PV voltage rise affecting other loads?

[PWDickerson]

I just did the math. The voltage drop from the inverter to the utility will be about 21%, so the voltage at the inverter will rise to about 290V if the utility voltage is 240V. Even if the utility voltage is on the low end, the inverter is going to shut down on every sunny day. This isn't going to work. How did you make the designer/owner aware? Did you show them the math? Even at half the rated power output, the inver will be operating at over 264V.

 

[Hv&Lv]

I figure about 255V the inverter is going to shut down anyway.

 

[Flicker Index]

The only way I can think of to salvage this (short of redoing the wire run) is to use transformers on each end to bump it up to 600 volts. That will keep voltage drop to about 2.6% which is pretty good, although you probably would still want to have taps on the transformers as you will have voltage drop in the two transformers so it's all going to add up.

However, like idling vehicles, those transformers will consume non-negligible amount of no-load watts 24/7 and the consumption varies drastically from one transformer to another. The watts per KVA consumption tends to be considerably higher for common transformers compared to utility grade transformers. Even if this is 60 watts each, that is 86kWh/month just to keep the transformer live

 

[jaggedben]

... Even if this is 60 watts each, that is 86kWh/month just to keep the transformer live

Which is ballpark 4% of what the solar system will produce. So it would be worth it if the solar system is otherwise producing close to zero.

 

[jaggedben]

I figure about 255V the inverter is going to shut down anyway.

Why do you not think it would be at the stated setpoint of 264V?

 

[Hv&Lv]

Why do you not think it would be at the stated setpoint of 264V?

Where does it say the set point is 264?
Maybe I missed that post...
The ones I’ve tested generally cutout at 255V sustained for a certain time. Once they get about 260 they should shut down.
Why would they ever need to produce 264 or higher?
Maybe for off grid but grid tie should not be that high.

 

[Barbqranch]

Could you put a buck boost transformer at the inverter to raise the voltage enough to compensate for the VD to the panel?

 

[electrofelon]

However, like idling vehicles, those transformers will consume non-negligible amount of no-load watts 24/7 and the consumption varies drastically from one transformer to another. The watts per KVA consumption tends to be considerably higher for common transformers compared to utility grade transformers. Even if this is 60 watts each, that is 86kWh/month just to keep the transformer live

I would be probably the first one to say the step up step down transformer scheme is a horrible solution and should be avoided at all costs due to the capital cost and the losses. In this situation, it may be a necessary evil.

No load losses have come down significantly in the last couple decades. A 15kva DOE2016 unit should be at about a quarter percent of name plate, but yes that is still about 40 Watts each for a 15kva. I have three 15kva at my house and they total 300 watts because some of them are older.

 

[jaggedben]

Where does it say the set point is 264?
Maybe I missed that post...
The ones I’ve tested generally cutout at 255V sustained for a certain time. Once they get about 260 they should shut down.
Why would they ever need to produce 264 or higher?
Maybe for off grid but grid tie should not be that high.

I've been reading forever that the IEEE standard these inverters typically go by is +10% to -12% of nominal voltage. +10% is 264V on a nominal 240V system. Sorry to not cite sources, those numbers have just been in my head for years. Maybe there are tighter standards in play than I realized.

 

[Flicker Index]

I would be probably the first one to say the step up step down transformer scheme is a horrible solution and should be avoided at all costs due to the capital cost and the losses. In this situation, it may be a necessary evil.

No load losses have come down significantly in the last couple decades. A 15kva DOE2016 unit should be at about a quarter percent of name plate, but yes that is still about 40 Watts each for a 15kva. I have three 15kva at my house and they total 300 watts because some of them are older.

I don't know your home's setup, but unless you're electric everything, 300W 24/7 is 216kWh, or a good 1/5 to 1/4 of total average monthly power use for many typical homes. Just to keep those transformers humming. Make it 400W 24/7 if they're in conditioned space during warmer month for added load on A/C. If this is all solar related asset, this energy consumption needs to be counted against the payback.

 

[winnie]

Not useful for the op with the connected 240V load, but are 600V single phase inverters commercially available? That could eliminate one transformer on a long single phase run.

