AC Disconnect Sizing and Type for Line Tap (no backfeed)

[reader2580]

Get a free account for SolarEdge's Designer. Then plot out your panels and put in your equipment. It will tell you also, recommended optimizers and figure out string length max for you.

Thanks for the tip on the SolarEdge Designer. I was not aware of this tool. I have a SolarEdge installer account so I was able to use that to access the Designer. The Designer would not let me create a string with more DC power than 6,000 watts per SolarEdge's requirements that a string cannot be larger than 6,000 watts.

[theoak]

Get a free account for SolarEdge's Designer. Then plot out your panels and put in your equipment. It will tell you also, recommended optimizers and figure out string length max for you.

[reader2580]

I would consult the installation manual. I know for the HD Wave 11.4kW inverter I installed there's a combiner box inside the inverter, the strings are combined before the inverter, there aren't separate channels. If the terminals inside the inverter can take ~30A then there's no difference between combining the strings at the inverter or at the array.

I read the installation manual and it doesn't mention much about actually connecting up the strings. It also doesn't mention the 6,000 watt limit per string. That limit is in other Solaredge documents.

The price difference for the wire to do two strings is about $50. I will follow Solaredge's guidance that strings not exceed 6,000 watts unless I find something official from Solaredge that says otherwise. A decent combiner box will eat up the $50 in savings on wire pretty easily.

[nwdiver]

Power is watts, not volts. Power is proportional to the square of the current through
a given resistance, voltage is only proportional. Bruce Roe

... yep... and Watts = (Volts)(Current). If current doesn't change then volts changes => Power loss due to resistance is experienced as voltage drop.

[bcroe]

.... power lost in the line to resistance IS voltage drop. Current doesn't change and energy is conserved => voltage must drop.

Power is watts, not volts. Power is proportional to the square of the current through
a given resistance, voltage is only proportional. Bruce Roe

[nwdiver]

Power lost in line resistance, yes, not voltage drop. Bruce Roe

.... power lost in the line to resistance IS voltage drop. Current doesn't change and energy is conserved => voltage must drop.

[bcroe]

. Voltage drop is a function of current squared.

Power lost in line resistance, yes, not voltage drop. Bruce Roe

[Mike90250]

..... My understanding has always been that AC has less voltage drop than DC.....

not correct. DC actually has lower drop than AC. New ultra HV lines are DC, because of capacitive and inductive loss over hundreds of miles with 60 hz AC

For the same wattage, high voltage has lower loss than lower voltage. Most of the loss is in the resistance of the copper wire.
For a 5kw PV array, a HV (500V) DC array feeding a string inverter has lower loss in the cabling than using a long AC (240V) line. You can add heavier copper wires to reduce the loss, but you have doubled your wire expense. Some use even heavier 1 or 2 OO aluminum feeder wire which is less expensive than copper

https://en.wikipedia.org/wiki/High-voltage_direct_current

[nwdiver]

Soalredge has a maximum string wattage of 6,000 watts for the 11.4KW model. You have to have two physically separate strings once over 6,000 watts.

I would consult the installation manual. I know for the HD Wave 11.4kW inverter I installed there's a combiner box inside the inverter, the strings are combined before the inverter, there aren't separate channels. If the terminals inside the inverter can take ~30A then there's no difference between combining the strings at the inverter or at the array.

I think the conversion losses are not going to change much whether it's before or after the line losses.
At least my assumption is that the conversion losses are basically a percentage of power - so whether you have X * 97% or 97% * X, it winds up being the same result.
I think all the gains you'd have in efficiency would be from less line losses.

I can see your point. I was thinking in terms of being taxed on net vs gross income. If your inverter is at the array you're being taxed on gross income vs at the meter is net (after line losses) income. The real benefit would be for oversized arrays. If the inverter is capped then line losses are irrelevant.

[foo1bar]

Not only is DC more efficient due to higher voltages but your conversion losses of DC=>AC are taken AFTER line losses not before.

I think the conversion losses are not going to change much whether it's before or after the line losses.
At least my assumption is that the conversion losses are basically a percentage of power - so whether you have X * 97% or 97% * X, it winds up being the same result.
I think all the gains you'd have in efficiency would be from less line losses.
Do you have a performance curve or something that shows less % losses when the DC voltage is lower? Because that would contradict my assumption.

[reader2580]

Soalredge has a maximum string wattage of 6,000 watts for the 11.4KW model. You have to have two physically separate strings once over 6,000 watts.

[nwdiver]

I would need to do two strings as Solaredge doesn't allow a single string that big. I calculated 10 AWG and 12 AWG for the two strings based on 350 volts that Solaredge uses and a 2% loss. One would be 13.1 amps and the other 17.4 amps.

You can combine the strings at the array and use a pair of #8 home run back to the inverter but a pair of #10 and a pair of #12 would work too.

[reader2580]

You could probably get away with running a pair #8 from the panels. My voltage drop calculator says that your line losses would be ~2% and if your 11.4kW inverter is maxed line losses are 0. VS if you were to mount the inverter at the panels and run AC you would need #4 to get a 2.3% voltage drop and you would also be required by code to run a separate #6 neutral.

I would need to do two strings as Solaredge doesn't allow a single string that big. I calculated 10 AWG and 12 AWG for the two strings based on 350 volts that Solaredge uses and a 2% loss. One would be 13.1 amps and the other 17.4 amps.

[reader2580]

Yes, the optimizers still work with new inverter. I replaced one of the two inverters with the 11.4KW one to make sure it worked. It generated power just fine until I disconnected the panels for the move recently.

[nwdiver]

I am relocating an existing solar system and also adding four more panels at the same time. I am changing to a single 11.4KW Solaredge inverter I got cheap since my two Solaredge inverters were maxed out. I want to keep one of the old inverters to add more panels at a later date. Garage is being torn down and replaced with much larger garage so panels have to come off roof and ground mount has to move.

Current config is combo of ground mount and panels on detached garage roof with two Solaredge inverters. AC runs 80 feet to house with about a ten foot run of DC to each string.

New config will be two ground mounts with a single inverter to start. Original plan was to run AC 200 feet to inverter with DC going about ten feet to panels. It sounds like I should mount the inverters close to meter so AC run is a lot shorter and DC run is much longer.

You could probably get away with running a pair #8 from the panels. My voltage drop calculator says that your line losses would be ~2% and if your 11.4kW inverter is maxed line losses are 0. VS if you were to mount the inverter at the panels and run AC you would need #4 to get a 2.3% voltage drop and you would also be required by code to run a separate #6 neutral.

Voltage drop is a function of current squared. At 11.4kW and ~400vdc / 240vac your inverter input is ~30ADC in and ~48AAC out. So line losses on the AC side all things being equal would be ~250% higher.

I would also double check to make sure your optimizers are compatible with your 'new' 11.4kW inverter.