I suspect my issue may be the clamp placement, but if anyone has had the same issue or can give some pointers it would be helpful to save time troubleshooting.
I have an emonpi with CT1 inside the consumer unit on meter tail in (fitted by electrician yesterday when PV commissioned) and CT2 on the tail from from the new sub-board the PV inverter is fed through. The electrician did not put in a henley block as they planned as the tails come out from inside the wall and he was not sure he had enough length (or it looked like a lot of hassle) - so the tails from sub-board go into the back of the main breaker on the consumer unit - but basically its like the "Type 1" diagram on the emonpi page. I am also using AC-AC adapter.
During darkness the house reading seems accurate, general reading and increases/decreases with load changes (follows my older current cost/perl/rrd setup). When the PV starts to output though I seem to get an accurate PV reading (~200w on grey xmas eve) but the house load goes from 200-300w down to double digits like 19w - and I even saw -17w earlier.
I am not that awake yet today but the behavior is a bit strange as the reading is accurate on CT1 when the PV is not generating and then inaccurate when it is.
Any suggestions? I will try to re-position clamps a bit later but the behavior on house and the negative reading seems strange and I am wondering if the clamps are directional (they are marked with arrows, and were installed in direction of current flow - i.e. from tails to house, and from inverter to supply).
Running "Powered by openenergymonitor.org | low-write 9.2 | 2015.12.10"
Re: Erratic Consumption readings during solar generation (new install)
It sounds like CT1 is measuring the net power flow into / out-of the property. At night, that's the same as your consumption, but by day it's your consumption minus your PV production, which can easily go negative on a sunny day when you're not consuming much.
Re: Erratic Consumption readings during solar generation (new install)
From your description, I think dBC is right and CT1 is on the grid side of where the PV infeed joins the house wiring. CT1 will then be measuring the nett power flow to or from the grid, so on a good sunny day with little or no power being used in the house, you should see -3 kW or so.
Re: Erratic Consumption readings during solar generation (new install)
No CT1 is between the grid and the consumer unit only, its on the main cable inside the CU. CT2 is on the sub-board from the PV.
I suspected CT1 was backwards, so I gingerly teased it out of the CU and swapped it (arrows were the wrong way) but now it reliably reads negative of expected reading so I am a bit lost. I put CT1 on the solar with CT2 and it reads (well +/- 10w) the same so clamp seems ok - all a bit confusing.
Arrows on the CT go in direction of flow right? So from inverter to grid, and from grid to CU?
Re: Erratic Consumption readings during solar generation (new install)
The way you originally described the wiring, I understood that you have a pair of tails coming from the meter into the main switch of your consumer unit. Also going into the live side of that main switch is another pair of tails (or a cable) that end up at your inverter via a sub-board.
Is (was) your CT1 on one of the tails between the meter and the main switch, i.e. is the point of connection to the inverter on the meter side of CT1 or on the consumer unit side?
If CT2 is on the PV feed, it will see only the PV generation (with a caveat, below).
If CT1 is on the meter side of the common connection, then it will see the nett power and energy flow (forget the sign for the moment). Any generated power that is consumed by the house will bypass it. When you get the sign right as per our convention, CT1 will show your nett power, CT2 will show the PV generation, and your house consumption is CT1 + CT2
If CT1 is on the consumer unit side of the common connection (and if I've understood your wiring correctly, this is not possible), then it will see only the house consumption. It will have no knowledge of whether part or all of that comes from the grid or from your PV. It will only see the part of your generated power that the house consumes, the remained that is exported to the grid will bypass it. In this case CT1 will show your house consumption, CT2 will show the PV generation as always, and the nett power flow from/to the grid is CT1 - CT2.
Our convention is that imported power and energy is positive, and generation too is positive. What determines the sign of the power is the relative phasing of both the ac adapter and the current transformer in relation to the voltage and current waveforms respectively. Turning the CT over on the cable won't affect anything other than the sign. My CTs don't have arrows, this http://openenergymonitor.org/emon/Current_Transformer_Installation tells you how to install the CT to honour our convention.
The caveat: Most inverters consume a few watts of "keep alive" power during darkness hours, but the power level is normally below that which we can reliably measure. So it may indicate a positive value when it really should be negative, i.e. your inverter is consuming power.
Re: Erratic Consumption readings during solar generation (new install)
Right, thanks for all the replies, especially for Robert for taking time to explain again in detail, sorry to go AWOL family stuff overrided tinkering.
This has been vexing me for some time... and after much head scratching I just started drawing a diagram and realised he has the clamp on the wrong cable! It is indeed measuring between utility meter and everything, not between junction and house board only, this explains exactly what is happening. Moral of the story - never let a qualified electrician install something and assume they know what they are doing!
So basically referring to the installation guide, I have ended up with a "Type 2" setup as the diagram shows setup: http://openenergymonitor.org/emon/modules/emonpi#ct_sensors
I did not see the CT installation page: http://openenergymonitor.org/emon/Current_Transformer_Installation although now I can see its linked from the installation guide - genuine comment that link could be more prominent in the install guide - I did my own tests as I could not find the details in the install guide.
This does explain why my "usage" does drop to 0 when generating, as we are exporting or close to 0 so net flow is minimal - what further confused the matter is that the hot water immersion heater is aiming to get close to 0 as possible - so the reading kept moving around. See attached chart from today (just over 5Kwh today - not bad for December I thought!) to see what I mean.
Question now is do I move the clamp onto the right cable inside the board, or do I just reconfigure emoncms inputs - I think I will look at moving the clamp as in its current location its very hard to remove so there is a risk of it being in place and unplugged which I believe is a no-no as even with the diode in the clamps form the shop there is a risk of damage to the CT (or yourself).
Re: Erratic Consumption readings during solar generation (new install)
"never let a qualified electrician install something and assume they know what they are doing"
Hmm. We engineers have a name for those people, and it's not complimentary. I don't have a lot of confidence in their trade's supervisory system, I can think of several installations or alterations I've had cause to look at, carried out by registered, or allegedly registered, firms or individuals, that had defects - one quite serious. And a very serious defect was put on my neighbour's installation by a bathroom fitter. He broke the seal on the supply authority's main fuse, inserted Henley blocks and in doing so managed to swap line and neutral upstream of the consumer unit. Fortunately for the neighbour, I found it before anything serious did happen. How serious could it have been? The only protection against a line-earth fault on any wiring including the lighting circuits was the 80 A supply authority's fuse.
<Rant over>
Personally, I'd leave the CT where it is, because there you'll get a much more accurate measure of your nett energy flow to/from the grid, and if at some point you want to add MartinR's or Robin's diverter to the system, it's in the right place.
Re: Erratic Consumption readings during solar generation (new install)
Yes, in the end I left it on the main incoming tail and reconfigured emoncms. The electricians were ok but as usual job was done to time and effort rather than best results. They clipped the 4mm to the woodwork where it comes down through the airing cupboard rather than using conduit (no mechanical protection from airing shelves against it) which I might address later and the conduit and the way they used the knock outs in the new sub board and isolator were not the neatest - but I am very OCD with my cable routing I concede!
All working now, I was going to mount the emon in a plastic box with the temp sensor breakout but just ordering an Alu case from the shop now... should have done that in the first place!
I was not aware of the diverters with emon when I did the solar order back in October, so I have an iBoost+ fitted I took the option on at the time which works to achieve the same goal - emon will have temprature sensors for top and bottom of the tank and I might yet add a sender to record the power put into the immersion heaters (big tank, have two immersions).