3 Phase with Solar input on 1 Phase - Reference Voltage Recomendaetions

I am looking at putting in a number of Iotawatt units on my property. In part of this I have a 3 Phase supply servicing multiple buildings. On three of these buildings I have individual solar single phase solar systems each feeding into one phase. So in the end, on the one three phase supply, there is a Solar input on each phase, but each of these input occurs in a different building.

To try and spell this out some:
Building 1 is three phase, it has a single phase 6.6kW solar system on Phase A,
Building 2 in three phase, it has a single phase 6.6.kW solar system on Phase B,
Building 3 is single phase, it has a single phase 6.6kW solar system on Phase C,
Building 4 is three phase (on same supply) with no solar system.
Building 5 is single phase (on same supply) with no solar system.

My main question in scoping the iotawatt configuration is whether to go with a single VT for reference (in the three phase buildings) or should I use three (one per phase). Since in a building the Solar feeds into one phase, and the other two phases don’t have solar (in that building), but do have solar from other buildings coming together at the main distribution board. Is this likely to cause accuracy issues in measurement if I only have one VT for reference in each building? If so what is the likely magnitude of the inaccuracy?

I know this is all a bit confusing when spelled out in words, so I can put together a schematic diagram if it helps to understand the situation.

Overall, I am keen to know detailed usage by circuit in each of the buildings, and be able to separate this from the solar generation data appropriately. Each Solar has SolarEdge Inverter, with its own CTs for all phases, plus the main distribution board has a SolarEdge gateway for feed in / grid use, but this only gives me part of the energy consumption picture.

I think what I need is an iotawatt in each building with CTs for building supply, Solar input, and then for each appropriate circuit I want to measure. I am basically planning on only the single VT and a power supply, but not sure on the accuracy impact of this.

Then, somehow I then need to bring all this together (which I undestand can be done a few ways), and blend this with the other solar and usage data from my main house and use all of this to optimise energy use, via Home Assistant, and maximise solar feed-in and where it is viable.

What kind of accuracy are you looking for? While I’ve not done any kind of direct survey, and don’t collect any metrics on users, I feel fairly confident in my sense that more than 90% of three-phase users use the “derived reference” method (one VT) with satisfactory results. That said, your mileage may vary.

Measuring the dispersed solar generation is not a problem. Measuring the individual circuits in each building, within the inputs limitation of each unit, is also not a problem.

Your greatest challenge will be aggregating the information into one data-store and developing tools to analyze it. Both influxDB (1.0 or 2.0) as well as Emoncms are capable of doing that, but will require some considerable effort to setup.

I’d be happy with anything better than a 5% error I think, since most of the analysis and use of the data will be in a relative sense.

In some buildings that operate as rentals, it may come to pass that I want to oncharge electricity usage, so in that scenario (overall building use of grid power) 2% would be better. It’s this bit where (to my non electrical engineering brain) distributed solar might complicate the measurement or calculations.

To keep the calculations simple it would probably be a lot easier if I just focused on on charging overall use in a building rather than grid use.

In terms of measurement it was the impact of the distributed solar on the VT that I was unsure about. With the solar feed going in on the load side of a single phase in each building. Would it mess with the voltage or phasing in some way that would make the measurements meaningless?

Given all the buildings are on the one supply, If each building has an iotawatt, would it help (or hinder) to have the reference VT on different phases so that across the overall system there was at least one reference VT on each phase? I doubt this will help since each iotawatt is making its own calculations/measurement in relation to only the VT(s) it sees directly I would guess?

You didn’t mention billing. I would be remiss if I didn’t point out that IoTaWatt is not a revenue meter and not certified for trade.

That said, if your standard is 5% accuracy, IoTaWatt will easily do that, and the question of voltage reference should have no bearing on achieving it

If you measure the mains and solar feed in each building, you will be able to get overall usage, net usage, energy self-consumed, energy exported, energy imported and energy generated.

For all intents and purposes, the voltage on each phase should be consistent across your network. You may see some local variation if the distances between buildings is far and/or the wiring is undersized. Inverters raise the voltage. They must do that in order to export surplus power. That’s probably one of the reasons why they were installed on different phases.

1 Like

Thanks. I’ve placed an order for 5 Aussie bundles. Locally Jaycar are out of the USB power supplies. Can any 5V 1Amp USB supply be used or does it need to be one from a supported list. Assume from the listed Jaycar one, the cable needs to be mini B on the iotawatt input end?