High Voltage over L1-L2

[sdscuba]

I've searched the forum to find similar issues with what I am trying to resolve, but didn't see an exact answer or resolution. Here's the problem I'm trying to resolve.

I have 22 - 250w solar panels with a Sunny Boy 5000 inverter.
The system has been flawless over the past 2 years.

Voltage at the service panel, that I have measured, ranges from 255v to 265v. :blink:
This is not a US installation..... install is in Mexico.

Recently, about 10 times over the past 2 months, the inverter will report grid fault errors when the voltage over L1-L2 exceeds 264v.

I've had the PoCo here several times to troubleshoot this. Each time they come out, they report no problems.
L1 and L2 would read between 128-132v. This they say is within their acceptable range.

Will a buck transformer do the trick, if installed at the service panel, to maintain 240v over L1-L2?
And will a transformer work with a PV system installed?

Any feedback will be greatly appreciated.

Greg.

 

[ggunn]

I've searched the forum to find similar issues with what I am trying to resolve, but didn't see an exact answer or resolution. Here's the problem I'm trying to resolve.

I have 22 - 250w solar panels with a Sunny Boy 5000 inverter.
The system has been flawless over the past 2 years.

Voltage at the service panel, that I have measured, ranges from 255v to 265v. :blink:
This is not a US installation..... install is in Mexico.

Recently, about 10 times over the past 2 months, the inverter will report grid fault errors when the voltage over L1-L2 exceeds 264v.

I've had the PoCo here several times to troubleshoot this. Each time they come out, they report no problems.
L1 and L2 would read between 128-132v. This they say is within their acceptable range.

Will a buck transformer do the trick, if installed at the service panel, to maintain 240v over L1-L2?
And will a transformer work with a PV system installed?

Any feedback will be greatly appreciated.

Greg.

You can get into the Sunny Boy's controls and open up that voltage range a bit. You'll need to get SMA customer support to help with it because that parameter is not something you can easily get to. Youll probably need to install a piggyback com card in the inverter and connect it to a laptop. CS can tell you how to calculate the password and they will likely ask you to sign a document swearing that you won't use the power this gives you for doing evil things with it.

 

[sdscuba]

You can get into the Sunny Boy's controls and open up that voltage range a bit. You'll need to get SMA customer support to help with that, that parameter is not something you can easily get to. Youll probably need to install a piggyback com card in the inverter and connect it to a laptop.

I did contact SMA initially and they directed me change a dip setting in how the inverter checks voltage. It was factory set to look at L1-N and L2-N. It was modified to check L1-L2. But I'm still getting grid fault errors. Not as many, but still too many. I was not informed that the range can be changed internally. I will contact them to see what this all entails.

However, the question still stands as to the buck transformer as we've seen CFL bulbs burn out (4 of them) since this problem started. I'm trying to get two things resolved.... inverter shutdowns and protecting everything else in the house.

 

[ATSman]

I my be way off base but if you removed one or more panels from the string wouldn't that lower the voltage
range the system operates in?

 

[jaggedben]

Will a buck transformer do the trick, if installed at the service panel, to maintain 240v over L1-L2?
And will a transformer work with a PV system installed?

I believe the answers are yes and yes, but be aware that this is rather unscrupulous. It can result in raising the voltage higher than is supposed to be allowed at the service, while your inverter still operates and thus raises the utility's voltage even further. Put another way, you could end up burning out the neighbors' CFLs instead of your own. Not really an above-board solution, and perhaps you could really piss off the utility if someone who understands this gets wind of it.

Best solution is probably to consult SMA as advised by ggunn. One other thing: You haven't mentioned anything about the distance from the inverter to the service, but if it is a long distance then upsizing the wires to lower voltage rise might work. Or alternatively move the inverter right next to the service and lengthen the DC run to reach it. Point being if they are routinely hitting 264V at the service then you don't really have a volt to spare for voltage drop on your side.

Bottom line though, as grid operator, they are supposed to maintain voltage within a window. Doesn't sound like they are really doing it and you should try to collect data or get them to collect data. Obviously if they just send out some guys for an hour and the problem doesn't happen to be occurring when those guys are out there then that's not really satisfactory on their part. Not the first time I've heard about these sorts problems in Mexico, though.

 

[sdscuba]

I my be way off base but if you removed one or more panels from the string wouldn't that lower the voltage
range the system operates in?

