NMDU service entrance cable sizing

romac55

Just a question about what size of NMDU cable to use as my service entrance cable ; my solar array is 5000 watts at 24 volts dc . All the wiring catalogues show 600 volt or 1000 volt cables in a variety of sizes for AC service entrance , but nowhere do I find any DC wire sizing .  My 250 watt panels are wired parallel to buss bars and a DC disconnect , but I'm faced with the same question with the DC disconnect . What size do I use ?  It's rated at 100 amps , but my array will be outputting over 200 amps of 24 VDC .  Am I making this too difficult ?  Just want to keep the electrical inspector happy at inspection time .  newbie romac55

BB.

More or less, current is current. 200 amps of 240 vac vs 24 vdc is 200 amps for wire sizing.

DC vs AC current has other fundimental differences (switch ratings, arc fault issues, safety, etc.) that need to be researched/designed for.

But I would like to understand why 200 amps at 24vdc solar?

We generally try to use higher voltage solar array's (100 to 400 vdc Imp-array).

24 vdc is not my first choice.

-Bill

romac55

I've been thinking and re-thinking this whole amperage thing , so will most likely bump my pv system voltage to 48 from the pv array , and save tons of amperage .  Only down-side will be buying 8 big 6-volt batteries instead of 4 ; the inverters are about the same price whether 24 or 48 vdc .  SO , the dc disconnect will be okay at 100 amps ; just need to know what size NMDU cable (50 ft long) to use for my service entrance , as the inverter will be inside the house . Pretty steep learning curve .  romac55

romac55

BB. said:

More or less, current is current. 200 amps of 240 vac vs 24 vdc is 200 amps for wire sizing.

DC vs AC current has other fundimental differences (switch ratings, arc fault issues, safety, etc.) that need to be researched/designed for.

But I would like to understand why 200 amps at 24vdc solar?

We generally try to use higher voltage solar array's (100 to 400 vdc Imp-array).

24 vdc is not my first choice.

-Bill

24vdc wasn't my first choice either , but I inherited a few old 24dc batteries from my years in the forklift biz , and the pv panels charge the batteries as well as supply the inverter with 24vdc power .  Now the latest system is an off-grid cottage were building far from the grid , so again the storage batteries will need charging ; the cottage system will see the pv array charging a 48 volt battery bank and supplying 220 vac to the main panel in the cottage . Thanks for the Christmas eve reply , Bill .   romac55

BB.

I suggest that you start from the beginning documenting your design... For example, to charge a 24 volt lead acid battery bank you need around Vmp-array of 35-40 VDC minimum (or optimum for a PWM controller). If you have larger solar panels, they may be Vmp~30 volts--Which is not high enough voltage (when all conditions are taken into account) to properly charge a 24 volt battery bank.

-Bill

romac55

Woah boy ; my Magnum inverter says input batt voltage 18.0-34.0 on my 24 volt system , and it's been charging batteries well for 8 years . It uses 12X 80 watt pv panels wired parallel . I think I'm okay in the battery department ; my question was about NMDU service entrance wire .  Maybe I'll just stay with my idea of changing up the system voltage to 48vdc to lower the amperage in the new system , and forget I asked the question about the NMDU .  I KNOW how to "properly charge a 24 volt battery bank" .  I'm an off-grid guy anyways , so will ask questions in off-grid forums . Thanks .

BB.

Sorry--there just are not many "24 volt" (Vmp~36 volt) solar panels these days (Evergreen is gone, only a few 72 cell panel mfg. out there at the moment--But that does change over time as mfg. make larger diameter cells/solar panels and update their ever changing product line). That was my concern.

If you go with a MPPT type charge controller vs PWM, you can run a much higher Vmp-array voltage and save on copper wiring (smaller diameter wire) from the array to the charge controller/battery installation (plus, you don't need near as many fuses either--1 fuse/breaker per solar panel string if you have 3 or more panel strings in parallel).

Need to know the Vmp panels, Vmp-array, and the type (brand/model) of solar charge controller, as well as the Imp-array (number of solar panels), battery bank voltage, and cable run length to figure out the allowable voltage drop for your array to charge controller. For a PWM controller, you can usually only have ~3% maximum voltage drop (Vmp~36 volts, 1.08 volt drop recommended maximum). For a Vmp-array of 90 volts (on a MPPT charge controller), you can have 2.7 volt drop (and almost 1/3rd the current) and use a much smaller gauge of wire.

Many times, in solar power systems, designing for low voltage drop (1-3% typically), we end up with larger awg cable than just the current requirement would require on longer cable runs. Especially for low voltage DC cable runs.

-Bill

Vic

Hello romac..,

Before you go too much farther,  you might want to think more about the overall system design.

First,  with 5 KW of PV,  you really do NOT want to wire all of the PVs in parallel,  and as BB Bill noted,  many of the cost common PV modules are NOT 24 V  --  they have nominal Vmps in the range of about 29 - 31 V.   This is many volts TOO LOW to allow full charging of Flooded batteries (and many other types as well).

You will probably want to use MPPT Charge Controllers (CCs),  with an array of this size,  and if you will be running more than one CC,  you will NOT want the entire PV array to enter your power facility (room,  or whatever)  as a single array.

Furthermore,  running each PV in parallel  will mean that your COMBINER  will need one circuit breaker (or fuse)  for EVERY PV module.

