voltage controlled switch to turn off AC from the remote GT inverter. Probably the most expensive, but you have to admire their quality and designs.

    Similar to above, but use other brand/models of GT inverters (or micro inverters) plus compatible AC inverters. Less money, more of a "bang bang" type control loop (GT inverter is either "ON" or "OFF", no tapering charge rate as SMA SI can do).

    Get a "high voltage" MPPT Charge Controller. Schneider is the only ~600 VDC rated (Vmp~400 volts) MPPT charge controller I am aware of. Don't bother to convert to AC, just send ~400 VDC to the MPPT controller at your home/main power shed. No AC conversion needed.

    Get a "medium high voltage" MPPT Charge controller from Midnite. Maybe ~180 VDC Vmp-array input... Not quite as good as the ~400 VDC (MPPT controller) or 240 VAC (GT based), but certaily better than the ~100 Vmp-array from a "typical" high end MPPT charge controller. This controller is almost 1/2 the price of the Schneider solution (look at costs of wiring, system integration, useful options/networking hardware, etc. to compare).

    Those are a few things I can think of off the top of my head. Again, what do you have now and need to integrate with (and your familiarity with different brands/suppliers of equipment) will weight the scales too.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
    · Share on Twitter
  • Organic Farmer
    Organic Farmer Solar Expert Posts: 128 ✭✭
    #4
    Re: DC to AC then back to DC

    This array has twenty Evergreen 220w panels, so roughly 4400w. It is not currently wired. As MPPT charge-controllers seem to change each year, I am waiting to buy until the last possible moment. But now I am sitting here in -17F temps, and wondering how these temps will effect the batteries. I played with a few ideas for making the room for the charge-controller and battery-bank insulated and heated, but then I would be wasting power on heat.

    Line losses to get DC back to the house concern me too.
    · Share on Twitter
  • BB.
    BB. Super Moderators, Administrators Posts: 33,613 admin
    #5
    Re: DC to AC then back to DC

    Insulating the battery bank with removable sheets of foam (take it out if you have warm summers/room) seems to work well for many of our cold weather friends.

    If the batteries are unused during the winter, the cold is pretty good for them (as long as they are charged). And if they are cycled, then the charging/discharging seems to keep them warm enough (charging in the 80-90+% state of charge range is less efficient so you get more battery heating as a byproduct of charging).

    Say we do this with 80+ amp MPPT charge controllers with a ~100 Vmp array... That would be (I don't know what Vmp your panels are) roughly:

    4,400 watts / 100 volts = 44 amps.

    Using a generic voltage drop calculator for 44 amp array and 3% (3 volt) voltage drop at 44 amps and 300 feet one way run would give us 1/O cable (3.1 volt drop). At $2 per foot, that would be $1,200 worth of wire (300' * 2 * $2 per foot). Still need trench/conduit (not direct burial).

    If we use a 400 volt Vmp array and 11 amps @ 300 feet and 12 volt drop, that would be ~12 AWG wire (12.6 volt drop). At $0.85 for 600 feet of 12/2 direct burial wire is ~$510.

    So, the cost savings of the copper wire would be put into the more expensive high voltage MPPT controller--More or less (need prices for your location, plus cost of burial+conduit, if needed, etc.). The Mid-Range MPPT controller (~180-200 VDC Vmp range) may be your most "cost efficient point"... All depends on the exact numbers and requirements (through air, buried, etc., exact solar array, etc.).

    Is the battery shed well insulated (no wind, in ground contact, etc.) where the batteries will be able to keep their heat? If you needed to start the bank up with a a sub-zero F bank and needed to cycle it heavily--You may need to do some initial heating to get the battery in the 40F+ range (all guessing on my part--A cold day here is below 50F).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
    · Share on Twitter
  • Organic Farmer
    Organic Farmer Solar Expert Posts: 128 ✭✭
    #6
    Re: DC to AC then back to DC
    BB. wrote: »
    ... Is the battery shed well insulated (no wind, in ground contact, etc.) where the batteries will be able to keep their heat? If you needed to start the bank up with a a sub-zero F bank and needed to cycle it heavily--You may need to do some initial heating to get the battery in the 40F+ range (all guessing on my part--A cold day here is below 50F).

    -Bill

    I would prefer to do a below-grade box. But our water-table is too high. In the summer, any 6" hole you make will fill with water in a day. The normal frost-line here is at 4' below grade. Anything in the ground, above the frost-line will freeze every winter. So a battery shed must be above grade. Obviously it must be insulated, my house is insulated to R-60. We heat our home with wood and a radiant heated floor. I was thinking about putting a radiant heat floor in the battery shed too. But again the losses in getting heated water out there [and back] might be a lot.

    Many homes here use an outdoor furnace, and circulate water through buried hoses to get the heat into their homes. But my woodstove is indoors, currently none of my heated water circs outside.

    The battery shed would need to be above-grade. I can easily do a few inches of foam and 8-inches of fiberglass batting.

    Our system will be used year-round. I am not sure if the heat from daily cycling a battery-bank will be enough to keep it warm inside.

