4 Solis 10k AC voltage spike

P

pvstall

Member
Location
Illinois
Occupation
Installer
We have 4 Solis 10k inverters, parrelleled, 4g thhn from inverter to breaker on each inverter. 4/0 aluminum wire runs from combiner box up to disconnect whis is 300 feet. This disconnect is tapped into lines below meter whis is all 4/0. If we power up 2 inverters, they get up to 8000 watts, the line voltage goes up to 258.0 volts at inverters, 252.0 volts at disconnect. If we power 3rd inverter up, voltage ramps up over 264.0 volts and shuts inverter down.
New wire, connections all tight, check line resistance in the 4/0 as well.

If no inverters are turned on, the voltage is 248.0 at both the disconnect and at combiner box at inverters.

Any ideas?
 
pvstall said:
We have 4 Solis 10k inverters, parrelleled, 4g thhn from inverter to breaker on each inverter. 4/0 aluminum wire runs from combiner box up to disconnect whis is 300 feet. This disconnect is tapped into lines below meter whis is all 4/0. If we power up 2 inverters, they get up to 8000 watts, the line voltage goes up to 258.0 volts at inverters, 252.0 volts at disconnect. If we power 3rd inverter up, voltage ramps up over 264.0 volts and shuts inverter down.
New wire, connections all tight, check line resistance in the 4/0 as well.

If no inverters are turned on, the voltage is 248.0 at both the disconnect and at combiner box at inverters.

Any ideas?
Click to expand...
Yes. You have voltage drop (rise) in your interconnection conductors and in the service conductors to the service. Your service voltage is already a bit high and the voltage rise is pushing the voltage out of the inverters' operating window. Can you program the inverters to open their voltage window enough to accommodate the rise?
 
Last edited:
pvstall said:
We have 4 Solis 10k inverters, parrelleled, 4g thhn from inverter to breaker on each inverter. 4/0 aluminum wire runs from combiner box up to disconnect whis is 300 feet. This disconnect is tapped into lines below meter whis is all 4/0. If we power up 2 inverters, they get up to 8000 watts, the line voltage goes up to 258.0 volts at inverters, 252.0 volts at disconnect. If we power 3rd inverter up, voltage ramps up over 264.0 volts and shuts inverter down.
Click to expand...
With two inverters on, is that 8kW total, or 16 kW total?

For reference, 8kW / 258V = 31A, 600 ft (round trip) of 4/0 Al (75C DC resistance of 0.1 ohms/kft) has a resistance of 0.060 ohms, and that accounts for 1.86V of voltage rise. You're seeing 6V with two inverters on, so there's either a lot or a little extra voltage rise (depending on 8 kW vs 16 kW). If just a little, may be attributable to the unspecified length of #4 Cu.

Anyway, diagram out all the wiring from the meter (or disconnect if it is very close by) to the inverters, sizes and (fairly) exact lengths. Use that to calculate round trip resistance for the premises wiring. Take simultaneous measurements of the voltage at the meter and the voltage and current at the inverter(s) for the cases of 0, 1, 2 and possibly 3 inverters running.

For any given case (number of inverters operating), the current times the premises wiring resistance should be the observed voltage rise between the meter and the inverters. If you get a significant discrepancy between the calculated and observed values, there's a problem with the premises wiring, or your diagram is not accurate.

You can determine the resistance between the utility transformer and meter (including the transformer impedance) by comparing for two cases the change of voltage at the meter to the change in current at the inverters. If that resistance is too high, you'll either need to arrange to have that wiring or transformer upsized, or have the utility drop the no-load voltage, or reconfigure the inverters to allow operation at a higher voltage, or some combination of those.

Cheers, Wayne
 
wwhitney said:
With two inverters on, is that 8kW total, or 16 kW total?

For reference, 8kW / 258V = 31A, 600 ft (round trip) of 4/0 Al (75C DC resistance of 0.1 ohms/kft) has a resistance of 0.060 ohms, and that accounts for 1.86V of voltage rise. You're seeing 6V with two inverters on, so there's either a lot or a little extra voltage rise (depending on 8 kW vs 16 kW). If just a little, may be attributable to the unspecified length of #4 Cu.

Anyway, diagram out all the wiring from the meter (or disconnect if it is very close by) to the inverters, sizes and (fairly) exact lengths. Use that to calculate round trip resistance for the premises wiring. Take simultaneous measurements of the voltage at the meter and the voltage and current at the inverter(s) for the cases of 0, 1, 2 and possibly 3 inverters running.

For any given case (number of inverters operating), the current times the premises wiring resistance should be the observed voltage rise between the meter and the inverters. If you get a significant discrepancy between the calculated and observed values, there's a problem with the premises wiring, or your diagram is not accurate.

You can determine the resistance between the utility transformer and meter (including the transformer impedance) by comparing for two cases the change of voltage at the meter to the change in current at the inverters. If that resistance is too high, you'll either need to arrange to have that wiring or transformer upsized, or have the utility drop the no-load voltage, or reconfigure the inverters to allow operation at a higher voltage, or some combination of those.

Cheers, Wayne
Click to expand...
Which is the long way of saying what I said. :D
 
wwhitney said:
Expanding on your response in several ways hopefully useful to the OP. Glad that makes you grin. : - )

Cheers, Wayne
Click to expand...
The first choice (simplest and cheapest) approach would be what we both suggested, which would be to see if the inverters' voltage window can be expanded to accommodate the voltage rise. I ran into this same situation a few years ago when in a small commercial system we installed some of the SMA inverters would drop off line on cool clear days. During those times the PV was running at maximum at the same time as the building's HVAC systems were off; the voltage rise in the service conductors to the pole mounted transformer over 100' away drove the service voltage up out of the inverters' operating voltage window until enough of them dropped out to bring it back down. In that case we were able to program the inverters to accept the higher voltage.
 
ggunn said:
The first choice (simplest and cheapest) approach would be what we both suggested, which would be to see if the inverters' voltage window can be expanded to accommodate the voltage rise.
Click to expand...
Agreed, once you are sure that the observed voltage rise is close to the expected voltage rise, and not a result of a bad connection or the like. So I would say diagramming out the premises wiring to determine the expected voltage rise is important.

