Inaccurate Voltage and Watt readings

[crazydane]

Good to know, thanks.

Yesterday my GEM3 voltage readings shot up higher for some reason. Recall this is the GEM that was reporting around 110 volts vs 120 until I power cycled it.

These are my variances now:



Note how around 12:30 yesterday, the grey line (GEM3) went from 123 volts to 127 volts and remained there.

During night time and on cloudy/rainy days, my voltages at all locations where I have energy monitors is right at 123 volts +/- 0.1 volts using my Fluke, which is what I would expect since all those locations are connected using very large cables (500 MCM down to 3/0).

I'm seeing variances of over 5 volts, which doesn't seem right.

[ben]

Good to know, thanks.

Yesterday my GEM3 voltage readings shot up higher for some reason. Recall this is the GEM that was reporting around 110 volts vs 120 until I power cycled it.

These are my variances now:


Note how around 12:30 yesterday, the grey line (GEM3) went from 123 volts to 127 volts and remained there.

During night time and on cloudy/rainy days, my voltages at all locations where I have energy monitors is right at 123 volts +/- 0.1 volts using my Fluke, which is what I would expect since all those locations are connected using very large cables (500 MCM down to 3/0).

I'm seeing variances of over 5 volts, which doesn't seem right.

Is there any chance you're connected to 2-phases of a 3-phase system? I had Paul look at your Main Panel and blue/red is usually used to identify 2 phases of 3-phase.

[crazydane]

Definitely just regular 1-phase 400A residential setup. I did all the wiring myself and decided to use those colors back when I did it a few years back.

The good news is that after power cycling GEM3 this morning, I'm now seeing what I would expect when comparing the Main Net CTs vs. the Solar Net CTs:



Voltages still look off, but rather then it being GEM 3 that has wide swings, today it was the 1240 in the server room (the one that doesn't like Split-100 CTs).



I'll spend some time redoing all my connections to make sure they are nice and tight and relive tension on that PT that you spotted and then just things run for a few days to see if things settle.

[ben]

Good to know, thanks.

Yesterday my GEM3 voltage readings shot up higher for some reason. Recall this is the GEM that was reporting around 110 volts vs 120 until I power cycled it.

These are my variances now:

Note how around 12:30 yesterday, the grey line (GEM3) went from 123 volts to 127 volts and remained there.

During night time and on cloudy/rainy days, my voltages at all locations where I have energy monitors is right at 123 volts +/- 0.1 volts using my Fluke, which is what I would expect since all those locations are connected using very large cables (500 MCM down to 3/0).

I'm seeing variances of over 5 volts, which doesn't seem right.

Definitely just regular 1-phase 400A residential setup. I did all the wiring myself and decided to use those colors back when I did it a few years back.

Gotcha.

If you can catch it in the act, can you probe the same outlet with the fluke? When solar comes on, it may bump up the voltage, the dipping down is strange and could be a couple of things.

Are all GEM PTs on the same phase? A loose neutral could cause the voltage balance to be off.

[crazydane]

I have been watching the voltages reported by my various energy monitors over the last week and see a distinct pattern.

Monday 3/1: (rainy)



Tuesday 3/2:



Wednesday 3/3:



Thursday 3/4:



Friday 3/5:



Saturday 3/6:



Sunday 3/7:



Except for Monday, it has been very sunny all week and when looking at the voltage graphs for the sunny days, the uplift from the power generated by the solar farm is very evident.

Per the above voltage curves, each morning around 10am the voltage drops on that one 1240 and it returns to normal around 5pm.

If we look at the power output from the solar farm, it reaches about 25,000 watts around 10am and dips below 25,000 watts again around 5pm:



I bet the next time I have a rainy/cloudy day, that one 1240 will behave normally.

[ben]

Are 500295/500296 on different legs? We have seen solar installs pull the voltage down on one leg when generating.

EDIT: Also, when it gets pulled down, you can check with your multimeter between the phase and neutral to see if it shows the same thing.

[crazydane]

I'm not sure if the 500295/500296 are on different legs or not. The 5000296 is just sitting in my office right now with no CT's connected just to track voltage.

Here's the voltage tracking output from another energy monitor connected directly to the solar sub-panel:



Both legs track pretty close to each other and at no point did either leg drop below 121.7 volts.

Tomorrow, assuming it is still sunny and I experience the issue, I'll check voltage in the server room where the 500295 is connected.

[ben]

I'm not sure if the 500295/500296 are on different legs or not. The 5000296 is just sitting in my office right now with no CT's connected just to track voltage.

Here's the voltage tracking output from another energy monitor connected directly to the solar sub-panel:



Both legs track pretty close to each other and at no point did either leg drop below 121.7 volts.

Tomorrow, assuming it is still sunny and I experience the issue, I'll check voltage in the server room where the 500295 is connected.

Yeah, looks like those curves on the other 3 are correct.

What F/W is that ECM-1240 running? I'll check with Paul tomorrow.

[crazydane]

I'm pretty sure I figured out the issue. I took a closer look at the receptacle that is powering the 1240 (and Etherport) and traced the feeder back to a UPS...

Took a look at the UPS logs and sure enough:



So once the input voltage reaches a certain value, it "compensates for a high input voltage". I modified the config to allow the output to rise to 135 volts (was set to 127 volts), so that should address the issue.

Sorry to have taken us on a wild goose chase on this one!

[ben]

I'm pretty sure I figured out the issue. I took a closer look at the receptacle that is powering the 1240 (and Etherport) and traced the feeder back to a UPS...

Took a look at the UPS logs and sure enough:

So once the input voltage reaches a certain value, it "compensates for a high input voltage". I modified the config to allow the output to rise to 135 volts (was set to 127 volts), so that should address the issue.

Sorry to have taken us on a wild goose chase on this one!

That's good to know. I'm pretty sure we've had this issue before and weren't able to figure it out.