48 Volts for the battery will reduce the number of CCs you will need,  and reduce the size of connecting cables that are needed.

You will really want to choose the PVs,  based on the overall system design,  if at all possible.   And,  one usually sizes  the entire system around the total load the system  needs to supply,  the peak loads,  and when these loads occur during the day,  and so on.

Opinions,   Good Luck,   Vic

romac55

Thanks to BB and Vic for your responses ; I've re-designed the cottage PV system to 48Vdc and now have 3 parallel strings of 4 X 250 watt panels in series . Now the system amperage is away down , and system voltage to the Morningstar charge controller is 120Vdc ; I can now use 2awg wire as Service entrance cable to the Magnum 4448PAE inverter inside the building . Also installing a 100 amp DC disconnect at the array . Thanks again for your helpful suggestions , and have a great year in 2016 .  Bob  (romac55)

BB.

Hi Bob,

Where is the cabin system installed? What is the Vmp/Imp and Voc/Isc ratings of your panels? And what MorningStar charge controller are you using?

The typical 250 Watt panel is a Vmp~30 volts and Voc~38 volts... Unless the system is installed on a relatively warm/tropical island, the Voc-array-cold is usually too high for the "typical" MPPT type charge controller (usually maximum of 140-150 VDC maximum).

For MorningStar, here is their string calculator:

http://string-calculator.morningstarcorp.com/

MorningStar does a ~600 VDC max input MPPT charge controller--But that is a very expensive unit vs the standard 150 VDC max input.

http://www.solar-electric.com/inverters-controllers-accessories/chco/mochco/admpchco.html

My guess is you should (quickly) rewire the array to 3 panels in series by 4 parallel stings. You run the risk of damaging the MPPT controller (and many log the maximum input voltage--And this will violate your warranty).

-Bill

romac55

Right as rain , Bill ; I misspoke ; 4 parallel strings of 3 X 250 panels will feed a TS60 pwm controller at about 90 Vdc . The TS60 recommends a Voc of 125 and I think we're right in that ballpark .  I can only wish for the warm , tropical island ; the location is in Northern Ontario where they don't even speak the "warm , tropical" language , or much English for that matter .  According to their string calculator , because of the extreme winter cold temps , I should be using 6 parallel strings of 2 X 250 watt panels to keep the TS60 happy ; The last thing you need on a cold winter morning is to watch the CC box fill with smoke , and we all know once the smoke gets out , it's hard to put it back in .   Bob  (romac55)

Vic

romac55 said:

Right as rain , Bill ; I misspoke ; 4 parallel strings of 3 X 250 panels will feed a TS60 pwm controller at about 90 Vdc . The TS60 recommends a Voc of 125 and I think we're right in that ballpark .  I can only wish for the warm , tropical island ; the location is in Northern Ontario where they don't even speak the "warm , tropical" language , or much English for that matter .  According to their string calculator , because of the extreme winter cold temps , I should be using 6 parallel strings of 2 X 250 watt panels to keep the TS60 happy ; The last thing you need on a cold winter morning is to watch the CC box fill with smoke , and we all know once the smoke gets out , it's hard to put it back in .   Bob  (romac55)

 

Hi Bob,

OK,  but it seems that you are proposing to run both or either strings of three,  or strings of two 30-ish volt PVs

A PWM CC is not very efficient dealing with a 90 volt input,   when trying to charge a 48 V battery.  AND,  strings of two PVs with about 60 volts into the TS-60  will not have enough voltage to fully charge  cool/cold 48 V Flooded batteries,  or to EQ them on many occasions.

An MPPT CC with strings of three PVs should probably still work OK in your Winter climate.

If you are planning on enlarging your system,  at some point,  would still try to run a pair of PV + and PV -- cables to your power room,  unless the distance between the PVs and the power room is short,  and your use a large conduit,  with a pull rope inside for possible future expansion.

IMO,  Vic

romac55

Thanks again , Vic . I'll probably run the 4 strings of 3 X 250 panels like BB suggested ; the Morningstar calculator may think the 6 X 2 configuration is "optimum" , but I have to agree with you , and will leave a large underground conduit for future expansion .    Bob (romac55)

Vic

Hi Bob,

OK,    BUT    a PWM CC will,  essentially turn  PV voltage at it input that exceeds the battery voltage into self-heating.

You mentioned that your 250 watt PVs in strings of three have a nominal string voltage of about 90-ish volts  --  this is quite typical for many/most 250-ish watt PVs.

Also,  typical,  is,  that with a 48 volt Flooded or AGM battery,  and an array of appreciable size,  that an MPPT CC is really REQUIRED,  as strings of two PVs will have insufficiently high input voltage and strings of three will have too high an input to the CC.  Strings of two PVs  will be able to do part of the Bulk charge stage,  depending upon the battery temperature,  and the PV's temperature.

Strings of three PVs,  with a sting voltage of about 90-ish volts is a very good compromise,  and seldom will the PV voltage input to the CC be excessive,  even in cold climates,  and will have sufficiently high string voltage to the CC even in the hottest PV temperatures.

72-cell PVs often have just (barely) enough PV voltage on hot days,  and will be fine on cold days,  allowing the use of a PWM CC,  efficiently.   Modern 72-cell PVs typically have a Vmp of about 36 - 36.5 V at STC.

FWIW,   Vic