    In most cases, I do not think of electric-heat as being the most economic source of heating. Electric is good for circuits, lighting and motors; but as fuel for heat, I think of wood as being much more efficient.

    I have grid access, which is what we are using now for our house. But it is not very reliable. We lost power four times last month, a couple times so far this month. We currently have 18,000+ grid customers without power now in this region. We expect to lose it any time now again, for a few days. About 1/4 of the homes in our township have gone off-grid entirely, so they can have some form of reliable power.

    Our plan is to begin a Solar-Thermal array as our next big project after we finish Photo-Voltaic. So in theory, if I did a radiant heated floor in the battery bank. Then at that time, I would be able to tie-in both our house and the battery-bank shed, with the Solar-Thermal array.

    Sorry if I ramble too much :)
    · Share on Twitter
  • BB.
    BB. Super Moderators, Administrators Posts: 33,613 admin
    #7
    Re: DC to AC then back to DC

    Not a problem--See me when your reply has to be broken up into two to three post due to charter limits. :blush::p

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
    · Share on Twitter
  • SolInvictus
    SolInvictus Solar Expert Posts: 138 ✭
    #8
    Re: DC to AC then back to DC

    Evergreen ES-E-220-fc3 (220W) Solar Panel
    Imp 7.54A
    Vmp 29.2V
    Isc 8.22A
    Voc 35.9V
    NOCT 45.4°C
    Temp. Coefficient of Voltage -0.111V/K
    Series Fuse Rating 15A
    Maximum System Voltage 600V
    At a temperature of -30 C, Voc = 35.9 V + (-30 C - 45.4 C) * (-.111 V/C) = 35.9 V + 8.37 = 44.3 V

    MidNite Solar Classic 250 MPPT Charge Controller, $680, operates up to 250 V DC allowing at least 5 Evergreen 220 W panels to be connected in series which yields 4 parallel strings outputting a combined short circuit current of about 8.22 A * 4 = 33 A requiring a minimum of 9 gauge copper wire.

    Resistance of 8 gauge copper wire at 0 C: .5788 mOhm / foot
    Resistance of 8 gauge copper, 2 conductor, cable 300 feet long at 0 C: 2 * 300 feet * .5788 mOhm / foot = .347 ohms
    Operating voltage drop across cable in cold weather: 7.54 A * 4 * .347 ohms = 10.5 V
    Operating power loss in cable during cold weather: (7.54 A * 4)^2 * .347 = 316 W (almost 1.5 PV panels, 7.2% loss)

    8/2 type UF-B with ground wire, solid copper, might cost $1.50/foot so: 300 feet * $1.50 / foot + (2 * $680 (MidNite Solar Classic)) = $1,810

    Converting DC to high voltage AC, conducting it 300 feet and converting back to DC would probably incur a loss greater than 7.2%. 120 V AC would be a lower voltage than the 188 V DC operating voltage during low temperatures. I doubt there is a less expensive way than sending it as high voltage DC from the PV panels to a charge controller located 300 feet away.

    Edit: Because you need two MidNite Solar Classics, it might work best running two 8 gauge cables, one for each controller, which would lower the power loss. You could use two 10-2 type UF with ground cables which should cost less than $1.00 / foot and have the combined resistance of 7 gauge copper wire.
    · Share on Twitter
  • Organic Farmer
    Organic Farmer Solar Expert Posts: 128 ✭✭
    #9
    Re: DC to AC then back to DC

    SolInvictus- Thank you.

    If a 7% loss then becomes the 'limbo-bar' [at a loss for the right metaphor] number to beat.

    7% of 4400w is 308w. I can have a heating element on a thermostat to burn 200w and still have less loss.
    · Share on Twitter
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    #10
    Re: DC to AC then back to DC
    BB. wrote: »
    Get a "high voltage" MPPT Charge Controller. Schneider is the only ~600 VDC rated (Vmp~400 volts) MPPT charge controller I am aware of. Don't bother to convert to AC, just send ~400 VDC to the MPPT controller at your home/main power shed. No AC conversion needed.

    Morningstar also has a new high voltage controller. boB (from Midnite) has looked at both of them, liked them, and indicated that it was unlikely that Midnite would undertake the considerable development effort to improve on the designs.

    I like Bill's analysis:
    BB. wrote: »
    Say we do this with 80+ amp MPPT charge controllers with a ~100 Vmp array... That would be (I don't know what Vmp your panels are) roughly:

    4,400 watts / 100 volts = 44 amps.

    Using a generic voltage drop calculator for 44 amp array and 3% (3 volt) voltage drop at 44 amps and 300 feet one way run would give us 1/O cable (3.1 volt drop). At $2 per foot, that would be $1,200 worth of wire (300' * 2 * $2 per foot). Still need trench/conduit (not direct burial).

    If we use a 400 volt Vmp array and 11 amps @ 300 feet and 12 volt drop, that would be ~12 AWG wire (12.6 volt drop). At $0.85 for 600 feet of 12/2 direct burial wire is ~$510.

