Learn from my mistake

Status
Not open for further replies.

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Consulting Electrical Engineer - Photovoltaic Systems
If you must interconnect a 480V inverter to a 208V service and you want to use an autotransformer to do it, be sure that your 208V service is actually 208V, because if it's on the high side of the acceptable range at the service, the transformed voltage plus the voltage rise in the conductors may result in the voltage at the inverter being too high for the inverter to operate consistently. Isolation transformers have (or at least can have) taps where the voltage can be adjusted but autotransformers (at least the ones I have found) do not.
 
If you must interconnect a 480V inverter to a 208V service and you want to use an autotransformer to do it, be sure that your 208V service is actually 208V, because if it's on the high side of the acceptable range at the service, the transformed voltage plus the voltage rise in the conductors may result in the voltage at the inverter being too high for the inverter to operate consistently. Isolation transformers have (or at least can have) taps where the voltage can be adjusted but autotransformers (at least the ones I have found) do not.


How would one calculate the deviation from the theoretical voltage ratio in an autotransformer, based on load?

In otherwords, given 215V operational on the theoretically 208V side, and 90% of total KVA worth of load through the autotransformer, and any commonly known performance parameters of the autotransformer (such as impedance or efficiency), how would you calculate the actual operating voltage on the theoretical 480V side?
 
Last edited:
How would one calculate the deviation from the theoretical voltage ratio in an autotransformer, based on load?

In otherwords, given 215V operational on the theoretically 208V side, and 90% of total KVA worth of load through the autotransformer, and any commonly known performance parameters of the autotransformer (such as impedance or efficiency), how would you calculate the actual operating voltage on the theoretical 480V side?

If you know the percent impedance of the transformer, you can calculate the voltage difference under load using that. I would ignore the magnetizing current of the transformer and the voltage difference in the POCO side wiring associated just with that.
Then just keep firmly in mind that everywhere (except the magnetizing current contribution) that you have a voltage drop in normal calculations you will have a voltage rise instead, since the PV will be the source of the power.
 
If you know the percent impedance of the transformer, you can calculate the voltage difference under load using that. I would ignore the magnetizing current of the transformer and the voltage difference in the POCO side wiring associated just with that.
Then just keep firmly in mind that everywhere (except the magnetizing current contribution) that you have a voltage drop in normal calculations you will have a voltage rise instead, since the PV will be the source of the power.

So if the above example has 5% impedance, would this be your calculation?
480/208 * 215 * 1.05 = 521V
 
If you know the percent impedance of the transformer, you can calculate the voltage difference under load using that. I would ignore the magnetizing current of the transformer and the voltage difference in the POCO side wiring associated just with that.
Then just keep firmly in mind that everywhere (except the magnetizing current contribution) that you have a voltage drop in normal calculations you will have a voltage rise instead, since the PV will be the source of the power.

The essential point I was making is that with at least some isolation transformers you can adjust the voltage but with most if not all autotransformers you cannot. I just learned an expensive lesson.
 
So if the above example has 5% impedance, would this be your calculation?
480/208 * 215 * 1.05 = 521V
Yes. That may or may not not be right, but it is how I would do it. I might multiply that by .9 if only partially loaded as you indicated.

Sent from my XT1585 using Tapatalk
 
The essential point I was making is that with at least some isolation transformers you can adjust the voltage but with most if not all autotransformers you cannot. I just learned an expensive lesson.
Not all. We put +/- 5 & or 10% taps on MOST autotransformer we spec & supply. Perhaps you just need to find a different supplier?

Sent from my SM-G900V using Tapatalk
 
Maybe when you find out you can share the multiplier; 15%, 20%, etc?

Here's what I can share. It was my idea to use autotransformers instead of isolation trannies for installing 480V inverters on 208V services because they are cheaper and easier to implement. What I didn't count on was that sometimes the utility voltage runs a bit high (which they are unable/unwilling to adjust), and that, combined with the voltage rise in the conductors, in some installations is causing the inverters to drop offline due to AC overvoltage, even when the AC voltage window is opened as far as it will go. Remediation is going to be expensive, be it the installation of a bucking transformer or replacement of the autotransformer with... something else.
 
Here's what I can share. It was my idea to use autotransformers instead of isolation trannies for installing 480V inverters on 208V services because they are cheaper and easier to implement. What I didn't count on was that sometimes the utility voltage runs a bit high (which they are unable/unwilling to adjust), and that, combined with the voltage rise in the conductors, in some installations is causing the inverters to drop offline due to AC overvoltage, even when the AC voltage window is opened as far as it will go. Remediation is going to be expensive, be it the installation of a bucking transformer or replacement of the autotransformer with... something else.

I'm sure you know this, but just in case...

Beware suggestion to just put in a 'input choke.' Yes, they are rated in voltage drop (they misnamed them years ago and it stuck; 3% impedance, 5% impedance, etc) but that is at FULL CURRENT LOAD rating; at idle, they will do NOTHING to reduce the Ac voltage the drive sees. That will be when any AC voltage overlevel detect circuit will see highest voltage of course.

