To earth the metalwork that supports the panels? I would say yes, if there’s a possibility that anyone can come into contact with them and another good earth at the same time. The thinking there is, if you’re isolated from earth, touching metalwork that’s live or has picked up a static charge will not be as serious as it would be if another part of your body is earthed. The reason electricians have fibreglass stepladders follows the same thinking.
I suspect we’re looking at a second or third fault before the frame does become live, so it’s most likely an extremely remote possibility, which is why an earth is deemed not necessary.
Thanks. Could the earth connection be to the consumer unit earthing point which would be very easy or should it be to a ground spike which would not be easy as the sheds are on concrete bases that extend away from the sheds?
I’d go for your main earth at the C.U. every time. But you could put a spike in temporarily and check for voltage between that and the main earth. If there is any, you’ll probably wish you hadn’t looked, as it would indicate that there’s a significant current flowing in the ground and there might be a problem somewhere.
So, back to my original question - what hardware do I need to measure the power generated from the PV panels? The inverter is a INVT iMars MG4K6TL-2M. The two panel areas will be connected to two separate MPPTs on the inverter, which in turn will provide single-phase 240VAC 50Hz. I guess terminated in a 20A fuse.
The inverter will allegedly provide “values” both on the DC and AC side via a Wifi connection and some cloud based service (alternatively wired, local ModBus connection). This has made me determined not to do anything more on the DC side.
Referring to the diagrams below “Sensor Installation” - how are these translated to my system?
I have a three-phase mains fuse, and the individual user fuses are obviously single-phase fuses. I already measure these fuses using the IotaWatt (with three CTs on the main fuse), storing data to my EmonCms database on my RPi3.
Do I have a Type 1 or 2 system?
Is CT2 in the diagram a new CT measuring on the new AC wire from the inverter?
What then is CT1?
I do not have any spare CTs on the IotaWatt, so I would need some new gear to connect the new CT (or CTs?). Also, I would like to measure temperature, and the optical sensor looks tempting for redundancy. Is the emonPi PV bundle what I need?
The difference is where you can put the c.t’s. Type 1 measures the PV infeed and the nett Grid power. Type 2 is appropriate where you know (or can assume) the direction of power flow, so it measures the PV infeed and the house consumption.
I think you have a Type 1, but you need to confirm that by finding where the PV infeed connects into the system. If your Ioatwatt c.t’s are on the supply side of that point, you indeed have a Type 1.
Unless you already have a c.t. there, then yes, it is.
That being the case, and in part because I have no knowledge of the Ioatwatt and few people here have a detailed understanding of it and its capabilities, I think that would be a wise decision.
One warning about the optical sensor - what the pulses it detects indicate depends on how your meter is configured. It’s likely to be nett imported energy only, something you should check.
Thanks!
The Iotawatt measures reference voltage and takes 14 CTs as input. It does not take any other types of inputs. It communicates over wifi. So I am thinking it must be in part the same as an emonpi, but with more CT inputs and without features like pulse or temperature.
I think I’ll go for the emonpi. The reason why I asked about CT1 and CT2 in the illustration is: Does it take more to measure the PV power than measure the AC + reference voltage? Yes, I see that a standalone Solar PV Emonpi setup also measures power on house usage. But I already have that covered by my Iotawatt. If it is so that the emoncms database can take data from multiple emon units (is it?), I guess I will have all the feeds I need to get the full picture, right?
I am not sure what you mean by that. Measuring PV power and measuring the Grid power are essentially the same thing. If you only need the emonPi to give a measure of the PV infeed, then you have a second channel free to measure something else connected to the same phase.
It is, and from Iotatwatt (as I understand it), and indeed from any source provided that the data can be sent in an acceptable format.
@OlavKristiansen
Your solar arrays will have an open-circuit voltage rating. What is it?
So I don’t raise your hopes, I’ll say it’s likely there is nothing here to meet your DC measuring wishes.
Without the basic information of system voltages, can’t help much really.
Don’t try measure using a multimeter. Sounds like you could have big enough arrays to kill you if you make a mistake!
@Robert.Wall
My explanation of what I will to was confusing, sorry. But you got my point.
I am curious if it is as simple as I think. Lets assume i have 3kW PV production, measured by the CT on the EmonPi, and the CT shows a positive number. In July, typical power consumption during day time is <<3kW. Let’s assume my main fuse (without PVs in operation, like last summer) is measured to 0.5kW consumption. Putting the PVs in operation would then give -2.5kW consumption on the main fuse. Are the CTs bi-directional, i.e. will the negative amp measurement on the main fuse wire (the same phase as the PV is connected to) measure accurately my household’s input to the external grid?
On their own, no. But when you combine that with the voltage reading in the emonPi (it MUST be the voltage reading from the emonPi, not one from the iotatwatt) then you will get a change of sign depending on whether the power is imported or exported. See Learn→Electricity Monitoring→AC Power Theory→Introduction→An Introduction to AC Power
