hot grids triggering inverter overvoltage protection

[nerdralph]

I don't know if it is part of the UL 1741 spec, but I've noticed string inverters will fail to startup if they sense a grid voltage 5% over nominal spec of 240V. If the grid voltage is <252, they will sync to the grid and start producing power, and continue to operate as long as the grid voltage is no more than 10% over nominal spec (i.e. 264V).
In eastern Canada (and many other regions where electric heating is common), power companies run the grid hot to account for voltage drop in the winter when power draw is high. No-load voltage of 245V-250V is common. While it isn't common to see a no-load voltage over 252V, on sunny days with 2 6kW inverters running, I almost always see over 252V. Even with just one of the inverters running, 252V is common, so if I shut down one inverter for any kind of maintenance or testing, it may not start back up. One of my inverters allows me to increase the overvoltage limit beyond 264V, but so far I haven't found a way to change it on Solis 1P-4G-US inverters.

I see this issue only getting worse as more people add solar PV. As more houses on the same power phase in a neighborhood (around here power is distributed in 3 phases), that will elevate the local distribution voltage, making it more common to encounter >252V.

[bcroe]

Besides the Power Co sending out a high voltage, your inverter(s)
pushing power back will raise it even higher. Some of this is due to
inverter current building up voltage across the resistance of your own
wiring, that was initially adding 9V here. Making sure that is
minimized is a first step. I had to use larger AC cable.

There are several solutions possible. One is ask the PoCo to lower
their voltage, that eventually worked here. Another is add an auto
transformer to knock a few volts off that seen by the inverter, a
clumsy solution at best. Another is to program the inverter voltage
monitors to a higher trip voltage. Here the inverters were capable of
programing for 277VAC operation, so the high line was of no danger
to them. good luck, Bruce Roe

[nerdralph]

Besides the Power Co sending out a high voltage, your inverter(s)
pushing power back will raise it even higher. Some of this is due to
inverter current building up voltage across the resistance of your own
wiring, that was initially adding 9V here. Making sure that is
minimized is a first step. I had to use larger AC cable.

Agreed. My comment about more PV on the street leading to higher voltages refers to the voltage drop (or gain) from the inverter to the 240V transformer on the street.
I also agree on the larger cable. I was occasionally seeing 24.1A on the #10 Cu circuit going to my garage. The CEC limit (same as NEC) for #10 continuous is 24.0A, so I'm upgrading the circuit to 6/3 Al & 8/3 Cu.

[peakbagger]

Fronius would supply a password to customers who were experiencing out of spec voltages to increase the high set point. A friend had to do it to his until the utility figured out what was wrong on their end. A properly designed and operated distribution system should not have this issue. There are devices called remote tap changers that allow the utility to modify circuit voltage to keep it in the acceptable range.

[bcroe]

I also agree on the larger cable. I was occasionally seeing 24.1A on the #10 Cu circuit going to my garage. The CEC limit (same as NEC) for #10 continuous is 24.0A, so I'm upgrading the circuit to 6/3 Al & 8/3 Cu.

Aluminum is an economical upgrade, if the proper terminations are
used. Using 4/0 AL Triplex for a 600 ft loop, I managed to reduce
voltage loss to 1% for 60A inverter output. A full sized distribution
box terminated this cable at one end. The other end required me
to splice in a short length of 1/0 copper to feed a 100A breaker.

Fronius IG+ inverters have 2 output voltage monitors. The line to
line monitor was nominally set to 264VAC, but could be adjusted
to 284. The line to neutral was nominally set to 132VAC, but could
be adjusted to 142. Another possible approach would be to
configure them for 277VAC operation. Bruce Roe