Newbie. Advice sought regarding most suitable equipment to match 3 phase setup in UK

Tags: #<Tag:0x00007f1fe9b49258>

(Tom Hall) #1

Hello all and thanks for taking the time to read my post. Apologies, if I’m asking questions covered elsewhere – but I think my setup is at least a little different to the norm to warrant a new post. Feel free to hurl abuse at me if not.

We have just had a domestic PV system installed and this has fuelled an immediate obsession with power usage.

As you’ll know there are various third party tools available but I wanted a DIY solution which a) I understood instead of just a mysterious box, and b) I had control over, particularly the generated data. When I arrived at this site today after many hours of fruitless searching I concluded I had likely found what I needed.

However, that last thing I want to do is rush into buying the wrong (or least appropriate) kit.

Thus, I guess I need your help in answering a couple of fundamental questions.

  1. Can I achieve what I need with the hardware/software detailed here?

  2. If so, which hardware solution is most appropriate?

OK, in no particular order some basics about the setup we have here and what I’m hoping to achieve.

A. I mention we have just installed PV. (A battery will be added in the future but for now it’s just panels and inverter.) Whilst a sophisticated solution which incorporated the solar into the whole view might be a ‘nice-if’ it isn’t essential. The three phase inverter (Solar Edge) has logging and associated portal built in. So that data will be covered.

B. My primary interest is in what energy is still being pulled from the grid and when. Even though this is just a domestic property we have a three phase supply. The three phase inverter splits a third of the generated PV power to each phase. So I want to monitor the draw from the grid on each phase so I can review which takes the most load and over what period.

C. Absolute precision is not essential, e.g. I wouldn’t need to try and estimate cost to the nearest penny. Instead, I just need broad accuracy to display usage of each phase.

D. I assume I need three CT sensors one for each phase cable? (I can see there is a fourth separate neutral cable but I assume I don’t need to monitor this?)

E. As I understand it (and excuse my ignorance here) but the CT sensors can’t understand flow direction? Thus, if the cable is periodically drawing current from the grid, then getting current pushed back by the PV the sensor can’t apply positive/negative of flow so just gives the quantity regardless of direction.

If this is the case do I need to “exit sub” at this point since it’s not possible to capture the data I need?

Any help and insights at this stage would be very welcome.

Cheers, Tom.

(Robert Wall) #2

Welcome, Tom.

I hope we don’t disappoint! There’s quite a lot of accumulated knowledge here, so feel free to tap into that whenever you wish.

I take it you’ve been looking at the emonTx?
There is an “approximate” sketch (Arduio-speak for program) that will record and display 3-phase power and energy. “Approximate” because the emonTx has only one voltage input so it can monitor only the voltage of one phase, so it must assume the other two are the same.

Statistically, out of the box, we should be able to get to within a couple of percent. With the proper instruments and careful calibration, you could probably improve a little on that.
But bear in mind, your tariff meter records only import, and your generation meter only records generation, so there will be no direct comparison with what I think you’re asking for at the moment.

That’s correct. The neutral current is actually the vector sum of the three phase currents, and in an ideal set-up, with all three phases perfectly balanced, it is zero. Knowing it is not going to give you any information you don’t have from the other three.

On their own, the c.t’s can’t understand direction. But when you monitor the voltage as well, it’s possible to determine the direction of power flow. Our convention is imported power is positive, exported power is negative. There’s an article in the ‘Learn’ section that explains this: Learn→Electricity Monitoring→AC Power Theory→Introduction→An Introduction to AC Power

Just wondering - what do you do for hot water? If you have an immersion heater and a hot water tank, you might be interested in my colleague Robin Emley’s PV Router, which will divert surplus PV energy to heating water. It was originally developed and published here, and there’s a lot in ‘Learn’ about it.

Assuming you only want to monitor your overall house consumption / export, and not what your PV is doing in isolation, then I think you want:
1-off emonTx,
1-off a.c. adapter,
3-off 100 A c.t.
1-off emonBase,
1-off 5V USB power supply & cable,
1-off Programmer & cable.
That will give you monitoring of only the nett power to/from the grid, and a record (stored locally on the emonBase - a Raspberry Pi) of that, which you access via a computer or smart phone.

That’s all you ask for, but before you part with money, do think about what you might want to monitor in the future, because you might want to expand the system. e.g. Does your generation meter have a means to give you its data, or might you want to put c.t’s on the PV infeed to measure that? Might you want to monitor individual circuits?