Cheers, Wayne
 
wwhitney said:
With two inverters on, is that 8kW total, or 16 kW total?

For reference, 8kW / 258V = 31A, 600 ft (round trip) of 4/0 Al (75C DC resistance of 0.1 ohms/kft) has a resistance of 0.060 ohms, and that accounts for 1.86V of voltage rise. You're seeing 6V with two inverters on, so there's either a lot or a little extra voltage rise (depending on 8 kW vs 16 kW). If just a little, may be attributable to the unspecified length of #4 Cu.

Anyway, diagram out all the wiring from the meter (or disconnect if it is very close by) to the inverters, sizes and (fairly) exact lengths. Use that to calculate round trip resistance for the premises wiring. Take simultaneous measurements of the voltage at the meter and the voltage and current at the inverter(s) for the cases of 0, 1, 2 and possibly 3 inverters running.

For any given case (number of inverters operating), the current times the premises wiring resistance should be the observed voltage rise between the meter and the inverters. If you get a significant discrepancy between the calculated and observed values, there's a problem with the premises wiring, or your diagram is not accurate.

You can determine the resistance between the utility transformer and meter (including the transformer impedance) by comparing for two cases the change of voltage at the meter to the change in current at the inverters. If that resistance is too high, you'll either need to arrange to have that wiring or transformer upsized, or have the utility drop the no-load voltage, or reconfigure the inverters to allow operation at a higher voltage, or some combination of those.

Cheers, Wayne
Click to expand...
8k is on each in inverter.

2 inverters up running at that time was 16k.

The inverter perimeters cab be adjusted larger, or wider. The thing is when the inverters 3 and 4 are turned on, the voltages will go above and beyond 268v. This is over the grid company high voltage setting.
 
pvstall said:
2 inverters up running at that time was 16k.
Click to expand...
Thanks for clarifying. I also missed that your no inverter voltage was 248V.

So 2 inverters gives you 4V of voltage rise to the disconnect (248V becomes 252V) and 6V more of voltage rise to the inverters (at 258V). So 60% of the problem is between the disconnect and the inverters, and 40% of the problem is between the utility and the disconnect.

Also, logically, with 3 inverters you should have 263V at the inverter, and with 4 inverters, 268V at the inverter. With the approximation that all the inverters are producing the same current.

If you're not able to do anything about the 2V / inverter voltage rise between the utility and the disconnect, then you either have to address the 3V / inverter voltage rise between the disconnect and the inverters (e.g. find a problem and fix it, or run another set of parallel conductors), or adjust the inverter's voltage window.

pvstall said:
The inverter perimeters cab be adjusted larger, or wider. The thing is when the inverters 3 and 4 are turned on, the voltages will go above and beyond 268v. This is over the grid company high voltage setting.
Click to expand...
Does it matter that the voltage at your inverters is elevated as long as the voltage at the disconnect is within the POCO guidelines? Based on 248V becoming 252V at the disconnect with 2 inverters running, with 4 inverters running it should become 256V at the disconnect. Which is only 6.7% above the 240V nominal.

Cheers, Wayne
 
Based on ANSI C84.1 the nominal service voltage is high, still within the A range of 228-252V, but high. You might be at the start of a high impedance distribution line and the utility is cranking up the voltage so at the end of the line the voltage is still in the A range, or the utility might have a bad setting on a line regulator. It's worth having a talk with the utility to see if they can reduce the service voltage closer to the 230V nominal value. This can also be a sign that the utility has a line regulator that is not back feed compatible. Those are still out there, and they freak when back fed.
If that does not work out then the other options are fat conductors to lower the voltage rise in the part of the system you have control over and increasing the upper voltage limit on the inverters, but not over the ANSI limit of 252V at the POCC. What you don't want to do is increase the site voltage, in theory you can go up to 252V but I would not push it that far. Good luck.
 
Last edited:
pv_n00b said:
Based on ANSI C84.1 the nominal service voltage is high, still within the A range of 228-252V, but high. You might be at the start of a high impedance distribution line and the utility is cranking up the voltage so at the end of the line the voltage is still in the A range, or the utility might have a bad setting on a line regulator. It's worth having a talk with the utility to see if they can reduce the service voltage closer to the 230V nominal value. This can also be a sign that the utility has a line regulator that is not back feed compatible. Those are still out there, and they freak when back fed.
If that does not work out then the other options are fat conductors to lower the voltage rise in the part of the system you have control over and increasing the upper voltage limit on the inverters, but not over the ANSI limit of 252V at the POCC. What you don't want to do is increase the site voltage, in theory you can go up to 252V but I would not push it that far. Good luck.
Click to expand...
What about a buck-boost transformer between the AC combiner and the service? It occurs to me, though, that such a thing might exacerbate the voltage rise at the service point.
 
ggunn said:
... a buck-boost transformer between the AC combiner and the service.. might exacerbate the voltage rise at the service point.
Click to expand...

Exactly what pv_noob is saying not to do here.

pv_n00b said:
...What you don't want to do is increase the site voltage, in theory you can go up to 252V but I would not push it that far. Good luck.
Click to expand...

If you do this you have to take responsibility for any over voltage damaging the loads. Utility may not like it either.
 
You must log in or register to reply here.
Top