PV System Clarification

[jaggedben]

Note that backfeeding PV generally raises the line voltage. So if the utility has high line voltage and you mask that from your own equipment with a buck transformer, you are likely still exacerbating the problem on their end of things, including for other customers. The real solution is lowering the impedence in the whole path.

 

[junkhound]

grid-connected inverter suddenly shuts down.....It continues this cycle until 3 PM at which point it remains on for the rest of the day. The most likely cause of this problem is

The answer, since OP list occupation as design engineer, is 'none of the above'; a correct answer would be that the inverter is not properly designed, it should have better hysteresis in the control loop for over or under voltage shutdown so that it does not cycle for 5 hours!

 

[electrofelon]

[/QUOTE]

The answer, since OP list occupation as design engineer, is 'none of the above'; a correct answer would be that the inverter is not properly designed, it should have better hysteresis in the control loop for over or under voltage shutdown so that it does not cycle for 5 hours!

Although that could be a nifty feature, if an inverter could throttle back it's output to keep voltage below a specified value, generally that would not be good solution.

 

[jaggedben]

The answer, since OP list occupation as design engineer, is 'none of the above'; a correct answer would be that the inverter is not properly designed, it should have better hysteresis in the control loop for over or under voltage shutdown so that it does not cycle for 5 hours!

People who design PV systems generally don't design the inverters.

 

[tallgirl]

So I have an update from the utility regarding the high/variable voltage for the PV system I have mentioned in this thread. But first, There seems to be some confusion about the voltage the utility is supposed to provide. It seems +/- 5% is what is typically required. I believe this is in ANSI C84.1. This is more inline with what the lineman said who showed up to look, said that it should be 114-123. I know that is a bit off from +/- 5%, either he was a little off or its a company policy? Anyway this seems to prove (not that I was ever in doubt) that this is a "POCO problem" not a "PV problem".

Is ERCOT different? I don't want to go scan years of logs, but 123 L-N would be a "nice" day in my old neighborhood. Between 125 and 130 was more often the case. That wasn't measured by the inverters, that was with with Continental Control Systems devices. I bought those things by the case.

https://ctlsys.com/product/wattnode-modbus/

 

[solardavez]

The answer here is 3 but the way the OP writes it, it sounds like a solar test question rather than one that is specific to a project. Obviously a test question doesn't provide every bit of detail for the system in question, but since you know it's sunny weather and it's the middle 5 hours of the day, you know the inverter will attempt to provide real power close to its nameplate, which is going to cause voltage rise with the high currents. If the utility voltage is high then the inverter will in this example see a voltage that is greater than Vnom + 10%. This could be caused by too much resistance in the output conductors (bad system design), or outlier cases like a high resistance caused by a bad connection or a failing breaker. It's not a utility 1547 issue because if the utility voltage were outside spec (say, Vnom + 12%), then the inverters would not ever try to turn on. (it would be a utility ANSI issue, but in my experience, that utility knows it has an ANSI issue and it absolutely does not care about how it affects your system)

I would disagree that there's an issue with inverter design here; the inverter sees an input within spec and it tries to turn on. The only other thing it could do would be to limit its current to try to keep the terminal voltage within spec, but as a designer, I'd rather see it keep trying so I could better troubleshoot from the logged errors.

On the policy side, DERs are the best and the majority of DER issues are because utilities aren't as interested as they need to be in finding solutions. Any reasonable solution has to be utility-coordinated because a single DER isn't seeing the whole picture. Should a DER limit its own output only because that home's air conditioner isn't running? No because that power should go somewhere else. There are cases when unmanaged DERs can cause issues, and utilities should be responsible for solving them while also being incentivized to let as much DER onto the grid as possible. Forcing inverters to run at unity power factor and then not figuring out how to communicate with them isn't a 50-year solution and it's not the fault of DER providers/owners because they are following the rules that utilities set.

 

[ggunn]

It's obviously a test question and the answer they want is #3. #1 and #2 are possible but much less likely. #4 is a red herring to weed out the clueless.

 

[electrofelon]

It's obviously a test question and the answer they want is #3. #1 and #2 are possible but much less likely. #4 is a red herring to weed out the clueless.

IMO it's a bad question. I agree number three is probably the answer they are looking for, but in my experience number one is actually the cause most of the time. Again, we have lots of long old 4800 volt lines around here and see this problem a lot. I don't think I have ever seen an inverter shut off due to undersized wiring, but I have seen them shut off due to high utility voltage half a dozen times.

 

[wwhitney]

IMO it's a bad question. I agree number three is probably the answer they are looking for, but in my experience number one is actually the cause most of the time.

If the utility's allowable voltage ceiling matches the inverter's allowable voltage ceiling, then (1) is ruled out, as a previous commenter noted. The inverter wouldn't keep trying every 5 minutes in that case.

If the utility's allowable voltage ceiling is lower than the inverter's, then I agree you can't distinguish between (1) and (3).

Oh, and since (1) references IEEE 1547, I think that voltage window is the same for the utility and the inverter, even if the utility is subject to some other tighter standard. So that would rule out (1).