And you are 100% sure it AC input voltage detection kicking them out? I can't say I have never heard of that detection; just dc bus voltage level. If I had time I would go back and reread all the posts to see what drive it is to check myself. Being from Missouri and all...
 
I'm sure you know this, but just in case...

Beware suggestion to just put in a 'input choke.' Yes, they are rated in voltage drop (they misnamed them years ago and it stuck; 3% impedance, 5% impedance, etc) but that is at FULL CURRENT LOAD rating; at idle, they will do NOTHING to reduce the Ac voltage the drive sees. That will be when any AC voltage overlevel detect circuit will see highest voltage of course.

And you are 100% sure it AC input voltage detection kicking them out? I can't say I have never heard of that detection; just dc bus voltage level. If I had time I would go back and reread all the posts to see what drive it is to check myself. Being from Missouri and all...

Input choke is not on my radar, but thanks for the warning. When the inverters drop off they display an error code that means AC voltage too high.
 
I'm sure you know this, but just in case...

Beware suggestion to just put in a 'input choke.' Yes, they are rated in voltage drop (they misnamed them years ago and it stuck; 3% impedance, 5% impedance, etc) but that is at FULL CURRENT LOAD rating; at idle, they will do NOTHING to reduce the Ac voltage the drive sees. That will be when any AC voltage overlevel detect circuit will see highest voltage of course.

And you are 100% sure it AC input voltage detection kicking them out? I can't say I have never heard of that detection; just dc bus voltage level. If I had time I would go back and reread all the posts to see what drive it is to check myself. Being from Missouri and all...
Are you aware that ggunn is talking about grid interactive PV inverters that are pushing power to POCO, not VFDs taking POCO power to drive motor loads?
Because of that direction of power flow any "input filter" or increased impedance will make the problem worse, not better.

Sent from my XT1585 using Tapatalk
 
...

Beware suggestion to just put in a 'input choke.' Yes, they are rated in voltage drop (they misnamed them years ago and it stuck; 3% impedance, 5% impedance, etc) but that is at FULL CURRENT LOAD rating; at idle, they will do NOTHING to reduce the Ac voltage the drive sees. That will be when any AC voltage overlevel detect circuit will see highest voltage of course. ...

To follow on Goldigger's post, since these are PV inverters it is full load that we are worried about. The problem doesn't present itself when the inverter hasn't started up and is just checking line voltage before it starts outputting current. (If the voltage is too high before startup then that is probably something the utility really needs to fix. )
 
Last edited:
To follow on Goldigger's post, since these are PV inverters it is full load that we are worried about. The problem doesn't present itself when the inverter hasn't started up and is just checking line voltage before it starts outputting current. (If the voltage is too high before startup then that is probably something the utility really needs to fix. )
To elaborate, the POCO voltage is high but still in their acceptable range. The voltage rise in the lines, the impedance of the transformer, and the higher than optimal line voltage add together to push the voltage at the terminals of the inverter out of its comfort zone as soon as the array comes up to near its peak power.
 
Unfortunately I too have run into this with strong inverters that have a long run to the POI.

I don't fully understand the transformer impedance portion of the voltage rise. Does it apply generally that the actual voltage change across a transformer will differ from the nominal windings ratio by the magnitude of the transformer impedance?



Sent from my iPhone using Tapatalk
 
Here's what I can share. It was my idea to use autotransformers instead of isolation trannies for installing 480V inverters on 208V services because they are cheaper and easier to implement. What I didn't count on was that sometimes the utility voltage runs a bit high (which they are unable/unwilling to adjust), and that, combined with the voltage rise in the conductors, in some installations is causing the inverters to drop offline due to AC overvoltage, even when the AC voltage window is opened as far as it will go. Remediation is going to be expensive, be it the installation of a bucking transformer or replacement of the autotransformer with... something else.

A few comments. Unless I am missing it, I don't really see this as an autotransformer caused issue. Sure there is a tad more voltage drop/rise through the transformer, but pretty minimal no? All other things being equal/proportional/whatever, I see this still happening without any transformers (again unless the transformer VD/VR contribution is the straw breaking the camel's back). We have heard of other people having issues with their voltage windows on this forum, perhaps the manufacturer's need to widen their acceptable parameters a bit.

Whether it be to solve this problem, or for next time, why not just size your autos slightly differently? I believe they come in quite a range of voltage ratios beyond the common nominal ones. How many kva is the bank in question?
 
Unfortunately I too have run into this with strong inverters that have a long run to the POI.

I don't fully understand the transformer impedance portion of the voltage rise. Does it apply generally that the actual voltage change across a transformer will differ from the nominal windings ratio by the magnitude of the transformer impedance?



Sent from my iPhone using Tapatalk
At full load, yes. Provided, of course, that the turn ratio corresponds exactly to the nominal voltage ratio.
Control transformers in the US and also some power transformers in Europe have what is called compensated windings. This refers to a change in the turn ratio that makes the output voltage equal to the nominal voltage at full load with nominal input.
As a side effect the unloaded voltage will be correspondingly too high.

Sent from my XT1585 using Tapatalk
 
Status
Not open for further replies.
Top