(Brian Orpin) #3

The EmonBase is a simpler approach, but depending on where things are, and whether you want any other RF based devices (such as the EmonTH), there are 2 other options:

  1. Just buy the RFM board if you already have a spare Pi.
  2. The EmonESP adapter, effectively connecting your EmonTX directly to the WiFi and run EmonCMS on a separate machine you might already have. If you are really adventurous, an ESP8266 from China or even Amazon can also reduce the cost (just needs a bit more work!).


(Tom Hall) #4

Thanks Robert, I really appreciate the swift and detailed response.

A couple of clarification questions/comments if I may?

Regarding the CT sensors and the ‘direction’ of flow, I understand, I think, the way the voltage can be used to generate a positive/negative value and thus know ‘direction’. (And when I say ‘understand’ it’s very much to the level of someone who was staring out of the window during double physics instead of paying close attention.) However, what I’m not clear on is how the voltage is determined? It seems as though the only data being collected would be that generated by the three CT sensors on each phase cable? So (hopefully without sounding too imbecilic) where does the voltage reading come from?

Regarding the sketch for a 3-phase system you mention is this in a format that will just work or will it require tinkering with? (I can programme to a certain extent but these days it’s always environments like Microsoft .NET/MSSQL so whilst I’m sure I can get my teeth into it there might be a learning curve. That said, when I started in IT it was on SCO Xenix so command lines should hold no fear :smiley: )

The hot water link is very interesting. We had been looking at third party (nicely boxed) solutions to do exactly this. Since our three phase inverter spreads the PV generation equally to the three phases there is likely one doing very little for a lot of the time so this type of approach would be sensible.

Your final question regarding future proofing is a good one, but it rather underlines the fact “I don’t know what I don’t know”. Would the hardware you’d listed (which mirrors what I’d thought myself with the exception of the Programmer) have any scope to support expansion? I doubt I’d need to monitor the PV in isolation since the inverter already supplies this data.

Thanks again taking the time to respond. Cheers, Tom :slight_smile:

(Tom Hall) #5

Hello, thanks for this. Looks interesting. I don’t have an existing Pi and I wasn’t thinking of additional data such as temperature. But who knows - I only found this site at lunchtime ~:D

Many thanks for info.

Cheers, Tom.

(Robert Wall) #6

That’s what the voltage input of the emonTx and the a.c adapter is for. It’s there primarily to measure the voltage. It can also be a power supply for the emonTx.

It will need a modicum of work - it’s a multi-purpose multi-choice beast, so you need to edit various configuration options to set it up how you need it. That’s why you needed the programmer - to load it and talk to it once it’s running. You’ll need to set up your computer with the Arduino IDE, but there are full instructions to do that, and you drive it through the IDE. (The inbuilt editor leaves much to be desired, but you can use another one - I use NP++.)

And I don’t know what you don’t know :wink:. But I felt I ought to mention it early enough. :grin:

Put it this way, it might turn out to be not the most economical approach. The competition for emonBase is the emonPi. That offers two channels of monitoring only, so no good for your three phases, but alternatively it’s two channels in addition to your emonTx. So if you think you might need only two more channels (on one phase - it can’t do two out of three phases) then I’d suggest that. But if you might need more, stay with the emonBase and multiple emonTx’s.

(Or, as Brian says/infers, an emonBase is a Raspberry Pi with an piggy-back radio to receive from the emonTx. An emonPi is that plus energy monitoring inputs and a small LCD display, in a case.)

Both emonBase and emonPi can receive data from multiple sources, emonTx’s or emonTHs, so in that respect, they are equal.

(Tom Hall) #7

I can feel a purchase coming on. Need to act swiftly before the wife can exercise her veto.

Thanks again. To quote that Schwarzenegger fellow. I’ll be back.

Cheers, Tom.

(Robin Emley) #8

TH: The hot water link is very interesting. We had been looking at third party (nicely boxed) solutions to do exactly this. Since our three phase inverter spreads the PV generation equally to the three phases there is likely one doing very little for a lot of the time so this type of approach would be sensible.

In the UK, a 3-phase meter should only register consumption after the energy flows on all three phases have been added together. So you can generate surplus power on one phase and consume a similar amount on a different phase without being charged.

When diverting surplus power on this basis, it is important that the energy flows are measured accurately on each of the phases using three sets of V*I measurements. That’s how the 3-phase version of my Mk2 PV Router design works, more details at

(Tom Hall) #9

Many thanks for this Robin. I’ll take a look.

Cheers, Tom.