No. The problem is the voltage coming in from the street. Having the PV system on or off doesn't change the grid voltage. The voltage out from the inverter is constant, only the watts being produced from the panels into the inverter is variable.

From what I understand, the inverter will simply match the grid voltage in order to backfeed the produced electricity. The inverter auto disconnects when the grid range is over 264v.

So I'm trying to figure out a way to lower the grid voltage coming into the house, since the power company is of no help.

 

[jaggedben]

I my be way off base but if you removed one or more panels from the string wouldn't that lower the voltage
range the system operates in?

You're way off base. The inverter matches utility voltage on the AC side, regardless of voltage on the DC side. Removing a panel might also drop the voltage below the functional parameters for the DC side.

It's true that injecting less power from the inverter will result in ever so slightly less AC voltage rise, but not enough to affect the situation. And lowering the system power would sort of defeat the purpose of finding a solution.

The real problem here is the utility voltage.

 

[sdscuba]

I believe the answers are yes and yes, but be aware that this is rather unscrupulous. It can result in raising the voltage higher than is supposed to be allowed at the service, while your inverter still operates and thus raises the utility's voltage even further. Put another way, you could end up burning out the neighbors' CFLs instead of your own. Not really an above-board solution, and perhaps you could really piss off the utility if someone who understands this gets wind of it.

Best solution is probably to consult SMA as advised by ggunn. One other thing: You haven't mentioned anything about the distance from the inverter to the service, but if it is a long distance then upsizing the wires to lower voltage rise might work. Or alternatively move the inverter right next to the service and lengthen the DC run to reach it. Point being if they are routinely hitting 264V at the service then you don't really have a volt to spare for voltage drop on your side.

Bottom line though, as grid operator, they are supposed to maintain voltage within a window. Doesn't sound like they are really doing it and you should try to collect data or get them to collect data. Obviously if they just send out some guys for an hour and the problem doesn't happen to be occurring when those guys are out there then that's not really satisfactory on their part. Not the first time I've heard about these sorts problems in Mexico, though.

Thanks for the feedback!

Inverter is approximately 30' from the service panel.

They have seen voltage in and around 133v per leg. This they say is within tolerance. They say power here in Mexico is not 120v but rather 127v.

They are routinely hitting 260v-264v.

 

[ggunn]

I my be way off base but if you removed one or more panels from the string wouldn't that lower the voltage
range the system operates in?

Maybe, maybe not. What he is probably experiencing is voltage rise due to voltage drop between his inverter and the grid source of power. It could be as simple as his conductors to the service being too small for the current he is pushing through them, or it could be that his installation is way out on the end of a branch of the grid and the voltage drop is in those lines.

I am pretty sure, though, that the voltage operating window is a parameter you can change if you can get into the onboard firmware. SMA sells an interface that is USB on one end and a piggyback plugin board on the other that one can use with a computer (you have to download some software from the SMA website) to get to parameters which are inaccessible to the casual installer. There is a password (some munge of the date) that SMA has to give you the formula for, and before they give it to you they make you sign off on a document that releases them from responsibility for anything bad you do with the power this access gives you.

This is the same access (different parameters) you need to have to get a Sunny Boy to play nice with a Sunny Island battery inverter.

 

[jaggedben]

...Having the PV system on or off doesn't change the grid voltage. ...
.

Well, strictly speaking it does. And if the utility is already right at that 264V borderline, you might see voltage ever so slightly below with the PV off, then see it creep above it when the PV is turned on (and thus trip the inverter off, possibly creating a cycle). For that matter, it could be affected by how much power is used in the house, depending on point of interconnection. But this is all academic: the exact amount of voltage change depends on grid impedance, and it should be very small (less than 1V) and I agree that the problem is the voltage from the street.

 

[augie47]

Unless I am sadly mistaken, if you have 120v loads, a buck & boost on the 240v line is not the answer and it might add to your problems on the low end.
A constant voltage transformer would solve the problem but they are pricey

 

[sdscuba]

Maybe, maybe not. What he is probably experiencing is voltage rise due to voltage drop between his inverter and the grid source of power. It could be as simple as his conductors to the service being too small for the current he is pushing through them, or it could be that his installation is way out on the end of a branch of the grid and the voltage drop is in those lines.