Cheers, Wayne

 

[tallgirl]

If the utility's allowable voltage ceiling matches the inverter's allowable voltage ceiling, then (1) is ruled out, as a previous commenter noted. The inverter wouldn't keep trying every 5 minutes in that case.

If the utility's allowable voltage ceiling is lower than the inverter's, then I agree you can't distinguish between (1) and (3).

Oh, and since (1) references IEEE 1547, I think that voltage window is the same for the utility and the inverter, even if the utility is subject to some other tighter standard. So that would rule out (1).

Cheers, Wayne

If the utility is running near the top of the their allowable window the inverter will disconnect when it raises the voltage above the high voltage, then the voltage drops, it times out, reconnects, repeats.

you can do the math with the designed voltage drop in the lateral, then calculate how close to the top the utility can be, and the inverters can't operate at nameplate output without tripping.

 

[wwhitney]

If the utility is running near the top of the their allowable window the inverter will disconnect when it raises the voltage above the high voltage, then the voltage drops, it times out, reconnects, repeats.

Right. So if the IEEE 1547 limits referred to in (1) are the same as the inverter's limits, (1) is ruled out, as the inverter keeps trying.

Cheers, Wayne

 

[ggunn]

A company I worked for had an installed system with 6 or 8 7kW SMA inverters that would run fine in the summertime, but on cool clear days some of them would trip off in the middle of the day and it was never the same ones. It turned out that the service conductors from the pole mounted transformer were undersized.

In the summer when the HVAC was running most or all of the PV output was consumed on the premises, but when it was cool the HVAC was off and the PV was running at its maximum. Most of the PV output was being pushed back to the service and there was enough voltage rise in the service conductors to push the voltage out of the inverters' operating window. The inverters would start tripping off until enough of them quit to reduce the line voltage to where the rest could handle it. Then the tripped ones would start coming back on line and the whole thing would start again.

I still say #3.

 

[Hv&Lv]

Interesting…

from our point of view we run a little hot at our station regulators. 124 with a 2V bandwidth. We want to make all the money we can also..
Our PV loads have to be spaced out, simply because we don’t want issues with voltages running out of control.
I posted in another thread about too much PV on a line.
Simply put, you get a saturation of PV at the end of the line it’s not hard to get way out of band with the voltages.

Say you live out in rural areas.. your about a mile from the end of the line. There are 10 people between you and the end.
6 of them have solar…Voltage is 123 at the end. the first inverter is boosting the voltage to 125, second is boosting the voltage up to 127, third, etc…
If the inverters all boost by 2 volts that’s 12 volts total over the line impedance that offers little drop at all. Inverter tops out at 132V, so now all of them are fighting all day long and we get a call the voltage is too high.

what are we supposed to do?… DER caused this problem.

The regulators are set to co-gen so they don’t really regulate when the voltage flows backwards, as it’s different than the bi-directional setting. That setting can’t be used because the regulators fight the inverters until they drop off line. The regulators drop back down, the inverter comes back online, the regulator tries to regulate and it’s a cycle that doesn’t work.
Also, partly cloudy days have my phasors flipping back and forth so much the transformer relay doesn’t know what to do.

Now this isn’t meant to be an “it’s all DERs fault” post, because there is also the problem of POCOs having some really sorry maintenance going on. I know of utilities that install regulators, set the relays, and forget them. They run them to failure. What this means is none of the linemen truly have an understanding of what the relay sees and tells the regulator to do..
It also means the regulators screw up often before they fail. Taking one out of service (200/288) and sending it in for maintenance is about $3000 per regulator, each time it’s sent off.

As far as Electrofelons problem, it could simply be a wrong setting in the relay now that DER has been introduced.

As for the OP.. it could be #5.. There is too much DER on the line already and all the collective inverters are pushing the voltage way too high.

 

[electrofelon]

Right. So if the IEEE 1547 limits referred to in (1) are the same as the inverter's limits, (1) is ruled out, as the inverter keeps trying.

Cheers, Wayne

Well first I will acknowledge I am confused on IEEE 1547 vs ANSI C84.1 and which applies and when. The former seems to be +/- 5% and the latter 5%.

But regardless, usually when I see "high voltage", its also "variable and high". Its bouncing all over the place, even without the PV on and even on other than partly cloudy days which should rule out effects of other DG on the line. The system we just completed that I have mentioned that currently has an issue, I actually have not been able to "catch" it getting high enough to trip the inverters. The highest I have seen is 259 at the inverters. But it does happen now and then.

 

[electrofelon]

It turned out that the service conductors from the pole mounted transformer were undersized.

My pet peeve about "undersized wiring" is what defines undersized? IF you have a problem or not? Take again this current system of mine that has the issue. We have 4 volts of drop on the service conductors between the meter and the PV. Its an 800 foot run and the wire (without conduit) was $3500. We could have done two runs for an extra $3500 plus labor and conduit, and that 2 volts MIGHT have been just enough. Is our wire undersized? Obviously in a perfect world, POCO voltage would be +/- 5 and the inverters are +/- 10 so we have a definite number to work with, but thats not the way it is and its sorta a guessing game and gamble as to where to draw the line (when you have long runs).