There are devices that detect when your PV panels are exporting power to the grid - so that you can divert it to heat water or a thermal store, for example.
These rely on a sensor that clips around one of the conductors at the grid connection.
Can anybody explain to me, in simple terms, how this is achieved with AC? I can understand measuring the magnitude, but not the direction.
I'm sure it's obvious, but I may have dissolved the necessary part of my brain over the years.
awneil 
These rely on a sensor that clips around one of the conductors at the grid connection
Do they?
Doesn't there have to be some sort of "system controller" - so that would know whether you're exporting or importing ... ?
I can understand measuring the magnitude, but not the direction.
A clamp on current probe, is a form of a transformer....so it monitors that AC current is flowing. By definition, AC is bidirectional. So if you look at the polarity of the sensor output signal, you can tell what the direction is at that moment.
Also, if you can measure the line voltage at the same time, you can calculate the present power, which will have a sing (of direction). So over many cycles you will probably have more power flowing in one direction than the other (towards the load).
awneil
So if you look at the polarity of the sensor output signal, you can tell what the direction is at that moment.
Indeed - but that doesn't help.
The question is how to detect whether there's a net flow of power (or energy) from the grid or to the grid ("export") over time.
Does measuring both the voltage & the current polarity simultaneously tell you ... ?
EDIT
looks like it should:
https://www.researchgate.net/figure/Cyclic-ac-power-component-phases-over-4-quadrants_fig3_353327351
https://openenergymonitor.org/forum-archive/node/10221.html
As I understand it, there is only a clip/clamp on sensor at the grid connection. No voltage measurement there.
The control box that runs the immersion heater would have access to the voltage, I suppose. Would that help?
awneil isn't there a control box that decides whether the PV output gets used locally, or exported ?
Yes, but that's not how the solar diverters work. They work by monitoring the big cables in the meter box. No connection with the grid tied inverter.
One manufacturer also scavenges enough power from the clamp sensor to self power.
gchapman 
No connection with the grid tied inverter.
AN1444 Grid-Connected Solar Microinverter Reference Design
[page 4, right]
This may help: https://web.archive.org/web/2017...
The transformer has a polarity that depends on the direction the power is flowing through the primary "winding", this can be used to show flow in or out!
Jim
The transformer has a polarity that depends on the direction the power is flowing through the primary "winding", this can be used to show flow in or out!
That statement is really making my head hurt. Surely a transformer, and more specifically a current transformer, isn't really concerned with power flow but with current flow.? And that current flow will reverse 50 or 60 times a second? So the detected direction of flow will reverse as well?
The transformer has a polarity that depends on the direction the power is flowing
You do this with the phase angle instead. Voltage and current do have a phase between them, and this phase can be detected.
Remember a lesson in the 3rd semester- how to synchronize a generator (or any AC-source) to a network:
1. the same frequency 2. the same voltage and 3. the same voltage phase 4. ZAP! you can activate a (3-section/phase) relay/switch and you are in.
Now, the most important: advance a Voltage phase a little so that you can start delivering a current.
It is all phase-related in AC.
ki0bk wrote:The transformer has a polarity that depends on the direction the power is flowing through the primary "winding", this can be used to show flow in or out!
That statement is really making my head hurt. Surely a transformer, and more specifically a current transformer, isn't really concerned with power flow but with current flow.? And that current flow will reverse 50 or 60 times a second? So the detected direction of flow will reverse as well?
It made my head hurt too, but I wasn't prepared to argue from a standpoint of extreme ignorance. So I ignored it and went on my merry way.
from a standpoint of extreme ignorance.
Sorry, I was stupid to not believe you were smart enough to get my point, and sorry I did not explain well enough to make my point clear...
Maybe this will help: https://learn.openenergymonitor....
And yes the current does move back and forth 50-60 Hz, but when compared with the voltage wave, the phase angle will tell you the power flow direction....
How could I be so stupid, Sorry guys, I'll try better next time. 
Jim
John_A_Brown wrote:from a standpoint of extreme ignorance.Sorry, I was stupid to not believe you were smart enough to get my point, and sorry I did not explain well enough to make my point clear...
Maybe this will help: https://learn.openenergymonitor....
And yes the current does move back and forth 50-60 Hz, but when compared with the voltage wave, the phase angle will tell you the power flow direction....
How could I be so stupid, Sorry guys, I'll try better next time.
Jim
Hey! That's the same link I posted!
But I'm not sure why I deserve the heavy handed sarcasm - I am extremely ignorant about AC theory. I don't have a degree, like most of you here. I don't even have an A level(or whatever the Over The Rainbow equivalent is. I was lucky enough to make a living of sorts writing embedded firmware for 45 years, owing to getting in near the ground floor and a few right-place-at-the-right-time experiences, and now I'm three quarters retired.
