EEVblog 1390 - NEGATIVE Household Solar Consumption? WHY?

[EEVblog]

Dave explains a common inherent measurement problem in the solar industry with having two seperate solar power systems and monitors on the one house.
Why is Dave's household consumption going NEGATIVE on both his older Solar Analystics system and new Enphase Envoy system?
And why does the consumtion follow the sun?
And what is the easy fix?

[firewalker]

How can we tell the direction (giving power to the grid vs consuming) of an AC system? Why would a simple current transformer produce negative values? Doesn't a simple magnetic current transformer produce an AC voltage?

Alexander.

[IanJ]

How can we tell the direction (giving power to the grid vs consuming) of an AC system? Why would a simple current transformer produce negative values? Doesn't a simple magnetic current transformer produce an AC voltage?

Alexander.

Measurement of the phase angle between voltage and current from the output of a CT will give you direction of the current and thus import or export.

Ian.

[Robert Smith Eco Warrior]

You were not joking about a rats nest of wires. 

[HattedSquirrel]

Maybe I'm pointing out the obvious here, but instead of paralleling current transformers, Dave could have run both (production) wires through the same transformer core. Magnetic fields add and voila. This way he could keep using the extra transformer for his A/C.

[madires]

In case you have a separate bus bar for the loads and clamp the load current transformers onto the feed wire for that bus bar, you'll have a similar problem since you measure the load twice. For example, PV system #1 and #2 produce 2A each and the load is 1A, as in the video. So monitor #1 thinks PV #1 generates 2A and the load is 1A, resulting in 1A for the grid. Same for monitor #2 and PV #2. Both added will give you 2A for the grid, while in reality it's 3A and 1A for the load. It's only logical.

[station240]

Actually there is another option Dave didn't look at.
The problem is each Solar Generator see's the other's power generation.

Lets say both systems generate the maximum power output, and the house is using minimal/no power.
The Solar Analytics system sees the +3kW from it's connected Solar Generator, but also sees -5kW 'consumption' the other system.
The Enphase Envoy system sees the +5kW from it's connected Solar Generator, but also sees -3kW 'consumption' the other system.

So what is needed is to add the power generated by the other system in both cases.

If you stuff a second wire into each Consumption Current Transformer (CT), that current thru that wire can be added/subtracted from the total.
So run the 3kW wire backwards thru the Consumption CT for the 5kW system.
Run the 5kW wire backwards thru the Consumption CT for the 3kW system.
Reality probably sucks in that you possibly cannot fit a second wire into some of those Current Clamps, or the wires would need to be made longer.

Or just go nuts, and rewire that big fat Active wire so after the main breaker it goes to the Solar system, then back to the loads.

[IanJ]

I had a similar'ish problem with my own Solar PV system which I sorted out just last week.

- Solar Inverter stand alone in terms of data output.
- Main electricity meter can't determine import versus export power.
- Unable to determine consumption.

One Arduino w/Wifi later......scrape the live generation data from the Inverter web interface.......fit a new main electricity meter with modbus connectivity, do some calcs, publish to PVOutput.org and I now have all the centralized data/graphs I need.



Ian.

[stuner]

Nice video, Dave, including some interesting mental exercise

Actually there is another option Dave didn't look at.
The problem is each Solar Generator see's the other's power generation.

Lets say both systems generate the maximum power output, and the house is using minimal/no power.
The Solar Analytics system sees the +3kW from it's connected Solar Generator, but also sees -5kW 'consumption' the other system.
The Enphase Envoy system sees the +5kW from it's connected Solar Generator, but also sees -3kW 'consumption' the other system.

So what is needed is to add the power generated by the other system in both cases.

If you stuff a second wire into each Consumption Current Transformer (CT), that current thru that wire can be added/subtracted from the total.
So run the 3kW wire backwards thru the Consumption CT for the 5kW system.
Run the 5kW wire backwards thru the Consumption CT for the 3kW system.
Reality probably sucks in that you possibly cannot fit a second wire into some of those Current Clamps, or the wires would need to be made longer.

Or just go nuts, and rewire that big fat Active wire so after the main breaker it goes to the Solar system, then back to the loads.

I was thinking in a similar direction. If you want to measure the production of the individual systems, that's also possible. Instead of adding the current from the other solar generator to the local generator current, you could subtract it from the grid current. That way you should also get the correct value for the power consumption on one or both measurement devices. You can do this by running the wires through the transformer in the other direction, but I think there's an even simpler solution: You can just switch the direction (switch the two cables) on the measurement device clamps and move it to the other measurement channel.

[EEVblog]

Maybe I'm pointing out the obvious here, but instead of paralleling current transformers, Dave could have run both (production) wires through the same transformer core. Magnetic fields add and voila. This way he could keep using the extra transformer for his A/C.

Wasn't physically possible unless the wires were extended.
This was not discovered until after installation. But yes, could have done that before installation. The parallel transformers was an easy fix.

[PSR B1257]

I wonder, if this is really that simple.
Do you have confirmed, that they give you the correct sum?
What would stop the secondary current of CT1 to flow (partially at least) into CT2 (or vice versa) instead of the measuring device. Typically they have very low input impedance. 

[floobydust]

CT's have poor bandwidth, not suitable with PWM due to their iron cores. Most I've seen roll off past a few kHz so I wonder what accuracy is going to be like.
LEM or Hall-based transducer might be interesting to scope the current waveforms.

[EEVblog]

I wonder, if this is really that simple.
Do you have confirmed, that they give you the correct sum?

To a near enough order. I turn on the known 2.2kW EV charger and the figure goes up by 2.2kW

[PSR B1257]

Who knew...

[kaz911]

I am amazed at the rats nest.....

anyway - to make a future safe panel - am I correct in saying "split it" so mains input / output - then to a "feeder panel" (Solar, Battery, Generator) - then to the "consumption panel"

Only issue there would then be power for battery charging would not show in "consumption"

So what would be the ideal setup for multiple input source(s) + battery.. ? incl. ability to run off-grid.

Rather relevant for me as I'm designing my new house specs at the moment. and I'm designing for redundancy and brownouts.

[EEVblog]

I am amazed at the rats nest.....
anyway - to make a future safe panel - am I correct in saying "split it" so mains input / output - then to a "feeder panel" (Solar, Battery, Generator) - then to the "consumption panel"
Only issue there would then be power for battery charging would not show in "consumption"
So what would be the ideal setup for multiple input source(s) + battery.. ? incl. ability to run off-grid.
Rather relevant for me as I'm designing my new house specs at the moment. and I'm designing for redundancy and brownouts.

The only issue is the consumption measurement. You have to ensure that there is place to put that before the load fuses but after the batteries and grid connections and fuses.

[EEVblog]

BTW I decided to modify the panel to make the Enphase system only measure its own production instead of both.
This requird me to feed the Sunnyboy production through the enphase consumption current transformer to cancel it out.
Will probbaly do a quick video on it.
Seemed kinda dumb to have both system measuring the total. So now only the Solar Anlytics system monitors the total Production and Consumption. And that's a good thing because it's resolution is way better than the Enphase system.