Jon

 

[GoldDigger]

I would be probably the first one to say the step up step down transformer scheme is a horrible solution and should be avoided at all costs due to the capital cost and the losses. In this situation, it may be a necessary evil.

No load losses have come down significantly in the last couple decades. A 15kva DOE2016 unit should be at about a quarter percent of name plate, but yes that is still about 40 Watts each for a 15kva. I have three 15kva at my house and they total 300 watts because some of them are older.

What do you think of putting in a time clock contactor to de-energize the transformer outside solar production hours? That could cut idling losses by a factor of two and should not impact the inverter operation if you do not care about telemetry during off hours. Or you could just use the contactor to bypass the transformer.

 

[synchro]

At the maximum output from the inverter, the 50V drop mentioned in post #21 will be dissipating (i.e., wasting) more than 3kW in the wires. So that's another thing to consider.

 

[GoldDigger]

At the maximum output from the inverter, the 50V drop mentioned in post #21 will be dissipating (i.e., wasting) more than 3kW in the wires. So that's another thing to consider.

That is a good argument for step-up/step-down with a high intermediate voltage.

 

[Hv&Lv]

Probably need to get out of this discussion because at this point I’m really confused..
Isn’t this a grid tied system? Is there not an interconnect agreement with the utility or other requirements that are in play here on the high voltages?

@jaggedben mentioned numbers above that are correct, and it’s what he does so I yield to his expertise, but I didn’t think voltage following inverters would produce voltages that high unless they first read voltages that high to follow.
The ones I’ve tested drop out at a certain voltage and won’t reconnect unless the voltage goes to about 126.5V
So how would you get the voltages that high anyway unless your increasing set points?

 

[winnie]

The issue here is the very long feed between the PV inverter and the solar array. When the system is quiescent the voltage is whatever the utility is providing at the service point. The PV inverter senses this voltage and matches it, pushing current into the line.

Normally we talk about voltage drop when current flows in a long conductor, but in this case the utility is a solid voltage source, so the voltage at the service point remains relatively constant, and the voltage at the PV inverter has to rise to compensate.

If it were not for the load in the middle, then I'd suggest running the raw PV DC for the long run, and putting the inverter near the service point. The PV string voltage can be considerably higher than the service voltage, meaning lower current, lower % voltage drop, and putting the voltage adjusting component (the inverter) right next to the service point.

-Jon

 

[electrofelon]

What do you think of putting in a time clock contactor to de-energize the transformer outside solar production hours? That could cut idling losses by a factor of two and should not impact the inverter operation if you do not care about telemetry during off hours. Or you could just use the contactor to bypass the transformer.

Yeah I had thought about that. It may well be worth it.

If it were not for the load in the middle, then I'd suggest running the raw PV DC for the long run, and putting the inverter near the service point. The PV string voltage can be considerably higher than the service voltage, meaning lower current, lower % voltage drop, and putting the voltage adjusting component (the inverter) right next to the service point.

-Jon

I thought about running the DC back too. With the existing wiring, I assume everything would need to be combined into one string, which is not ideal but is what it is. Fronius has a15kw inverter which is about right assuming that 17.3kw is DC KW. That fronius does 1kv strings, but unfortunately the OP likely has only 600 V conductors to work with.

Without running more numbers, I'm not sure if running DC or step up step down with 600V is the better option.

 

[ggunn]

I would be probably the first one to say the step up step down transformer scheme is a horrible solution and should be avoided at all costs due to the capital cost and the losses. In this situation, it may be a necessary evil.

No load losses have come down significantly in the last couple decades. A 15kva DOE2016 unit should be at about a quarter percent of name plate, but yes that is still about 40 Watts each for a 15kva. I have three 15kva at my house and they total 300 watts because some of them are older.

I agree. We built a couple of systems that way. They were expensive to build, the installations were a PITA, and the losses were a wash.

 

[pv_n00b]

The inverters are going to have a maximum operating AC voltage imposed by the UL 1741 listing. The utility has a maximum voltage they are supposed to stay under as defined by ANSI. The allowed voltage drop is the difference between the two and while the OP has not provided much information it seems like there is no way this system will not shut down due to AC overvoltage every time it tries to start up. Do the math and see if it's even worth breaking ground.