I am pretty sure, though, that the voltage operating window is a parameter you can change if you can get into the onboard firmware. SMA sells an interface that is USB on one end and a piggyback plugin board on the other that one can use with a computer (you have to download some software from the SMA website) to get to parameters which are inaccessible to the casual installer. There is a password (some munge of the date) that SMA has to give you the formula for, and before they give it to you they make you sign off on a document that releases them from responsibility for anything bad you do with the power this access gives you.

This is the same access (different parameters) you need to have to get a Sunny Boy to play nice with a Sunny Island battery inverter.

The power company transformer is just across the street, however we are at the end of our complex. So this might be causing higher voltage from them. But again, they checked the transformer and reported no issues.

It's a new house using #6 wire from the street to the meter to our breaker panel.
What I did find odd was 40amp breakers for each leg at the service panel. ????

From the street connection to our breaker panel is less than 50'.

Will try contacting SMA for added assistance.

 

[jaggedben]

Thanks for the feedback!

Inverter is approximately 30' from the service panel.

They have seen voltage in and around 133v per leg. This they say is within tolerance. They say power here in Mexico is not 120v but rather 127v.

They are routinely hitting 260v-264v.

If the official grid standard is truly higher then SMA should simply help you change the inverter settings accordingly. Tolerance is typically 10% over to 12% under, hence the 264V threshold based on 240V. If its 254/127 then the threshold should be 179V? It would certainly be nice to have firmly established what the nominal grid voltage is officially supposed to be.

 

[ggunn]

The power company transformer is just across the street, however we are at the end of our complex. So this might be causing higher voltage from them. But again, they checked the transformer and reported no issues.

It's a new house using #6 wire from the street to the meter to our breaker panel.
What I did find odd was 40amp breakers for each leg at the service panel. ????

From the street connection to our breaker panel is less than 50'.

Will try contacting SMA for added assistance.

#6 sounds awfully small to me for service entrance conductors.

 

[SolarPro]

The OP may benefit from reviewing this article on voltage rise considerations for interactive PV systems. There are any number of other potential causes of nuisance tripping due to overvoltage on the ac side of a PV system. But one of the most common causes is excessive voltage rise when the inverter is operating at or near its nameplate power rating.

If the problem originates with the utility voltage range, you'll likely have to get into the inverter settings as suggested earlier.

 

[ggunn]

I've searched the forum to find similar issues with what I am trying to resolve, but didn't see an exact answer or resolution. Here's the problem I'm trying to resolve.

I have 22 - 250w solar panels with a Sunny Boy 5000 inverter.
The system has been flawless over the past 2 years.

Voltage at the service panel, that I have measured, ranges from 255v to 265v. :blink:
This is not a US installation..... install is in Mexico.

Recently, about 10 times over the past 2 months, the inverter will report grid fault errors when the voltage over L1-L2 exceeds 264v.

I've had the PoCo here several times to troubleshoot this. Each time they come out, they report no problems.
L1 and L2 would read between 128-132v. This they say is within their acceptable range.

Will a buck transformer do the trick, if installed at the service panel, to maintain 240v over L1-L2?
And will a transformer work with a PV system installed?

Any feedback will be greatly appreciated.

Greg.

How far (as the wire flies) is your inverter from your service? What size conductors are the AC conductors for this run?

 

[Sierrasparky]

Mexico ,,,, and what do you expect?

 

[sdscuba]

How far (as the wire flies) is your inverter from your service? What size conductors are the AC conductors for this run?

The inverter is ~50' to the Breaker Panel. breaker panel is then ~50' to the service panel.

I did notice that it is #8 wire being used throughout the path... street-breaker panel-inverter.

Today I'm reading 262v, an inverter tripped three times over th epast 2 days, plus 1hr ago. Seams that the AM is most susceptible to voltage variations.

 

[sdscuba]

Mexico ,,,, and what do you expect?

It's clearly a problem with either a substation or the transformer. And just a matter of time until the get it right! In theory, that is.

 

[jaggedben]

..

I did notice that it is #8 wire being used throughout the path... street-breaker panel-inverter.

...

All the way to the transformer? Or just to the service? I thought you said above the service was #6?

I calculate that your voltage drop along the inverter output will be a volt or two higher with #8 than with, say, #4. If this were the US and it was a 100A service with a #2 feeder to the sub then you'd probably have a couple extra volts of cushion. Still, it doesn't change the nature of the problem. It's about 1% Vdrop. I'm sure I've ever seen an nominal 240V supply read higher than about 255V in my work. (Which has been all in fairly urban areas in the states, to be sure.)