Array wiring, conduit, and grounding questions

[sunnyguy]

My layout has 4 arrays.

1 - 15 panels (String A = 9 panels, String B1 = 6 panels)
2 - 3 panels (String B2 = 3 panels)
3 - 8 panels (String C)
4 - 8 panels (String D)

What is the best way to design the wiring, taking into account grounding and conduit issues?

If I want to skip the combiner box and just wire all 4 strings to the inverter, are there advantages to running the wiring in separate or shared conduits?

The physical layout lends itself to combining wiring to a single conduit in the following way:

Array 1 & Array 3 conduits combine at Array 2. Single conduit from Array 2 to Array 4. Single conduit from Array 4 to inverter.

[sensij]

You are going to need to share more info for help. What are the electrical specs of your panels? What inverter (or charge controller) are you using?

[solarfrank]

looks like you can run all 4 string with no 10 and also ground no 10 or 8 depends of your inverter specs in a 3/4" conduit from the roof , but we need more details for more accurate information.

[SunEagle]

Unless your inverter already has a built in "combiner box" I don't believe you can run 4 different "strings" to it without installing some type of over-current device for each string.

As the others have asked, what are your panel and inverter specs?

[solarix]

What inverter are you using?
Most inverters work with strings of between 8 and 13 panels.
You'll probably be dividing up your arrays to meet the inverter's string requirements instead of trying to connect a bunch of odd sized strings to the inverter.

[sunnyguy]

Sorry for the omission of details.

Panels:
Pmax = 295W
Imp = 8.13A
Isc = 8.62A
Voc = 44.7V

Inverter
Fronius Primo 8.2 Transformerless Dual MPPT
2 sets of DC input terminals per MPPT (4 total)
Max DC input 18A per MPPT
Ground Fault Protection : Yes

String B1 & B2 are in series, resulting in 2 strings of 8 and 2 strings of 9.

Wondering about EGC sizing, combining array EGC in junction boxes, ability to use standard NEMA 4 junction boxes since there is no string combining, and other relevant 2011 NEC requirements.

Adjustment factors (NEC 310.15):

Conduit 0.5-3.5in above roof temperature adder: 22C
Number of conductors ampacity derating (4-6): 0.8
Number of conductors ampacity derating (7-9): 0.7
Temperature derating factor of 90C conductor @ 65C (43C+22C): 0.65

Non-Adjusted Ampacities @ 30C:
12AWG 90C copper: 30A
10AWG 90C copper: 40A
8AWG 90C copper: 55A
6AWG 90C copper: 75A
4AWG 90C copper: 95A

Adjusted Ampacities @ 65C:
12AWG 90C copper (1-3 conductors): 19.5A
12AWG 90C copper (4-6 conductors): 15.6A
12AWG 90C copper (7-9 conductors): 13.6A

10AWG 90C copper (1-3 conductors): 26.0A
10AWG 90C copper (4-6 conductors): 20.8A
10AWG 90C copper (7-9 conductors): 18.2A

8AWG 90C copper (1-3 conductors): 35.7A
8AWG 90C copper (4-6 conductors): 28.6A
8AWG 90C copper (7-9 conductors): 25.0A

6AWG 90C copper (4-6 conductors): 39.0A
6AWG 90C copper (7-9 conductors): 34.1A

4AWG 90C copper (4-6 conductors): 49.4A
4AWG 90C copper (7-9 conductors): 43.2A



Max EGC Current = Isc * 4 * 1.25 = 43.1A ?
Max String Current = Isc * 1.25 = 10.8A ?

So how does this apply to the conductors and different segments of conduit?

For example, the only conduit carrying 4 pairs of DC wires would be from array 4 to the inverter and would require a 4AWG EGC. However, the inverter only accepts up to 6AWG, so it seems 2 conduits are required.

So I guess I can go with 2 conduits, one for strings A&B and one for strings C&D.

1/2" EMT 37% fill
DC Conductors - 4 x 12AWG (upsized for <1% voltage drop)
EGC - 1 x 8AWG

However in the conduit segments that carry wiring from only 1 string can I use a 12AWG EGC?

