Best Approach for 1,000' Home Run?

[IceGuy]

I'd like your input on the best way to handle a 1,000' home run length to achieve 3% max voltage drop on a 90 KW array.
1. Use inverter 208V output at array and many, big Al cables (my numbers est $50,000, incl conduit)
2. Use inverter 208V output at array and step up/down using 480v transformers. (my numbers est $38,000, incl conduit)
3. Use inverter 480V output at array and step down to building 208v using 480v transformer. (my numbers est $33,000, incl conduit)
4. Direct bury #6 AWG PV cable for each of 18 strings. Allowable by NEC? Do I need conduit? I don't have any numbers yet.

 

[petersonra]

I'd like your input on the best way to handle a 1,000' home run length to achieve 3% max voltage drop on a 90 KW array.
1. Use inverter 208V output at array and many, big Al cables (my numbers est $50,000, incl conduit)
2. Use inverter 208V output at array and step up/down using 480v transformers. (my numbers est $38,000, incl conduit)
3. Use inverter 480V output at array and step down to building 208v using 480v transformer. (my numbers est $33,000, incl conduit)
4. Direct bury #6 AWG PV cable for each of 18 strings. Allowable by NEC? Do I need conduit? I don't have any numbers yet.

My bet would be to go with the highest voltage you can get out at the PV end and xfmr it down to the system voltage.

Putting in two sets of xfmrs seems undesirable to me.

You might want to consider a higher voltage inverter at the PV end if that is allowed.

Is it possible to get the utility to run lines out there for you?

 

[jaggedben]

4. Direct bury #6 AWG PV cable for each of 18 strings. Allowable by NEC? Do I need conduit? I don't have any numbers yet.

A couple people have mentioned the direct burial idea on this forum. To me it is highly questionable to direct bury PV DC, even if it might be technically allowed. In my opinion, PV wire is not acceptable for direct burial unless it is crosslisted as something that is.

Beyond the code issues, it could be an O&M nightmare to bury 18 source circuits, directly or in conduit. Suppose you have to diagnose a bad string? The option you haven't mentioned, which would be much better, is to use a combiner at the array(s) and run the combined DC output 1000' to the inverter. That or your option 3 are probably the best options.

Note that from a functional point of view it's a bit less risky to run long DC conductors. If you undersize DC conductors the risk is merely that you'll lose some energy harvest, as opposed to an inverter shutting down due to AC voltage being out of range. I even recall a white paper that put forward the somewhat novel view that oversizing DC conductors for voltage drop is not economical. (But that might be out of the scope of your responsiblity.)

 

[GoldDigger]

The option you haven't mentioned, which would be much better, is to use a combiner at the array(s) and run the combined DC output 1000' to the inverter. That or your option 3 are probably the best options.

A big yes to that as long as the PV array and inverters can work with DC at 500V or more. The results would be as good as running 480AC but without the transformer losses and skin effect issues on large diameter wires. However, remember that the voltage drop will be that of two lengths of conductor compared to the one length for three-phase AC.
I would recommend that if you take that route you go belt and suspenders (i.e. that the wires are kept dry and are rated for wet use), since moisture-related problems can be much worse with DC.

 

[IceGuy]

A couple people have mentioned the direct burial idea on this forum. To me it is highly questionable to direct bury PV DC, even if it might be technically allowed. In my opinion, PV wire is not acceptable for direct burial unless it is crosslisted as something that is.

Beyond the code issues, it could be an O&M nightmare to bury 18 source circuits, directly or in conduit. Suppose you have to diagnose a bad string? The option you haven't mentioned, which would be much better, is to use a combiner at the array(s) and run the combined DC output 1000' to the inverter. That or your option 3 are probably the best options.

Note that from a functional point of view it's a bit less risky to run long DC conductors. If you undersize DC conductors the risk is merely that you'll lose some energy harvest, as opposed to an inverter shutting down due to AC voltage being out of range. I even recall a white paper that put forward the somewhat novel view that oversizing DC conductors for voltage drop is not economical. (But that might be out of the scope of your responsiblity.)

How would MPPT be affected by combining string wiring?

 

[jaggedben]

How would MPPT be affected by combining string wiring?

Unless you have an inverter with multiple MPPT inputs, it wouldn't be affected in any meaningful way. I assumed you're using a single central inverter with one MPPT input, since that's how almost all of them are.

 

[Zee]

1.
You may know this:
Use Vmp and Imp for DC V drop calcs.

2.
Isc (times other factors) and Vmax-DC (Voc times cold temperature factor) for ampacity calcs and string sizing respectively.

3.
The longer your strings, the better for V drop. There is a double benefit as Amps (combined amps of parallel strings) will be lower and V will be higher.

4. Combining dc will result in MUCH less copper. Altho very beefy and hard to pull.......
Spaghetti, aka multiple dc source circuits, in conduit needs to be seriously derated.....in your case of 36 wires, subtracting 60%.

 

[BillK-AZ]

Consider using long DC runs from remote DC Combiners to the inverters located close to the AC service panel. Most inverters have a wide DC input range, making voltage drop mostly a concern about power loss. If the power loss is of concern, use PV modules that are rated higher to make up for the power loss.

With many PV modules priced at about a dollar per watt, and wiring at historic high prices, this is usually the winning choice.

Also consider multiple conduits so as to keep conduit fill derating low, but space the conduits apart as required.

 

[beanland]

Combiner DC back to Inverter

Combiner DC back to Inverter

I have worked on a slew of PV systems. It is fine and common to direct bury the DC. For the run you are using, run from the combiner back to the inverter using direct buried XHHW-2 AL. Conduit is your option. Keep the AC link from inverter to panel short. Most inverters have transformers in them so adding any more transformers adds to losses. Get a 208V inverter and run the DC.