    So, the cost savings of the copper wire would be put into the more expensive high voltage MPPT controller--So, more or less (need prices for your location, plus cost of burial+conduit, if needed, etc.). The Mid-Range MPPT controller (~180-200 VDC Vmp range) may be your most "cost efficient point"... All depends on the exact numbers and requirements (through air, buried, etc., exact solar array, etc.).

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
    · Share on Twitter
  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    #11
    Re: DC to AC then back to DC

    Bill's option one (ac coupled) has some useful attributes, but is without a doubt the most expensive way to do it. It means two inverters, and, to get the features they need, tend on the more expensive/quality end of the spectrum.

    Whether you went that route would depend on whether you anticipate further growth, with other arrays or other power sources distributed over a wider area. The beauty of ac coupling is that you effectively have your own grid, and can tap into it anywhere, either as production or consumption. An example would be a farm complex with a number of suitable roofs.
    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


    · Share on Twitter
  • SolInvictus
    SolInvictus Solar Expert Posts: 138 ✭
    #12
    Re: DC to AC then back to DC

    Redoing the calculation using a Conext 80 Amp MPPT 600 VDC Charge Controller, $1,123,, you could connect 10 PV panels in series and have 2 series strings. I assume the battery array is 48 V.

    At a temperature of -30 C, Voc = 10 * 44.3 V = 443 V.
    At a temperature of -30 C, Vmp = 10 * (29.2 V + (-30 C - 45.4 C) * (-.111 V/C)) = 10 ( 29.2 V + 8.37) = 376 V.
    Both are within the input voltage range, 195 VDC to 600 VDC (open circuit), and the PV array start voltage, 230 VDC, of the charge controller.
    Your Vmp equals the start voltage when the temperature of the PV panels is 101 C which seems adequate.

    With both strings in series, the short circuit current is about 8.22 A * 2 = 16.44 A, and Imp = 7.54 A * 2 = 15.08 A.

    In these cold conditions you get 376 V * 15.08 A = 5,670 W with 100% efficiency. At 48 V you get 118 A, and at 57.6 V you get 98 A which both exceed the 80 A rating of the charge controller. If the efficiency of the wiring and charge controller is 85%, then you get 5,670 W * .85 / 48 V = 100 A and 84 A respectively. This charge controller probably will not be able to use all the power output by the PV panels on cold days. There is up to ~1,000 W of wasted power that you could use to heat the batteries if they are located near the PV array.

    The short circuit current of 16.44 A indicates 13 gauge copper wire is the minimum size for connecting the PV array to the charge controller.

    Resistance of 10 gauge copper wire at 0 C: .9203 mOhm / foot
    Resistance of 10 gauge copper, 2 conductor, cable 300 feet long at 0 C: 2 * 300 feet * .9203 mOhm / foot = .552 ohms
    Operating voltage drop across cable in cold weather: 15.08 * .552 ohms = 8.32 V
    Operating power loss in cable during cold weather: (15.08 A)^2 * .552 = 126 W (2.9% loss)

    If 10 gauge type UF-B cable costs $.88 / foot, then the cost of this configuration is $1,123 + ($.88 / foot * 300 feet) = $1,387 which is a better price than using the Midnight Classic 250 because you can use smaller wire and can make it work with one charge controller.

    As for locating the battery array, charge controller and inverter near the PV array and sending 120 VAC a distance of 300 feet to your house, you will have to use large wire to minimize the voltage drop and power loss in the wire. Assuming this system will be equipped with a 4,800 W inverter, you will need to use 4 or smaller gauge copper wire to deal with the surge currents of about 80 A. NAWS has 4 gauge copper wire listed at $1.68 / foot for which you would need conduit. 2 * 300 feet * $1.68 / foot = $1,008 not including the cost of conduit and constructing a shed. The Conext 80 A MPPT looks like your best option.
    · Share on Twitter
  • offgrid me
    offgrid me Solar Expert Posts: 119 ✭✭
    #13
    Re: DC to AC then back to DC

    Adding to what Solinvictus said you would also have additional savings in mc4 extension cables, not needing a combiner box and not needing to fuse the strings of panels because you will only have two. I ran the math on my system and found that the 80 600 saved me a few hundred in BOS componants. It usually runs at over 92% efficiency and never has gotten hot.
    Ned
    · Share on Twitter
  • Organic Farmer
    Organic Farmer Solar Expert Posts: 128 ✭✭
    #14
    Re: DC to AC then back to DC

    Thank you. everyone.
    · Share on Twitter
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    #15
    Re: DC to AC then back to DC
    offgrid me wrote: »
    I ran the math on my system and found that the 80 600 saved me a few hundred in BOS componants.

    If it only cost a few hundred more to go with Classic 250 Controller and thicker copper cable, I would go for the Classic and the thicker cable. That thicker cable will outlast any controller. (and you may need to replace your controller someday).

    As 'offgrid me' points out, the high voltage strings may save you the cost of a combiner box, but I find a combiner box is desirable even with only two strings. The combiner box is a good place to put a lightning arrester and circuit breakers on each string are helpful for troubleshooting and allow you to de-energize the long cable run.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
    · Share on Twitter

Categories