Since I am not combining strings I can use standard NEMA 4 JBs and not expensive "listed for purpose" solar boxes, right?

[sunnyguy]

So am I on the right track or do I have it wrong and need OCPDs?

[sunnyguy]

Because 2 strings are paralleled to each separate MPPT input then I am unsure whether it is necessary to consider them as possible backfeed sources for one another.

If they are then a fault could result in up to 3 strings backfeeding current through a DC conductor meaning OCPD devices are necessary because the current exceeds the fuse rating of the module and to avoid conductor upsizing.

If they are not then a fault could result in only 1 string backfeeding current through a DC conductor meaning OCPD devices are not necessary because conductor size can accommodate for this and it does not exceed the fuse rating of the module.

[solarix]

Each MPPT input is totally separate and you will not have a problem with backfeeding between the strings.

[solarix]

Each MPPT input is totally separate and you will not have a problem with backfeeding between the strings.

[SunEagle]

Each MPPT input is totally separate and you will not have a problem with backfeeding between the strings.

So the way the inverter is set up it looks like you can put 2 strings in parallel to one MPPT and 2 strings to the second MPPT. So technically you are not exceeding the max of 2 panels/strings to the inverter and may not need OCPDs.

[sunnyguy]

Thanks. So I'll plan on doing the 2 runs of conduit 1/2" EMT, butt splices from USE-2 to THWN-2 in NEMA 4 junction boxes, EMT grounding lugs.

Do my wire size calculations look right? 12AWG DC conductors and 8AWG EGC? What type of connectors are recommended for combining/transitioning ground wires in the junction boxes?

Any suggestions on conduit supports on tile roofs? These: http://www.ecofastensolar.com/store/...p?idproduct=27 ?

Thanks

[sensij]

Your calcs aren't quite right.

Required ampacity of PV Source or Output circuits (series strings before or after combining) is 1.25 * 1.25 * Isc (see 690.8(A)(1) and 690.8(B)(1)(a)).
Source circuit 8.62 * 1.56 = 13.44 A. With 0.65 temp derate, 90 deg 14 AWG is good for 25 A * 0.65 = 16.25 A.
Output circuit (after combining) = 2 * 8.62 * 1.56 = 26.88 A. With a 0.65 temp derate, 90 deg 8 awg is good for 55 A * 0.65 = 35.75 A

Typically, the source circuit is not in conduit, run with PV wire.
Typically, after combining, put all 4 current carrying conductors together in conduit. That corrects ampacity by 0.8.

8 AWG good for 35.75 (from above) * 0.8 = 28.6 A, still ok.

Also, if you want to run the EGC bare and exposed, it can be no smaller than 6 awg (see 690.46)

Since your inverter has ground fault protection, EGC size (if in conduit) between panels and inverter comes from Table 250.122 (see 690.45(B)). Since no OCPD is in PV circuits, assume Isc is max overcurrent (output circuit Isc = 2 * 8.62 = 17.24 A), which means 12 AWG for EGC would be ok.

[sunnyguy]

Your calcs aren't quite right.

Required ampacity of PV Source or Output circuits (series strings before or after combining) is 1.25 * 1.25 * Isc (see 690.8(A)(1) and 690.8(B)(1)(a)).
Source circuit 8.62 * 1.56 = 13.44 A. With 0.65 temp derate, 90 deg 14 AWG is good for 25 A * 0.65 = 16.25 A.
Output circuit (after combining) = 2 * 8.62 * 1.56 = 26.88 A. With a 0.65 temp derate, 90 deg 8 awg is good for 55 A * 0.65 = 35.75 A

Typically, the source circuit is not in conduit, run with PV wire.
Typically, after combining, put all 4 current carrying conductors together in conduit. That corrects ampacity by 0.8.

8 AWG good for 35.75 (from above) * 0.8 = 28.6 A, still ok.

Also, if you want to run the EGC bare and exposed, it can be no smaller than 6 awg (see 690.46)

Since your inverter has ground fault protection, EGC size (if in conduit) between panels and inverter comes from Table 250.122 (see 690.45(B)). Since no OCPD is in PV circuits, assume Isc is max overcurrent (output circuit Isc = 2 * 8.62 = 17.24 A), which means 12 AWG for EGC would be ok.

Thank you very much for your answers.

So when 690.8B(2) says:

(2) Conductor Ampacity. Circuit conductors shall be sized to carry not less than the larger of 690.8(B)(2)(a) or (2)(b).
(a) One hundred and twenty-five percent of the maximum currents calculated in 690.8(A) without any additional correction factors for conditions of use.
(b) The maximum currents calculated in 690.8(A) after conditions of use have been applied.

conditions of use does not mean conductor derating per 350.15? So the conductor must meet both 1.56 multiplier and derating instead of 1.25 multiplier and derating?

I was trying to avoid using combiner boxes on the roof to avoid buying expensive boxes that comply with 690.4(D). Instead I want to use standard NEMA 4 junction boxes to transition but not combine. If I do this I can use the 12AWG since the circuits are not combined until they reach the inverter?

The AWG6 exposed ground to the racking would also transition to THWN-2 in the JB. EGC does not have to comply with the 1.56 multipliers and temperature or conduit fill derating? If not then maybe I can use a single conduit and combine the grounds to a single 10AWG ground.

Thanks

[sensij]

Thank you very much for your answers.

So when 690.8B(2) says:

(2) Conductor Ampacity. Circuit conductors shall be sized to carry not less than the larger of 690.8(B)(2)(a) or (2)(b).
(a) One hundred and twenty-five percent of the maximum currents calculated in 690.8(A) without any additional correction factors for conditions of use.
(b) The maximum currents calculated in 690.8(A) after conditions of use have been applied.

conditions of use does not mean conductor derating per 350.15? So the conductor must meet both 1.56 multiplier and derating instead of 1.25 multiplier and derating?

Yeah... I think you are right, the extra 1.25 and the conditions of use corrections do not get stacked.

OK, I think we are agreed that 14 AWG PV Wire is OK for the open air runs from the panels to the pass-though into conduit, and you would prefer 12 AWG to limit the voltage drop. Your question is whether it is better to combine and transition to THWN-2, running 4 conductors (+ ground) in the home run, or whether you should not combine and run 8 conductors (+ground).

Let's look at the conduit fill:
1/2" EMT can be used with up to 9 x 12 AWG THWN-2 conductors and meet the 40% fill requirement. No difference there.

Now look at conductor ampacity.
For the 4 conductor (combined) version:
Isc of each string = 8.62 A
690.8(A)(2) = 2*8.62*1.25 = 21.55 A
690.8(B)(2)(a) = 1.25 * 21.55 A = 26.9 A. Uncorrected, you are limited by the 75 deg rating of the terminals (assuming the terminals are not 60 deg), which requires 14 AWG. OK
690.8(B)(2)(b) = Carry 21.55 A after conditions of use applied. 10 AWG @ 90 deg = 55 A * (0.65 * 0.8) = 28.6 A. Terminals at 75 deg = 50 A. OK. Maximum of 6 x 10 AWG THWN-2 in 1/2" EMT. OK.

So, for this version, 10 AWG is required.

For the 8 conductor (uncombined) version:
Isc = 8.62 A
690.8(A)(2) = 8.62*1.25 = 10.8 A
690.8(B)(2)(a) = 1.25 * 10.8 = 13.5 A. Uncorrected 14 AWG OK.
690.8(B)(2)(b) = Carry 10.8 after conditions of use applied. Conduit fill correction is 0.7. 14 AWG @ 90 deg = 25 A * (0.65 *0.7) = 13.6 A. Terminal at 75 deg = 20 A. OK.

So, for this version, 14 AWG required, although as you said, 12 AWG would be better for voltage drop.

Based on this, it looks like your choices are 8 x 12 AWG without a combiner, or 4 x 10 AWG with a combiner. In either case, you should be fine with 1/2" EMT, if you can pull the wires.

Really, before you buy anything, you should submit the permit application and make sure what you propose is accepted by your AHJ.